Textbook of Medicine (MRCP Study Guides)

Edited by R

L

Souhami

MD FRCP FMedSci

Emeritus Professor of Medicine, University College London, UK J Moxham MD FRCP Professor of Respiratory Medicine and Vice-Dean, Guy's King's and St Thomas' School of Medicine, King's College Hospital, London, UK

CHURCHILL LIVINGSTONE

EDINBURGH LONDON NEW YORK OXFORD PHILADELPHIA ST LOUIS SYDNEY TORONTO 2002

CHURCHILL LIVINGSTONE An imprint of Elsevier Limited © Pearson Professional Limited 1996 Assigned to Harcourt Brace and Company 1998 © 2002, Elsevier Science Limited. All rights reserved. © 2004, Elsevier Limited. All rights reserved.

Commissioning Editor: Laurence Hunter Project Development Manager: Barbara Simmons Project Manager: Nancy Arnott Designer: Erik Bigland

The right of Robert Souhami and John Moxham to be identified as editors of this work has been asserted by them in accordance with the Copyright, Designs and Patents Act 1988. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without either the prior permission of the publishers or a licence permitting restricted copying in the United Kingdom issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London WIT 4LP. Permissions may be sought directly from Elsevier's Health Sciences Rights Department in Philadelphia, USA: phone: (+1) 215 238 7869, fax: (+1) 215 238 2239, e-mail: [email protected]. You may also complete your request on-line via the Elsevier Science homepage (http://www.elsevier.com), by selecting 'Customer Support' and then 'Obtaining Permission First edition 1990 Second edition 1994 Third edition 1997 Fourth edition 2002 Reprinted 2004 ISBN 0-443-06464-4

British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library

Library of Congress Cataloging in Publication Data A catalog record for this book is available from the Library of Congress

Note Medical knowledge is constantly changing. As new information becomes available, changes in treatment, procedures, equipment and the use of drugs become necessary. The authors and the publishers have taken care to ensure that the information given in this text is accurate and up to date. However, readers are strongly advised to confirm that the information, especially with regard to drug usage, complies with the latest legislation and standards of practice. your source for books, journals and multimedia in the health sciences www.elsevier hea lth.com

ELSEVIER

The publisher's policy is to use paper manufactured from sustainable forest!

Printed in China

Preface The Fourth Edition of this textbook continues the approach, first begun fifteen years ago, to provide a comprehensive account of internal medicine that incorporates the ideas of modern medical science. The intervening years have strengthened our belief that an understanding of cellular and physiological processes makes medicine more exciting and less arbitrary. The emphasis of the book is clinical, giving a full description of the presentation, investigation and treatment of disease. We have attempted to interpret these clinical aspects by reference to the underlying mechanisms whenever possible. Modern integrated curricula demand this approach. The textbook is primarily intended for students of medicine, but we hope and anticipate that it will continue to be of value to trainee doctors and health care professionals in many fields. The Fourth Edition has been completely revised to incorporate all major developments and advances, leading to an expansion of many sections especially in the areas of infectious disease including AIDS, and cardiovascular disease. Many of the illustrations have been replaced or extended to include the most up-to-date epidemiological data. We hope that these will give the reader a sense of the way medicine is developing and where future advances in practice may lead. The most important innovation, however, affects the entire book. By linking the book to an easily-accessed internet site we

have been able to incorporate case problems, multiple choice questions and clinical images throughout the text. The illustrative case problems, some of which are in the textbook itself, are designed to allow the reader to test his or her knowledge by practical application in clinical problem solving. Knowledge and understanding can also be assessed by multiple choice questions, formerly in a companion volume but which are now located on the associated internet site. In addition the website contains 500 supplementary clinical images. This extra information, all of which is clearly cross-referenced in the text, represents for the reader an important additional compendium of knowledge, which expands the versatility of the book as a learning aid. The range and depth of specialised medical knowledge necessary for a major textbook has led us to select our authors from a wide range of medical schools. They have been chosen for their acknowledged reputation in clinical and scientific medicine. In this Fourth Edition we have again expanded the team of authors. All of us would be pleased to receive suggestions which will help the book change and develop for future editions. London 2002

RLS JM

V

Acknowledgements As in previous editions we are greatly indebted to Laurence Hunter for his help in the planning and realisation of the new edition. He has a very effective production team that have made our task much easier. Our editor Barbara Simmons has, as usual, been helpful, efficient and outstandingly courteous even in the face of occasionally rather provoking circumstances. We owe her a lot and are very grateful.

vi

The multiple choice questions and answers for the website have been developed by Dr Andrew Freedman, Dr Richard Watts and Dr Richard Evans. We are very grateful to them for their painstaking work, which has added greatly to the educational value of the book. We are also grateful to Dr Donncha O'Gradaigh for writing the rheumatology case studies.

Contributors Geoffrey J Bellingan PhD MRCP Senior Lecturer, Department of Medicine, University College London Medical School; Consultant, Department of Intensive Care Medicine, University College London Hospitals, London, UK Brian T Cooper BSC MD FRCP Consultant Gastroenterologist, City Hospital, Birmingham; Senior Clinical Lecturer in Medicine, University of Birmingham, UK John Costello MD FRCP FRCPI Consultant Physician and Clinical Director of Medicine, King's College Hospital, London, UK

Jeremy Holmes MD MRCP FRCPsych Consultant Psychotherapist, Devon Partnership Trust; Senior Lecturer, University of Exeter, Exeter, UK Cameron T C Kennedy MA MB BChir FRCP Consultant Dermatologist and Clinical Senior Lecturer, Bristol Dermatology Centre, Bristol, UK Michael E J Lean MA MD FRCP Professor of Human Nutrition, University of Glasgow; Consultant Physician, Glasgow Royal Infirmary; Non-executive Director, Health Education Board for Scotland, Glasgow, UK David C Linch BA MB BChir FRCP FRCPath FMedSci

John Cunningham MD FRCP Professor of Renal and Metabolic Medicine, Barts and the London, Queen Mary's School of Medicine and Dentistry, University of London, UK J David M Edgar BSc(Hons) FRCP MRCPath Consultant Immunologist, Regional Immunology Service, The Royal Hospitals Belfast, N Ireland

Professor of Haematology, Royal Free and University College London Medical School, London, UK Richard G Long MD(Lond) FRCP (Lond) Consultant Gastroenterologist, City Hospital and Queen's Medical Centre, Nottingham; Clinical Teacher, Nottingham University Medical School, Nottingham, UK

Jeremy M Gibbs MD FRCP Consultant Neurologist and Honorary Senior Lecturer, Royal Free Hospital, London, UK

John P Monson MD FRCP Professor of Endocrinology, Department of Endocrinology, Barts and the London, Queen Mary's School of Medicine and Dentistry, London, UK

John Goldstone MD FRCA Consultant in Intensive Care Medicine and Consultant Anaesthetist, University College London Hospitals, London, UK

John Moxham MD FRCP Professor of Respiratory Medicine and Vice-Dean, Guy's King's and St Thomas' School of Medicine, King's College Hospital, London, UK

Robin A C Graham-Brown BSc MB FRCP Consultant and Honorary Senior Lecturer in Dermatology, University Hospital of Leicester NHS Trust and Leicester Royal Infirmary, Leicester, UK

John O'Grady MD FRCPI Consultant Hepatologist, King's College Hospital, London, UK

Brian Hazleman MA FRCP Consultant Rheumatologist, Addenbrooke's Hospital, Cambridge; Director Rheumatology Research Unit; Fellow, Corpus Christi College, Cambridge, UK

Peter W Overstall FRCP Consultant in Geriatric Medicine, Hereford Hospital, Hereford, UK

VII

Philip A Routledge MD FRCP FRCPE Professor of Clinical Pharmacology, University of Wales College of Medicine, Cardiff; Honorary Consultant Physician, Cardiff and Vale NHS Trust, Cardiff, UK

John Wilding DM FRCP Reader in Medicine and Honorary Consultant Physician, University of Liverpool and University Hospital Aintree, Liverpool, UK

John W Scadding MD FRCP Consultant Neurologist, The National Hospital for Neurology and Neurosurgery and Whittington Hospital, London, UK

Gareth Williams MA MD FRCP Professor of Medicine and Honorary Consultant Physician, Clinical Sciences Centre, University Hospital Aintree, Liverpool, UK

Herb F Sewell MB ChB MSC PHD FRCP FRCPath FMedSci Professor and Consultant Immunologist, University of Nottingham, Queen's Medical Centre, Nottingham, UK Martin Smith MBBS FRCA Consultant Neuroanaesthetist, Department of Neuroanaesthesia and Intensive Care, The National Hospital for Neurology and Neurosurgery, London, UK Robert L Souhami MD FRCP FMedSci Emeritus Professor of Medicine, University College London, UK L B Tan DPM FRCP FESC Consultant Cardiologist, Leeds General Infirmary, Leeds, UK J Malcolm Walker MD FRCP Consultant Cardiologist and Clinical Director of the Hatter Institute of Cardiovascular Studies, University College Hospitals, London, UK Mark H Wansbrough-Jones MSC FRCP Consultant Physician and Senior Lecturer in Infectious Diseases, St George's Hospital and Medical School, London, UK Antony P Weetman MD DSC FRCP Professor of Medicine, University of Sheffield; Honorary Consultant Physician, Sheffield Teaching Hospitals Trust, Sheffield, UK

viii

Ian G Williams BSC MRCP Senior Lecturer, Department of Sexually Transmitted Diseases, University College London Medical School, London, UK Robin M Winter FRCP FMedSci Professor of Dysmorphology and Clinical Genetics, Institute of Child Health, London, UK Stephen G Wright MB FRCP Consultant Physician, Hospital for Tropical Diseases, London; Honorary Senior Lecturer, London School of Hygiene and Tropical Medicine, London, UK

MULTIPLE CHOICE QUESTIONS Richard H Evans MRCP Specialist Registrar in Infectious Diseases and General (Internal) Medicine, University Hospital of Wales, Cardiff, UK Andrew Freedman MA MD FRCP Senior Lecturer in Infectious Diseases, University of Wales College of Medicine; Consultant Physician, University Hospital of Wales, Cardiff, UK Richard A Watts MA DM FRCP Consultant Rheumatologist Ipswich Hospital NHS Trust, Ipswich, UK

Contents 1. Therapeutics and toxicology

14. Critical care medicine

1

725

J Goldstone, G J Bellingan, M Smith

P A Routledge

2. Acute physical and environmental disorders 45

15. Gastrointestinal disease 751 R G Long, B T Cooper

J Moxham, R L Souhami

16. Liver and biliary tract disease 835 3. The genetic basis of disease

57

J O'Grady

R M Winter

17. Endocrine disease 4. Immunological disorders

79

883

A P Weetman

J D M Edgar, H F Sewell

5. Nutrition in clinical medicine

105

18. Metabolic bone disease and mineral metabolism 953 J Cunningham, J P Monson

M E J Lean

6. Cancer medicine

19. Diabetes mellitus and lipid metabolism

145

979 J Wilding, G Williams

R L Souhami

7. Ageing and disease 171 20. Renal and urinary disease 1031 J Cunningham

P W Overstall

8. Psychological medicine

205 21. Salt and water homeostasis and acidbase balance 1099

J Holmes

9. Infectious, tropical and parasitic disease 259 M H Wansbrough-Jones, S G Wright

10. Skin disease

J P Monson

22. Musculoskeletal and connective tissue disease 1119

379

B Hazleman

C T C Kennedy, RAG Graham-Brown

23. Haematological disorders 1201 11. AIDS and genitourinary disease

439

D C Linch

I G Williams

24. Neurological disease 12. Cardiovascular disease

463

1283

J W Scadding, J Gibbs

J M Walker, L B Tan

Index 13. Respiratory disease J Moxham, J Costello

1443

605 ix

This page intentionally left blank

n Therapeutics andToxicology Philip A Routledge

Pharmacodynamic variability refers to differences in response to the same drug concentration at its effector site. This is sometimes related to differences in the number and/or activity of specific receptors or other effector sites (e.g. ion channels). The very young and the elderly appear to have an increased sensitivity to several drugs, particularly those affecting the central nervous system. An increased sensitivity to drugs acting on the central nervous system is also seen in patients with chronic renal and liver disease. The elderly appear also to have less adaptable homeostatic mechanisms and are therefore more prone to drug-induced postural hypotension or hypothermia. Hyperthyroid patients are more sensitive to the effects of oral anticoagulants and less sensitive to digoxin.

Causes of variability in drug response 1

Poisoning with gases and volatile solvents 33

PHARMACOKINETIC VARIABILITY

Adverse drug reactions 5

Overdose with alcohols 35

Drug interactions 7

Poisoning with corrosive agents 36

This occurs when the handling of a drug is altered in disease or in different physiological states. The causes of variability are shown in Summary 1 (p. 4).

Therapeutic drug monitoring 11 Specific drugs 12 Drug regulation and development 13

Drug overdose 15

Poisoning with insecticides, herbicides, fungicides and rodenticides 37 Poisonous plants and fungi 39 Poisonous animals 41

Diagnosis of poisoning 16 Principles of management of poisoning 18

Poisoning by metals 43

Management of specific drug overdose 23

Therapeutics is the branch of medicine that deals with the treatment and cure of disease and ill health. Medicines are the most important tool in therapeutics, and an understanding of their pharmacology in normal dose and excess dose (toxicology) is essential in ensuring their safe and effective use. Therapeutics developed as a specialty because of the need to understand the causes of biological variability in drug response in humans, particularly in disease states.

CAUSES OF VARIABILITY IN DRUG RESPONSE PHARMACODYNAMIC VARIABILITY There are three major ways in which a drug can produce pharmacological effects: • By combination with specific receptors; • By alteration of physiological enzyme processes; • By direct physical or chemical action.

Absorption Drugs are usually given by mouth and must therefore pass through the bowel wall in order to enter the bloodstream. This occurs via four different mechanisms: • Passive diffusion, which involves diffusion down a concentration gradient from the gut into the bloodstream. For this to occur the drug must dissolve in the gastrointestinal fluids, traverse the cell membrane and enter the bloodstream. The rate at which it does so depends on both the concentration gradient and the lipid solubility of the drug. This is the most important mechanism of absorption for most drugs. • Active transport, a less important mechanism by which some drugs, because of their resemblance to naturally occurring substances, use existing transport systems. Thus, levodopa is absorbed by the active process responsible for absorption of the amino acid tyrosine. • Filtration through pores, a route generally limited to molecules of molecular weight less than 100 (e.g. urea). Most drugs are too large to traverse the pores between cells. • Endocytosis, by which drugs or particles are engulfed by cells in the bowel wall. This is of relatively little importance to drug absorption. Factors affecting absorption Absorption is affected by the chemical nature and formulation of the drug, as well as the physiological characteristics of the absorption site. Thus, some drugs cannot be administered orally because they are broken down by gastrointestinal enzymes (e.g. the polypeptide insulin) or are unstable at extremes of pH (e.g. benzyl penicillin at gastric pH). Most drugs are either weak acids or weak bases and thus exist in two forms in solution - as ions or as undis-

1

sociated molecules. At the low pH encountered in the stomach, a weakly acidic drug exists mainly in its undissociated form; it is thus more lipid soluble and will tend to be more easily absorbed. Basic drugs, on the other hand, are largely ionized at low pH and exist predominantly in the undissociated form only in the alkaline medicine of the small bowel. Despite these considerations, both acidic and basic drugs tend to be absorbed predominantly in the small intestine, because they are there for longer and because the inner surface of the intestine provides a much greater absorptive area than the stomach. Formulation factors are important in determining the rate and extent of drug absorption. Most drugs consist not only of an active agent but also of diluents, granulating and binding agents, lubricants and disintegrating agents, any of which may have a marked effect on absorption. Particle size of the active constituent may also be important, and sustained- or delayed-release preparations will also delay absorption rate. Gastrointestinal motility affects the rate of absorption. Delays in gastric emptying will result in delayed drug absorption, although the extent of absorption is normally unchanged. Food has a variable effect. Some drugs (e.g. propranolol) have greater bioavailability in the presence of food because of reduced presystemic elimination, whereas others (e.g. rifampicin) are more poorly absorbed. Such factors may be important in determining the times of administration of drugs. The route of administration is also important. Alternatives to the oral route are:

2

• Intramuscular injection, either to speed the onset of effects or to enhance bioavailability when the oral route is suboptimal. Use of the intramuscular route can, however, cause local pain, and some drugs (e.g. phenytoin and diazepam) are poorly and variably absorbed when administered in this way, because of poor solubility at tissue pH. Absorption rate varies with the muscle group used and may be slow in situations of poor tissue perfusion. • Rectal administration can sometimes be used, but because the haemorrhoidal veins drain partly into the portal tract, presystemic elimination is not completely avoided. The suppository may also be variably retained or have poor and erratic absorption. • Buccal administration can completely avoid presystemic elimination and is used for drugs (e.g. glyceryl trinitrate) that are rapidly metabolized after the oral route. Unfortunately, not all such drugs can rapidly penetrate the buccal mucosa and enter the systemic circulation. • Transdermal administration is an alternative to the buccal route for drugs such as glyceryl trinitrate or the antihistamine hyoscine. The rate of absorption varies with the skin site used and the extent of absorption is limited, so that only very potent drugs (effective dose <10mg) can be usefully administered by this route. Absorption is normally rapid, but membranes are being developed to control the rate.

• Intrapulmonary route. Some drugs, and all anaesthetic gases, are administered by the intrapulmonary route. Non-gaseous compounds are difficult to introduce into the lungs. Only 10% of a drug administered by pressurized aerosol actually enters the lungs: the remainder is deposited in the buccal cavity and subsequently swallowed. Absorption of drugs is normally so efficient that it is unusual to see differences in the extent of absorption of most modern drugs, even in the presence of extensive bowel resection. However, reduced absorption may occur in severe malabsorption syndromes or in severe diarrhoea, or in severe congestive cardiac failure because of oedema of the small intestine. Differences in bowel motility may affect the rate (often more than the extent) of absorption. Gastric emptying is slow and erratic in the very young and in adults who are shocked or in pain. In such circumstances (e.g. myocardial infarction) threshold concentrations for drug effect may not be achieved rapidly, or at all, and the drug (e.g. analgesic treatment) may be more effective when given by an alternative, non-alimentary route.

Distribution If drug distribution did not occur, only those drugs with an action within the vascular compartment would provoke biological effects. Like absorption, distribution is dependent on the physicochemical characteristics of the drug as well as upon biological factors. The major physicochemical properties determining drug distribution are lipid solubility and (if the drug is a weak electrolyte) the degree of dissociation (ionization) in tissues. Highly lipid-soluble drugs are generally distributed throughout all fluid compartments and, because they can cross lipid membranes readily, can reach most organs, including the brain. Poorly lipid-soluble drugs are normally able to enter the interstitial fluid, but if they are weak electrolytes only the undissociated form can penetrate the cells, and then only to a limited extent, depending on the lipid-water partition coefficient. Biological factors determining the distribution of drugs in the body include the degree of both plasma protein binding and of uptake by tissue, either by binding or by active transport mechanisms. The blood contains approximately 180 g of protein, half of which is albumin. Albumin is thus the most abundant single protein in the plasma and, because of its structure, is able to reversibly bind many drugs by Van der Waals forces, hydrogen bonds or hydrophobic bonds. However, its binding capacity is relatively limited and it has been estimated that only 200 mg of drug of average molecular weight (300) would be bound to albumin in plasma if there were only one binding site available on each albumin molecule. For some drugs there may be more than one binding site and the binding capacity will be relatively greater. Major sites of binding are the so-called site 1 (warfarin site), which also binds the endogenous compound bilirubin, and site 2 (diazepam

1

FIG. 1.1 Relationship between the percentage of diazepam and theophylline free in plasma and plasma albumin concentration in different disorders See text for details.

FIG. 1.2 Relationship between the percentage of lidocaine (lignocaine) free in plasma and plasma AAG concentration in several disease states and in neonates

site), the site at which the endogenous amino acid tryptophan also binds. Drugs may bind to either or both of these sites and can competitively displace each other. The relationship between the serum albumin concentration and the proportion of two drugs in the free (unbound) form is shown in Figure 1.1. It can be seen that the differences in protein binding are likely to be relatively, small over the range of serum albumin concentrations seen in health. In situations of hypoalbuminaemia, however, even small further decreases in serum albumin concentration may cause marked increases in the proportion of drug free in the plasma. Another plasma protein, a,-acid glycoprotein (AAG) also binds drugs, particularly those of a basic nature. Although its plasma concentration is much lower (c. 2%) than that of albumin, its affinity for drugs is high and some drugs (e.g. lidocaine (lignocaine), erythromycin and disopyramide) may bind appreciably more to this protein than to albumin. Because of its low concentration, however, its binding capacity is relatively limited and, for several drugs, protein binding falls with increasing drug concentration. The tissues also contain proteins which combine with drugs (60% of the total exchangeable body albumin is outside the vascular compartment), whereas body fat can serve as an important reservoir for lipid-soluble agents such as thiopentone. Finally, active transport mechanisms - such as the uptake of adrenergic neuron-blocking drugs by the adrenergic neuron - may also contribute to high tissue concentrations of a drug. Extensive tissue binding tends to result in the drug having a large apparent volume of distribution (Vci), as most of the drug is outside the intravascular space. Distribution of drugs may be altered by changes in regional blood flow in disease. The volume of distribution of lidocaine (lignocaine), for example, appears to be lower in patients with cardiac failure. Protein binding also determines drug distribution. Albumin concentrations are low in the neonate, begin to decline over the age of 40, and are reduced during pregnancy. Severe hypoalbuminaemia also

occurs in nephrotic syndrome and chronic liver disease. In patients with chronic renal failure the proportion of drug in the free form is greater than would be expected from the reduction in serum albumin alone, indicating either structural changes in albumin or the presence of endogenous inhibitors of binding. AAG is reduced in neonates and in pregnancy but tends to be normal in healthy elderly individuals. It is markedly increased in patients with rheumatoid arthritis and other inflammatory diseases, including chronic renal disease, after myocardial infarction, and in cancer. As with albumin, it may be reduced in patients with nephrotic syndrome or chronic liver disease. Concomitant changes in plasma protein binding of drugs also occur, the effects depending on the relative strength of binding of the drug to these two proteins and on the disease state in question (Fig. 1.2), which may alter the amount bound to free drug. Thus the total plasma concentration of drug (free + bound), which is generally measured in clinical practice, may not always reflect the free (active) concentration. Therapeutic ranges for total drug may therefore need to be altered in diseases where plasma protein binding is affected (see Therapeutic drug monitoring).

Metabolism The liver is the main organ of drug metabolism, although some may also occur in the gastrointestinal tract, lung, blood or kidney. Metabolism is particularly important for lipid-soluble drugs, which more readily enter cells and are also prevented by their lipid solubility from being excreted unchanged by the kidney. The routes of metabolism are relatively non-specific and are shared by a variety of drugs. In most instances, more water-soluble metabolites (which are more easily excreted by the kidney) are produced. There are two groups of metabolic reaction: phase I and phase II. Phase I reactions involve the enzyme mono-oxygenase, situated in the smooth endoplasmic reticulum of cells with metabolic function, especially the liver. These are called

3

cytochrome P450 enzymes and are responsible for oxidation, reduction, hydrolysis and dealkylation of drugs with the relevant molecular structure. A given drug may be metabolized by one or more of these routes to produce more water-soluble (and thus biologically active) metabolites. Phase II reactions are conjugation reactions of more water-soluble molecules to the drug or metabolite. Conjugation with glucuronide (glucuronidation) is the commonest pathway, but conjugation can also occur with sulphate, amino acids such as glutathione, or acetylCoA (acetylation). Most of these conjugated metabolites (except for some acetylated compounds) are biologically inactive. Although most metabolic processes are controlled by environmental factors, acetylation is under genetic control: 45% of the UK population are slow acetylators and the remainder either intermediate or fast acetylators. Approximately 7% of the British population have a reduced ability to hydroxylate several drugs (e.g. debrisoquine, nortriptyline and phenformin), which may lead to an increased risk of adverse reactions to drugs metabolized by this route. Metabolism of drugs may be diminished at the extremes of age. Neonates have a reduced capacity for phase I, and most phase II reactions. Thus the effect of caffeine (a phase I pathway) is markedly reduced. Reaction of phase II pathways such as the glucuronidation of chloramphenicol, for example, may lead to grey baby syndrome in the neonate. Immaturity of these processes is greatest in the premature neonate, who is therefore at greatest risk of adverse drug reactions. In the healthy elderly, impairment of drug metabolism is only slight and tends to be overshadowed by changes in pharmacodynamic response. However, marked reductions in drug metabolism may be seen in the frail elderly. Dietary and environmental factors, such as smoking, may affect the rate of metabolism of some drugs; and chronic liver disease is often associated with an impairment of metabolism, particularly of phase I pathways. In liver disease the serum albumin concentration correlates roughly with the degree of impairment of drug metabolism (i.e. the lower the albumin the more severe the impairment). The effect of impairment of metabolism in liver disease will be greater for those drugs with extensive presystemic metabolism. The bioavailability of propranolol is much greater in patients with chronic liver disease, not because of better absorption (absorption is virtually complete in health) but because it cannot be efficiently metabolized on its first passage through the liver. This is due not only to the reduced hepatocyte metabolic activity, but also to intra- and extrahepatic shunting of blood past the liver cells and a consequent reduction in effective liver blood flow.

A MCQ1.1-1.5

4

SUMMARY 1 Pharmacokinetic causes of variability in drug response Absorption Extent of drug absorption reduced only in rare circumstances Distribution Altered in disease by changes in regional blood flow, altered plasma protein binding (e.g. hypoalbuminaemia in hepatic cirrhosis), or changes in activity of energy dependent transporter systems (e.g. P-glycoprotein) Metabolism Diminished in neonates (particular if premature) and the elderly (especially if frail). Affected by genetic factors (e.g. genetic polymorphisms), environmental factors (e.g. smoking) and liver disease Excretion Immature in neonates (particularly if premature) Impaired in renal disease and declines with advancing age over 40 years

In cardiac failure and shock, cardiac output is reduced and a disproportionately greater reduction in hepatic blood flow (normally around 1.5L/min) may result in a marked reduction in the systemic clearance of normally efficiently cleared compounds (e.g. lidocaine (lignocaine)), for which hepatic blood flow is the major determinant of delivery of drug to and clearance by the liver. In addition, the liver can be affected by the increased venous backpressure caused by failure of the right side of the heart in biventricular (congestive) cardiac failure. This results in increased size and congestion of the liver and derangement of liver function, with reduced hepatocyte clearance of several drugs. Genetic factors play a major role in determining drug response and handling, and susceptibility to adverse drug reactions. Allelic variations affecting pharmacokinetics with a frequency of at least 1% in the population are often termed genetic polymorphisms (e.g. acetylation) and are associated with a bimodal or trimodal (rather than the normal unimodal) distribution of clearance values. Some clinically important pathways of drug metabolism showing such polymorphisms are ./V-acetylation, oxidation (hydroxylation) via the cytochrome (CYP) P450 isoform, CYP2D6, succinylcholine hydrolysis (de-esterification) and thiopurine s-methylation. More are being recognized as detection techniques improve.

Excretion The vast majority of drugs must first be metabolized to more water-soluble compounds before they can be excreted in the urine. Those drugs (e.g. cimetidine, digoxin, aminoglycosides, lithium, cephalosporins, penicillins, atenolol and chlorpropamide) that are largely excreted unchanged tend to be relatively water-soluble and are not appreciably bound to plasma proteins and so can enter the glomerular filtrate. Lipid-soluble drugs may also enter the glomerular filtrate, provided they are not extensively plasma protein bound, but they then readily pass back

through the proximal tubular cells into the bloodstream by passive diffusion down a concentration gradient. Only when they have been metabolized to more water-soluble compounds will they have difficulty in passing back through the renal tubular cells and therefore be excreted in the urine. Weak electrolytes can also be actively secreted in the proximal renal tubules - weak bases sharing one mechanism and weak acids another - and there is some evidence for an active transport system promoting the excretion of digoxin in the distal renal tubule. Drugs may also be excreted in the bile, normally as conjugates and particularly if they have a high molecular weight. The drug may, however, be reabsorbed from the intestine, either directly or after deconjugation by intestinal microflora, to create an enterohepatic cycle. This may reduce the clearance of the drug and thus prolong its effect. Excretion of drugs is reduced in neonates (particularly if premature) because of a reduced glomerular filtration rate (GFR) relative to their body surface area, and also because their active renal tubular secretory mechanisms are not fully developed. These processes mature relatively rapidly and are fully functional after 6-12 months, but GFR begins to fall again over the age of 40 (the renal tubular secretory mechanisms appear to be less affected by age). Thus, the dose of digoxin or gentamicin (drugs excreted predominantly by glomerular filtration) may need to be reduced in the elderly or the very young. Renal disease is associated predominantly with a fall in GFR, with active renal tubular secretion mechanisms being less affected. Measurement of serum creatinine or creatinine clearance is therefore useful in calculating the optimum dose of drugs excreted predominantly by glomerular filtration, but it is of less value for drugs such as penicillins, which are actively secreted at the proximal renal tubule. 1

ADVERSE DRUG REACTIONS It has been estimated that approximately 5 % of all hospital admissions are because of adverse reactions to a drug; 1 in 10 patients admitted for other reasons will develop an adverse reaction to a drug while in hospital; and perhaps 1 in 1000 deaths in hospital are due to an adverse reaction to a drug, making adverse reactions between the fourth and sixth most common cause of death in the USA, for example. Adverse reactions are also common in general practice. The drug groups most commonly associated with adverse reactions are antihypertensives, anticoagulants, antirheumatics (particularly the non-steroidal antiinflammatories), cytotoxic agents, corticosteroids and digoxin. Although adverse reactions can affect all systems of the body, they are most often described in the skin and central nervous system, perhaps because detection tends to be easier. The very young (i.e. under 1 year of age) and the elderly are particularly at risk, and women appear to be at greater risk than men, although the reason for this is

unknown. Patients with a history of atopy, or who have had a drug reaction in the past, are more likely to develop further reactions, even to a drug in some other therapeutic group. The presence of disease, particularly heart failure or liver or renal disease, predisposes to a risk of adverse reactions, as may genetic factors, such as enzyme deficiencies.

1

TYPES OF ADVERSE DRUG REACTION Adverse reactions have been classified into two major types.

Type A The most common type is the dose-dependent adverse reaction (type A), which is an 'augmentation' or 'accentuation' of a known pharmacological effect of the drug not necessarily always the wanted effect. Thus, fisympathomimetic agents can cause increased tremor, a recognized pharmacological effect of the drug but one that may limit their use in bronchodilator therapy. Type A reactions account for about 75% of all adverse reactions and, because they are often insidious in onset, may go unrecognized for some time. They may therefore produce considerable morbidity, but the mortality from type A reactions is generally low. They may occur because of altered pharmacokinetic or pharmacodynamic factors and are therefore commonest in diseases (e.g. renal and liver disease) where these are most affected.

Type B The occurrence of type B reactions is not dependent on the dose given. They are unpredictable or 'bizarre' reactions to a drug that may be present only in tiny concentrations, although the severity of the reaction may be greater if a larger dose is administered rapidly (e.g. after intravenous administration). Because of their lack of predictability and their often acute onset they have a proportionally higher mortality than type A reactions, but associated morbidity is relatively low if the patient survives the initial event. Like type A reactions, type B reactions can be related to pharmacokinetic or pharmacodynamic factors. There are many ingredients in medications other than the active

SUMMARY 2 Types of adverse drug reaction (ADR) Type A Dose-dependent 'accentuation' of known pharmacological effect. Account for 75% of ADRs. Cause considerable morbidity but mortality is low. Can sometimes be treated by dose reduction rather than cessation of treatment Type B Non-dose dependent, 'bizarre' and unpredictable. Higher mortality. Type B reactions can be related to hereditary enzyme deficiencies or allergic mechanisms. Usually treated by discontinuing treatment immediately

5

compound: reactions may occur to the preservative, filler, dye, binder or solubilizing agent in the tablet or injection. Type B reactions may occur because individual subjects metabolize the drug through a novel pathway that produces a toxic metabolite. In other cases the presence of hereditary enzyme deficiencies may predispose the patient to the toxic effects of drugs or their metabolites. Thus, haemolysis caused by dapsone and other oxidizing agents is commoner in patients with glucose 6-phosphate dehydrogenase (G6PD) deficiency. Similarly, many agents can precipitate an attack of acute intermittent porphyria (p. 1027) in those carrying a gene for this disorder. Type B reactions are also often related to allergic mechanisms, and may involve any or all forms of antibody- or cellmediated tissue damage. Thus, anaphylaxis may occur as a result of immediate (type I) hypersensitivity associated with cell-bound IgE, after the administration of drugs such as penicillin. Cell surface (type II hypersensitivity) damage may occur as a result of the combination of methyldopa with red cell (or quinine with platelet) membranes and subsequent haemolysis or thrombocytopenia, 1 respectively. Toxic immune complexes may be formed in patients receiving gold or penicillamine, resulting in skin toxicity by a type III hypersensitivity mechanism. Cell-mediated (type IV) tissue damage is thought to be the mechanism of the relatively common (but usually mild) rash associated with ampicillin therapy 2 or the more serious erythroderma sometimes associated with drugs such as gold. 34

DIAGNOSIS AND MANAGEMENT OF ADVERSE DRUG REACTIONS Diagnosis It is rarely possible to make a diagnosis of drug-induced adverse reaction with complete certainty. Either an infective agent or the antibiotic with which the patient was being treated, for example, may cause erythema multiforme. It is therefore necessary to weigh up all circumstantial evidence carefully before making the diagnosis. This includes taking a full drug history, remembering that patients may not consider over-the-counter and herbal remedies, or even the oral contraceptive pill, as medicines or drugs. The brand names of the agents may also be important, as different pharmaceutical formulations may contain different dyes, binders, fillers, preservatives or lubricants, and reactions may occur only with certain formulations. Physical examination provides only limited information. Some physical signs may be directly attributable to drug therapy, e.g. the blue discoloration of the skin

1

Fig. 1.1

4

Case 1.1 Case 1.2

2 Fig. 1.2

3

© Fig. 1.3 6

MCQ1.6

Fig. 1.4

associated with amiodarone therapy 0 is rarely caused by disease. Similarly, an eruption occurring at a fixed site on the skin may suggest a drug-induced fixed drug eruption. 6 However, in most cases the organs affected have a limited number of responses to noxious stimuli. Biochemical tests are also of limited value, but measurement of plasma drug concentration may help to confirm the suspicion of a dose-dependent (type A) adverse reaction. In some cases measurement of metabolic phenotype may also be helpful. It is extremely rare, for example, for patients who are fast acetylators to develop drug-induced lupus erythematosus after hydralazine therapy. Probably the most helpful indication of the drug responsible for the adverse reaction is the time course relative to drug treatment, particularly the rate of resolution of symptoms on stopping a specific drug (dechallenge). Subsequent re-exposure (rechallenge) associated with a recurrence of the symptoms or signs gives the strongest evidence of causation, but it is rarely ethically justifiable unless the patient's need for the drug outweighs the potential risk of the adverse reaction.

Management The only choice with most type B reactions is to stop the suspected drug completely. With type A reactions it is sometimes possible to reduce the dose to prevent the unwanted effects while still obtaining therapeutic benefit. If an adverse reaction does occur, the patient's medical notes and prescription chart should be clearly marked with an eye-catching label to indicate the nature and date of the reaction. Any serious adverse reaction should also be reported to the Committee on Safety of Medicines using a yellow card. In addition, even minor adverse reactions to drugs that have been recently introduced (marked with an inverted black triangle in the British National Formulary) should be reported.

PREVENTION OF ADVERSE REACTIONS It must be remembered that although drugs provide enormous benefit to patients, all are potentially toxic and should be used selectively, particularly in those at greatest risk of adverse reactions, i.e. the very young, the very old and those with severe disease. The risk of developing an adverse reaction appears to increase disproportionately with increasing number of drugs prescribed concomitantly. Prescriptions should therefore be reviewed and, whenever possible, drug therapy rationalized at frequent intervals. When a drug is prescribed, the prescription should be written legibly, in block letters, and preferably using the approved name. Failure to do so may lead to errors in prescribing. In some cases the risk of adverse reactions to certain drugs is such that they should be avoided if at all possible. There are particular dangers during pregnancy or breastfeeding, when certain drugs may cause an adverse reaction in the fetus or breastfed child (Table 1.1).

TABLE 1.1 Some drugs to avoid during pregnancy or breastfeeding Pregnancy

Breastfeeding

Antibiotics

Tetracyclines Sulphonamides Aminoglycosides Chloramphenicol Trimethoprim Nitrofurantoin Quinolones

Tetracyclines Sulphonamides Isoniazid Chloramphenicol Metronidazole Nitrofurantoin Vancomycin

Antithyroids

All, including radioiodine

All, including radioiodine

Analgesics

Opiates NSAIDs

Aspirin NSAIDs (several)

Anticoagulants

Warfarin Acenocoumarol (nicoumalone)

Phenindione

Antidepressants Antihypertensives

SSRIs (several) Lithium ACE inhibitors Angiotensin receptor antagonists Thiazide, loop and potassium-sparing diuretics Hydralazine Minoxidil

Angiotensin receptor antagonists Calcium channel blocker (some)

Antineoplastics

Most

Most

Antidiabetics

Biguanides Sulphonylureas

Use with caution

Antirheumatics

Gold Penicillamine

Antiulcer drugs

Misoprostol Most proton pump inhibitors

DRUG INTERACTIONS

1

Drug interactions have been recognized for over 100 years and were described under the classic headings of antagonism, synergism or potentiation. Many of these interactions are beneficial and have been used to therapeutic advantage, but unwanted or adverse interactions also account for 10-20% of all adverse reactions to drugs. There are around 2000 potentially interacting drug pairs, but the number of clinically significant adverse interactions is much lower, largely involving drugs with a relatively low margin between safety and toxicity (i.e. a narrow 'therapeutic ratio'). Thus, they are most often described with oral anticoagulants, some antidiabetic drugs, anticonvulsants, antiarrhythmics (especially digoxin), tricyclic antidepressants, antihypertensive drugs, NSAIDs, neuroleptic drugs, lithium, anticancer/immunosuppressive agents and theophylline (Tables 1.2-1.4). The elderly, particularly if frail and suffering from multiple medical conditions, are more likely to experience an interaction and more likely to have serious adverse consequences. As with adverse reactions in general, the risk of developing an adverse interaction increases disproportionately with the number of drugs being prescribed concomitantly; the mechanisms of interaction are also similar, in that both pharmacodynamic and pharmacokinetic interactions may occur. Interactions may also take place, however, before the drugs even enter the body, and these have been termed 'pharmaceutical' interactions. 7

PHARMACEUTICAL INTERACTIONS Misoprostol Proton pump inhibitors

Antivirals

Most

Anxiolytics

Benzodiazepines Chloral derivatives

Sex hormones

Androgens, oestrogens and high-dose progestogens

Androgens, oestrogens and high-dose progestogens

Hypolipidaemics

Fibrates Statins

Fibrates

Miscellaneous

Vitamin D and its analogues Live vaccines Podophyllum resin and podophyllotoxin Tamoxifen Thrombolytics

Anthraquinone laxatives and phenolphthalein Amiodarone

NB Absence of a drug from this list does not imply safety. Further information can be obtained by consulting the manufacturer's datasheet, the British National Formulary, or by contacting the National Teratology Centre (0191-232-1525).

Interactions may occur prior to the drug entering the body when drugs are mixed together for infusion, or when a drug interacts with the infusion material itself. Guides to intravenous mixtures are available and should be consulted before compounds are given by intravenous infusion or mixed together. In general terms, it is advisable to avoid adding drugs to infusion fluids unless absolutely necessary, and certainly to avoid adding more than one drug if at all possible. Infusion fluids particularly likely to be associated with pharmaceutical incompatibilities include blood and blood products, amino acids and lipid solutions, mannitol and sodium bicarbonate.

PHARMACODYNAMIC INTERACTIONS Pharmacodynamic interaction occurs where one drug alters the response to another by interacting either at the receptor site or at a different site to enhance or diminish the primary drug effect. Partial agonists initiate a minor response but, by occupying a significant fraction of the receptors, they antagonize the action of more potent agonists. Nalorphine, for example, is a partial antagonist

7

TABLE 1.2 Some drug interactions of clinical importance resulting in increased drug effect Drug (A)

May interact with B

Effect of interaction

Mechanism of interaction

ACE inhibitors Antidepressants (tricyclic) Antihypertensive agents

NSAIDs Enzyme inhibitors Vasodilators (e.g. nitrates for angina) neuroleptics and some antidepressants NSAIDs

Hyperkalaemia, reduced renal function Increased effect of A Postural hypotension

Additive nephrotoxic effects Reduced clearance of A Combined hypotensive effects

Peptic ulceration

Enzyme inhibitors Verapamil NSAIDs (including aspirin) Enzyme inhibitors Amiodarone Diltiazem Verapamil Diuretics (loop and thiazide) ACE inhibitors Potassium supplements NSAIDs Thiazide diuretics Anticholinergic drugs (e.g. some antihistamines/tricyclic antidepressants) Enzyme inhibitors NSAIDs

Increased effect of A

Increased risk of peptic ulceration Reduced clearance of A

Increased risk of peptic ulceration Increased effect of A Increased effect of A

Corticosteroid prevents healing? Reduced clearance of A Reduced clearance of A

Increased effect of A (e.g. arrhythmias) Hyperkalaemia, impaired renal function

Diuretic-induced hypokalaemia Combined potassium-elevating effects

Increased effect of A

Reduced clearance of A

Excessive anticholinergic effects

Combined anticholinergic effects

Increased effect of A Seizures

Reduced clearance of A Pharmacodynamic interaction at CNS effector site Pharmacodynamic interaction Reduced clearance of A Reduced clearance of A

Aspirin (low-dose) Carbamazepine Corticosteroids (oral) Ciclosporin Digoxin Digoxin Diuretics (potassium-sparing) Lithium Phenothiazines and butyrophenones Phenytoin Quinolones Sildenafil (Viagra) Theophylline Warfarin

Nitrates Saquinavir Enzyme inhibitors (e.g. macrolides and 4-quinolones) See Table 1.4

Increased hypotensive effect of A Increased hypotensive effect of A Increased effect of A

TABLE 1.3 Some drug Interactions resulting in reduced drug effect Drug (A)

May interact with B

Effect of interaction

Mechanism of interaction

Antidepressants Antihypertensives (e.g. ACE inhibitors, thiazides and p-blockers) Calcium channel blockers Corticosteroids (oral) Ciclosporin Digoxin

Enzyme inducers NSAIDs

Reduced effect of A Reduced effect of A

Enzyme inducers Enzyme inducers Enzyme inducers Colestyramine (cholestyramine) Colestipol Enzyme inducers Broad spectrum antibiotics Enzyme inducers Colestyramine (cholestyramine) Colestipol Antacids Enzyme inducers Enzyme inducers Colestyramine (cholestyramine) and colestipol See Table 1.4

Reduced effect of A Reduced effect of A Reduced effect of A Reduced effect of A

Increased clearance of A Pharmacodynamic antagonism of anti hypertensive effect of A Increased clearance of A Increased clearance of A Increased clearance of A Reduced absorption of A

Reduced effect of A Reduced effect of A Reduced effect of A Reduced effect of A

ncreased clearance of A nterruption of enterohepatic recycling of A? ncreased clearance of A Reduced absorption of A

Reduced effect of A Reduced effect of A Reduced effect of A

ncreased clearance of A ncreased clearance of A Reduced absorption of A

Oral contraceptives Phenytoin Quinolone antibiotics

Theophylline Thyroxine

8

Warfarin

TABLE 1.4 Important drug interactions involving warfarin

Interacting drug

Effect of interaction on anticoagulant effect

Colestyramine (cholestyramine) Colestipol

Reduced anticoagulant effect

Impaired absorption and increased elimination of warfarin. Long-term treatment may cause impaired vitamin K absorption and enhance anticoagulant effect

Barbiturates Carbamazepine Griseofulvin Phenytoin (see also below) Primidone Rifampicin Ritabutin St John's wort

Reduced anticoagulant effect

Induction of warfarin metabolism

Amiodarone Azapropazone Chloramphenicol Cimetidine Ciprofloxacin Clarithromycin Dextropopoxyphene Erythromycin Fluconazole Itraconazole Ketoconazole Mefenamic acid Metronidazole Miconazole Nalidixic acid Norfloxacin Ofloxacin Phenylbutazone Sulfinpyrazone Sulphonamides (e.g. in co-trimoxazole)

Increased anticoagulant effect

Inhibition of warfarin metabolism

Anabolic steroids Bezafibrate Clofibrate Danazol D-thyroxine Gemfibrozil L-thyroxine Phenytoin (see also above) Salicylates/aspirin (high-dose) Stanozolol Tamoxifen

Increased anticoagulant effect

Pharmacodynamic potentiation of anticoagulant effect

NSAIDs (including aspirin at all doses) clopidogrel and ticlopidine

Increased risk of bleeding

Additive effects on coagulation and haemostasis

Oral contraceptives, oestrogens and progestogens Vitamin K (e.g. in some enteral feeds)

Reduced anticoagulant effect

Pharmacodynamic antagonism of anticoagulant effect

Probable mechanism(s)

but may add to the respiratory depression produced by opioids, whereas naloxone, despite having no opioidlike effects itself, antagonizes the pharmacological effects of most opioids. Many pharmacodynamic interactions occur when the physician forgets that a drug known to be acting at one receptor may also act at another. Thus, antihistamines, phenothiazines and tricyclic antidepressants may have muscarinic anticholinergic activity in addition to their desired pharmacological effects: administration of two or more of these agents (particularly in the elderly) may lead to pronounced anticholinergic symptoms and signs. Pharmacodynamic interaction may also occur when two drugs act at different sites, e.g. the interaction of digoxin and diuretics. Here, hypokalaemia caused by diureticinduced reduction of potassium reabsorption in the kidney may potentiate the effects of digoxin on the heart.

1

PHARMACOKINETIC INTERACTIONS Absorption Drugs may interact with each other to prevent either the rate or the extent of absorption of another compound. Drugs that increase gastric emptying (e.g. metoclopramide) will increase the rate of absorption of many other drugs, so that the peak concentration of that drug will tend to be higher. The opposite effect is achieved by agents - such as the anticholinergic compounds atropine, trihexyphenidyl (benzhexol) and propantheline - that diminish gastric emptying. These changes are likely to be more important after single-dose administration in a situation where a threshold concentration for drug effect exists, e.g. for analgesics. Under these circumstances (particularly if the rate of elimination of the drug is high), a delay in absorption may result in an inability to reach therapeutic drug concentrations. The extent of absorption of the drug may be impaired by the presence of another through the formation of chelates, ion pairs or complexes. Ion exchange resins such as colestyramine (cholestyramine) and colestipol combine with warfarin, levothyroxine (thyroxine) and digitalis glycosides to im-pair absorption. Similarly, the calcium, magnesium or aluminium ions in antacids or milk may interact with 4-quinolones such as ciprofloxacin. In most cases these interactions can be prevented by ensuring a time interval of at least 2 hours between administrations of the two compounds. In the case of warfarin and digitoxin, however, colestyramine (cholestyramine) or colestipol may also affect enterohepatic recirculation and reduce concentrations of these drugs, even if they are administered after absorption has been completed.

Distribution Drugs may compete for both plasma protein-binding sites and active transport processes in the body. The importance of plasma protein-binding displacement interactions has

9

TABLE 1.5 Some drugs causing induction of drug metabolism Barbiturates and primidone Carbamazepine Cigarette smoking* Phenytoin Rifampicin and rifabutin St John's wort * Selective induction (e.g. of theophylline metabolism).

FIG. 1.3 Factors determining the distribution of drugs in the body

been much exaggerated (Fig. 1.3). For most drugs, only the free form in plasma is available for metabolism. Thus, an increase in the percentage of free drug will result in a concomitant increase in total clearance and the rapid return of the free (biologically active) drug concentration to its original value. (The net result is a lower total drug concentration, but a larger percentage of this is in the free form.) The temporary increase in free drug concentration is only likely to be important if the drug is normally highly protein bound (e.g. >90%) and has a low apparent volume of distribution. If the displacing agent is given by a rapid intravenous injection, the temporary increase in free drug concentration of the displaced agent may also be important, but clinically significant examples of this interaction are difficult to find. At present, whenever significant clinical interactions have been shown to be associated with protein-binding displacement, other mechanisms (e.g. competition for metabolism) have been shown to be the major causes of the interaction. It is therefore possible that plasma protein-binding displacement merely reflects the physicochemical similarities of the two compounds, which leads to their competing for transport or metabolic processes. Inhibition of active transport of one drug by another can result in significant interactions. P-glycoprotein is a surface phosphoglycoprotein that functions as an energydependent efflux. It is coded by the human MDR1 (multidrug resistance) gene and is present in a wide variety of tissues, including the gut, kidney, liver, brain, testes, placenta and adrenals. Human P-glycoprotein has been shown to transport a wide range of structurally unrelated drugs such as digoxin, quinidine, ciclosporin and HIV-1 protease inhibitors. Competition for this pump between quinidine and digoxin, for example, can result in increased concen-

O MCQ 1.7-1.11

10

trations of digoxin entering the CNS and corresponding toxicity. If potent and safe inhibitors of P-glycoprotein and other similar transporters (multidrug resistance-associated proteins or MRPs) can be found, the interaction could possibly be used to therapeutic effect in overcoming the intrinsic or acquired multidrug resistance of human cancers.

Metabolism Compounds in cigarette smoke and a variety of other drugs (Table 1.5) can stimulate drug metabolism by the hepatic mono-oxygenase system. Enzyme induction is related to an increase in the amount of cytochromes present, and may therefore take 7-14 days to develop fully and a similar lag period to wear off after the inducing agent is discontinued. This time lag tends to disguise any causal relationship between the drug interaction and administration of the inducing agent. Furthermore, if the dose of the active drug has been increased to maintain adequate effects, marked increases in drug concentration and toxic effects may be seen when induction wears off. Other agents can impair the metabolism of concomitantly administered drugs, either by competing for metabolism or in a non-competitive manner (Table 1.6). Cimetidine is an interesting example in that, although not itself metabolized (it is excreted largely unchanged in the urine), it inhibits the metabolism of drugs by the liver. As the effect of enzyme inhibitors is generally direct, the lag time for this effect to occur is shorter than with enzyme inducers and the effects disappear more quickly after discontinuation. The effects also tend to be more selective than with enzyme induction. Thus, allopurinol has little effect on drugs metabolized by the mono-oxygenase pathway but, by inhibiting the enzyme xanthine oxidase, can impair the metabolism of azathioprine and 6mercaptopurine and potentiate their toxicity. If a drug has a carbon atom with four different substituents, it exists as two optical isomers or enantiomers, which are mirror images of each other. Most chemical syntheses will result in equal amounts of the two steroisomers, and this is termed a racemic mixture. One isomer may be much more potent than the other (e.g. S-warfarin is at least five times more potent than R-warfarin), or the toxicity

1

TABLE 1.6 Some drugs causing inhibition of metabolism Allopurinol Amiodarone Cimetidine Metronidazole Monoamine oxidase inhibitors Sulphonamides Macrolides (e.g. erythromycin, clarithromycin, azithromycin) Some antifungals (e.g. ketaconazole, fluconazole, itraconazole, miconazole) Some quinolones (e.g. ciprofloxacin, norfloxacin, ofloxacin, nalidixic acid) Some HIV-protease inhibitors (e.g. saquinavir, indinavir, nelfinavir, ritonavir) Some SSRIs (e.g. fluoxetine, fluvoxamine, nefazodone, paroxetine, sertraline)

of the two isomers may differ. The enantiomers may also differ in metabolic route or interact differently with other compounds. Cimetidine inhibits the metabolism of Rwarfarin, metronidazole inhibits the S enantiomer, and amiodarone inhibits the metabolism of both enantiomers.

Excretion It is unlikely that drug-induced changes in GFR will markedly alter the excretion of other drugs by the kidney. However, drugs that alter urinary pH may have a marked effect on the renal clearance of weak bases or acids, because of the effects on passive reabsorption in the proximal tubule. The administration of bicarbonate, for example, may enhance the removal of weak acids (such as phenobarbital and aspirin) by increasing the proportion of drug in the ionized form in the urine, thus preventing passive reabsorption. Similarly, acidification of the urine (using ammonium chloride) reduces tubular reabsorption of weak bases such as amphetamine, an effect that was once used in overdose to enhance renal excretion. Acidic drugs may compete with each other for the active renal tubular secretory pathway. Salicylates and thiazides may potentiate the toxicity of methotrexate in this way. No clinically important interactions involving basic drugs at this site have been reported. Quinidine and verapamil may compete with digoxin for an active tubular secretory mechanism in the distal convoluted tubule. This mechanism is now thought to be P-glycoprotein (see earlier), so that the digoxin-quinidine interaction results in reduced renal clearance of digoxin (as well as the increased CNS penetration of digoxin described earlier), thus causing increased plasma and CNS digoxin concentrations and increased risk of toxicity. P-glycoprotein may also be a factor in the excretion of some drugs into the gut via the intestinal wall and the bile, and may thus determine the apparent bioavailability of certain drugs such as ciclosporin. Like CYP450 enzymes, P-glycoprotein activity can be enhanced by enzyme induction, and this may well be an important determinant of rifampicin-induced reduction in ciclosporin bioavailability, for example (Fig. 1.4). 1

FIG. 1.4 Schematic representation of the effect of P-glycoprotein activity on the disposition and bioavailability of drugs

AVOIDANCE OF DRUG INTERACTIONS There are a number of strategies that can be adopted to decrease the risk of potential clinical problems. The number of drugs prescribed for an individual should be limited to as few as is necessary. The drug list (including over-the-counter and herbal preparations) should be reviewed on a regular basis, potentially unnecessary medicines withdrawn and the patient's progress monitored subsequently. Patients should be encouraged into a 'prescribing partnership', alerting doctors, pharmacists and other healthcare professionals to symptoms that occur when new drugs are introduced. The prescriber should be alert for the presence of medicines (e.g. particularly likely to be associated with interaction (e.g. enzyme inducers and inhibitors)) and avoid such combinations. He or she should also be prepared to check (there are excellent tables in the British National Formulary} before prescribing any new or unfamiliar drug, particularly in those already receiving drugs with a low therapeutic ratio. Clinicians should also report via the appropriate postmarketing surveillance scheme any drug-drug interactions they feel they have encountered.

THERAPEUTIC DRUG MONITORING Therapeutic drug monitoring refers to the measurement of drug concentrations in plasma to aid in optimizing drug efficacy and reducing the risk of toxicity.

11

TABLE 1.7 Drug concentration associated with optimal efficacy in most patients Usual half-life (hours) Carbamazepine Ciclosporin Digoxin Gentamicin

2-10 30-40 2-3

Lithium Phenytoin

7-20 20-60

Theophylline

4-15

Usual therapeutic plasma concentrations

Toxic concentrations

Principal route of elimination

4-1 2 mg/L 1 00-200 ^g/L 0.8-2 ng/L Trough <2mg/L Peak 4-8 mg/L 0.4-1 .0 mmol/L Plasma 10-20 mg/L Saliva 1-2 mg/L Plasma 10-20 mg/L Saliva 6-1 2 mg/L

>12mg/L >200^g/L >2 g/L Trough >2 mg/L Peak >12 mg/L >1.5mmol/L >20mg/L >2mg/L >20mg/L >12mg/L

Hepatic metabolism Hepatic metabolism Renal excretion Renal excretion

The ideal way to monitor drug therapy is, of course, to measure wanted and unwanted effects directly: where good measures of these exist (e.g. in anticoagulant therapy), measurement of plasma drug concentrations has a very limited role. There are, however, situations in which drug efficacy or toxicity is more difficult to measure and, both here and where the drug is being given prophylactically to prevent serious events (e.g. antiepileptic and antiarrhythmic drugs), therapeutic drug monitoring may have a role to play (Table 1.7). Measurement of plasma (or urine) drug concentration is also of value in detecting overdose (either deliberate or accidental) and in monitoring compliance with therapy in patients who may not be responding normally to the drug. For therapeutic drug monitoring to be of value in optimizing therapeutic benefit, the drug should work by a reversible mechanism without the development of tolerance. The possibility of accumulation of active metabolites that are not detected by the drug assay must also be considered. Procainamide, for example, is metabolized to Nacetylprocainamide, which also has antiarrhythmic activity and, ideally, should also be measured.

Renal excretion Hepatic metabolism Hepatic metabolism

tration taken just before the next dose will correlate most closely with the steady-state concentration. As it takes approximately five half-lives before steady state is achieved, sampling before this time will tend to underestimate the subsequent steady-state concentration and is only indicated if toxicity due to drug accumulation is possible. Details of sampling time, time and extent of last change in dose, and time of administration of the most recent dose should always be given on the assay request form to enable the plasma concentration to be interpreted correctly. Total drug concentrations are normally measured and these generally mirror the free plasma concentration. This may not be true in diseases where changes in plasma protein binding of drugs is likely, such as hepatic and renal disease. For some drugs (e.g. phenytoin and theophylline) the concentration in saliva may be a closer reflection of the free plasma concentration: in young children, saliva collection (stimulated by a drop of citric acid solution on the tongue) may avoid venepuncture.

SPECIFIC DRUGS

Sample collection

12

Unlike many endogenous compounds, such as creatinine, which are produced at a continuous rate, drugs are normally administered at fixed intervals and the plasma concentration will therefore vary markedly with time after administration. Early after administration the plasma concentration may be affected by the rate of gastric emptying. However, if there is extensive distribution of the drug early samples may be higher than the average, or steady-state, concentration during the dosing interval, and may not accurately reflect the concentration at the site of action of the drug. It is therefore preferable to wait at least 8 hours after administration of an oral dose of digoxin, or 12 hours after lithium, before sampling. If peak levels are important, samples should be taken 2-3 hours after a conventional tablet, or later if a sustained-release formulation has been administered, or 30-60 minutes after intramuscular injection. In most cases, however, a trough concen-

Anticonvulsants Phenytoin metabolism shows saturation rather than firstorder kinetics, so that clearance falls with increasing dose and the risk of toxicity increases. Half-life also increases with increasing dose, and so it will take longer for steady state to be achieved. Most patients tend to be given too small a dose of phenytoin, but therapeutic monitoring has reduced the tendency to prescribe a second antiepileptic drug in many patients who initially received subtherapeutic doses. Therapeutic monitoring of other anticonvulsants except carbamazepine is of limited value except in the identification of poor compliance.

Antiarrhythmics Because of their serious toxicity and the narrow margin between therapeutic and toxic doses, monitoring of several

antiarrhythmic drugs may help to reduce adverse effects and improve response. The half-life of most antiarrhythmics is short, but that of amiodarone is around 30-45 days. A loading dose is therefore necessary to achieve early therapeutic concentrations of this drug, and prolonged monitoring will help to detect the accumulation of amiodarone and its major (and possibly toxic) metabolite desethylamiodarone.

and regular monitoring of plasma concentration and renal function may be helpful in reducing the risks of nephrotoxicity.

Digoxin

The importance of drug regulation has grown with the rapid increase in the number of therapeutic agents developed over the last 50 years. Increased efficacy of new compounds is often associated with a risk of serious adverse effects.

The plasma digoxin concentration is one of the determinants of toxicity of the drug, particularly noncardiovascular toxicity, causing lassitude, depression and anorexia. Monitoring is particularly important in patients with reduced renal function (e.g. in the elderly), whose digoxin clearance is reduced, and will also help to identify poor compliance.

Theophylline Patients vary greatly in metabolic clearance of this potentially toxic bronchodilator, and dose adjustment is facilitated by the measurement of plasma theophylline concentrations.

Antibiotics The aminoglycosides (gentamicin in particular, but also amikacin, kanamycin, netilmicin and tobramycin) are the most frequently monitored antibiotics. The risk of ototoxicity can be reduced by preventing excessive plasma drug concentrations, but the nephrotoxic effects of the drugs are less closely related to the plasma concentration, so that renal function must also be carefully monitored during therapy. The antibiotic vancomycin is used systemically in the prophylaxis and treatment of endocarditis and other serious infections caused by Gram-positive cocci, including multiresistant staphylococci. Vancomycin (added to dialysis fluid) is also used in the treatment of peritoneal dialysis-associated peritonitis. After parenteral administration, vancomycin may be associated with renal damage and ototoxicity, and plasma concentrations are of help in minimizing toxicity.

Lithium Lithium is used in the treatment of hypomania, and because toxicity may develop insidiously (p. 27) therapeutic monitoring is essential.

Ciclosporin Ciclosporin is an immunosuppressant used in organ transplantation to prevent graft rejection and for the prophylaxis of graft-versus-host disease. It is also used in the treatment of severe atopic dermatitis, psoriasis and rheumatoid arthritis. It may cause renal damage,

1

DRUG REGULATION AND DEVELOPMENT

THE UNITED KINGDOM MEDICINES ACT (1968) AND THE MEDICINES FOR HUMAN USE REGULATIONS (1994) In the late 1950s thalidomide was introduced as a new hypnotic drug and was widely used in Europe, including in pregnant women. In 1961 an epidemic of appalling proportions occurred in which hundreds of the babies were born with limb abnormalities (phocomelia). The connection with thalidomide administration was soon realized and the drug was soon withdrawn. The nature of this adverse reaction and the public pressure it provoked resulted in the formation, in the UK, of the Committee on Safety of Drugs (the Dunlop Committee) in 1965, and the subsequent development of strict legislation concerning all aspects of drug manufacture, testing and selling. This legislation is contained in the United Kingdom Medicines Act (1968). In 1994, the European Union regulations were implemented in the UK by statutory instrument (the Medicines for Human Use Regulations) and the European Medicines Evaluation Agency (EMEA) was set up in London in 1995 to develop a centralized European procedure for drug evaluation. Each EU member state is required to have a national system for drug safety surveillance (pharmacovigilance) in place. In the UK this system is administered by the Medicines Control Agency (MCA), an executive agency of the

SUMMARY 3 Drug development Synthesis/isolation of new chemical entity Animal studies Pharmacological/toxicological studies Clinical trials Phase 1 Dose-finding tests in healthy volunteers (c 50) Phase 2 Tested in carefully selected patients (c 200) Phase 3 Formal, larger-scale trials (normally up to 2000) Efficacy, safety and comparison with other established therapies Phase 4 Post-marketing studies on a much larger population of patients

13

Department of Health, which is tasked to safeguard public health by ensuring that all medicines on the UK market meet appropriate standards of safety, quality and efficacy. It is advised by an independent Committee, the Committee on Safety of Medicines (CSM), which gives advice to the Licensing Authority (in this case the UK government) on whether new products (new active substances) submitted to the MCA should be granted a marketing authorization. It also helps to monitor the safety of marketed medicines.

DEVELOPING NEW DRUGS

14

Virtually all new drugs are developed within the pharmaceutical industry, because of the enormous financial investment (about £560 million for a new product in 2001) and the huge amount of work involved. Because a patent taken out on a new compound lasts for only 17 years, the 10-12 years (or longer) required for most drugs to reach the stage of a product licence erodes the patent life alarmingly. The industry therefore has to look very carefully at the likely returns on its investment over the short period that it alone can market the drug under patent. Developing drugs for less common diseases becomes commercially unattractive under these circumstances. Apart from the basic chemistry and pharmacology required to develop a new drug, a large amount of toxicological data from animal testing is also required from the pharmaceutical company to satisfy the licensing authority that the drug is unlikely to cause serious adverse reactions in humans. If it is so satisfied, the authority will grant a Clinical Trials Exemption (CTX) or Clinical Trials Certificate (CTC), both of which allow the drug to be tested in patients before an application for marketing authorization is made. In Europe, authorization for a marketing (product) licence can be sought by either a centralized or a decentralized procedure. The former process gives approval throughout Europe and is compulsory for medicinal products derived from biotechnology. The decentralized procedure (or mutual recognition procedure) applies to the majority of conventional medicines and is based upon the principle of mutual recognition by member states. The EMEA provides any arbitration between the member states before the final decision is made, normally by the European Commission. Member states can still, if they wish, give authorization for medicines to be marketed only in their own country. Clinical trials are usually divided into four phases. In phase 1 a small number of healthy volunteers (usually around 50) are given the drug for the first time. The initial doses are very low, and are titrated up to doses likely to show activity, with very careful monitoring for possible toxicity. In the UK, a CTX or CTC is not required for studies in healthy volunteers. If the results are satisfactory, around 200 or so carefully selected but clinically stable patients are given the drug to discover whether pharmacological and therapeutic effects occur in humans (phase

2). In the third phase, formal and larger-scale clinical trials are performed in order to confirm the drug's therapeutic benefit in a broader range of patients (around 2000 on average) with other concomitant conditions, to evaluate its adverse effects and compare it with alternative established therapies. At this point, assuming that the results are satisfactory, an application will be made for a product licence so that marketing and promotion of the drug can begin. It is thus unusual for more than 2500 individuals to have received the drug up to this point in time. Therefore, postmarketing surveillance is crucial to detect any rare (incidence <1 in 1000) ADRs that are unlikely to have been detected during the drug development process. Companies may institute formal (phase 4) prospective cohort studies at this stage, particularly if they wish to examine any potential drug safety issues on a much larger population of patients.

NATIONAL INSTITUTE FOR CLINICAL EXCELLENCE (NICE) This body was set up in 1999 to assess the cost-effectiveness and clinical effectiveness of new and existing technologies, including certain drugs, and to provide patients, health professionals and the public in England and Wales with authoritative, robust and reliable guidance on current 'best practice'. When a cost-effective treatment already exists, this is used as a benchmark against which any existing or newly marketed treatment is appraised, as to whether it is a clinically effective and cost-effective use of NHS resources. Details of the appraisals can be found at http://www.nice.org.uk.

MONITORING ADVERSE DRUG REACTIONS One of the MCA/CSM's most important tools in monitoring safety of marketed medicines is the yellow card system. Set up in 1964 in the wake of the thalidomide tragedy, it now has more than 480000 reports of suspected ADRs that have been submitted on a voluntary basis by doctors, dentists, pharmacists and coroners, and (under statutory obligation) by pharmaceutical companies. If a serious reaction is suspected, or any possible reaction to a newly marketed drug (marked with an inverted triangle in the British National Formulary), a card should be sent to the Medicines Control Agency (MCA), in certain areas via the appropriate Regional Monitoring Centre. A substantial data bank is thereby accumulated and analysed regularly for associations between suspected adverse reactions and individual drugs. Data from the UK scheme and from more than 40 other countries is shared with the WHO Collaborating Centre for International Drug Monitoring in Uppsala, Sweden. Each year more than 150000 such reports are received and the cumulative database now contains over 1.4 million records, so that potential ADR

1

FIG. 1.5 Drugs involved in self-poisoning Number and percentage of admission to the Cardiff Poisons Treatment Unit in 1999 involving the seven most frequently taken drugs/drug groups. (Some patients take more than one drug).

FIG. 1.6 The relationship between age and the incidence of poisoning in children Figures are based on enquiries concerning accidental poisoning made to the National Poisons Information Service (Cardiff Centre) in 1999.

signals can be detected more sensitively. Such spontaneous reporting schemes are relatively inexpensive and extremely cost-effective. Unfortunately, less than 10% of even serious suspected ADRs are reported via these schemes, and they could be even more effective if more health professionals regularly reported their drug-safety concerns.

Whereas death from carbon monoxide poisoning predominates in adults, there are approximately 20 accidental deaths per year from all other agents, particularly tricyclic antidepressants, iron, aspirin and barbiturates in young children. The incidence of accidental poisoning in childhood is partly determined by availability.

DELIBERATE SELF-POISONING DRUG OVERDOSE Approximately 4000 people die each year of poisoning in England and Wales, a quarter of them of carbon monoxide poisoning. Approximately two-thirds of the remaining deaths occur outside hospital. Nevertheless, more than 100000 patients are admitted to hospital in England and Wales each year with poisoning, accounting for 10% of all acute admissions. The unadjusted admission rates in the USA range from 80 per 100000 in patients aged 65-74 years to 400 per 100000 in patients aged between 15 and 24 years. Toxic compounds may be ingested or inhaled either accidentally or deliberately, or may be administered accidentally or deliberately. This forms the basis for a classification of poisoning incidents, separating them into broad epidemiological categories. The major causes of admission due to poisoning are shown in Figure 1.5.

ACCIDENTAL POISONING Accidental poisoning can occur at any age but is much more likely to occur in children under 5 years of age (Fig. 1.6). Peak incidence is around 2 years, when children have become independently mobile, and boys appear to be more at risk. This contrasts with other categories of poisoning, in which women are involved more than men. Fortunately, the mortality from accidental poisoning is relatively small.

Deliberate self-poisoning is commoner in adults who, unlike children, ingest medicinal agents more readily than they do household chemicals and other non-medicinal compounds. Women are more often involved than men, and the vast majority of instances are not intended to be ultimately fatal. Generally they represent a response to an apparently insoluble personal predicament (Ch. 8), but some have a definite suicidal intent. Patients who successfully commit suicide tend to be older and may have been recently divorced or bereaved, may be living alone, suffering from a physical illness, unemployed or abusers of alcohol.

NON-ACCIDENTAL POISONING Non-accidental poisoning usually involves children and is a form of abuse. The parents of such children may have suffered child abuse or deprivation themselves, and often have marital or financial problems or difficulties in managing the child at the time of the poisoning episode.

HOMICIDAL POISONING Although often featured in novels and plays, homicidal poisoning is uncommon. When it does happen, it is often

15

caused by a close relative or friend and generally involves the use of poisons producing chronic, rather than acute, symptoms.

DIAGNOSIS OF POISONING Laboratory tests cannot identify every agent to which an individual might be exposed. A full history and clinical examination is therefore an integral part of the diagnosis. As the patient may be unwilling or unable to give a history on presentation, it is important to talk to relatives or friends, the GP or the ambulance personnel who brought the patient into hospital. Particular difficulties are encountered with the patient who is found alone and unconscious and in whom poisoning is just one of the differential diagnoses. The first priority in the unconscious patient is maintenance of the airway, ventilation and cardiovascular function. When this has been assured, the question arises as to whether the patient is poisoned. Other causes of unconsciousness are discussed in Chapter 24. In deep coma due to drug overdose the pupillary reflexes are often preserved, but the oculovestibular responses (e.g. 'doll's eye response') are absent. Focal neurological signs are rare in overdose and, unless they can be accounted for by previous pathology, these should alert the clinician to the possibility of another cause. Papilloedema due to drug overdose is rare, unless the overdose has been associated with cerebral hypoxia and oedema. If the patient shows no improvement in level of consciousness by 12 hours after admission, another pathology should be suspected.

TABLE 1,8 Physical signs suggestive of overdose or poisoning: head and neck Sign

Possible cause

Dilated pupils

Anticholinergics Tricyclic antidepressants Amphetamines Ecstasy Carbamazepine Sympathomimetics Cocaine LSD Glutethimide Methanol

Constricted pupils

Opiates Organophosphates and carbamates

Strabismus

Carbamazepine

Nystagmus

Barbiturates Carbamazepine Phenytoin Phencyclidine

Perforated nasal septum

Cocaine abuse

Acneiform rash

Solvent abuse

Erythema/burns

Corrosives Paraquat

Mouth dryness

Anticholinergics

Hypersalivation

Parasympathomimetics Organophosphates and carbamates Arsenic Strychnine Clormethiazole Clozapine

CLINICAL EXAMINATION It is sometimes possible to identify the poison by the clinical signs on examination, although it may be more difficult if the patient has taken several agents. The most helpful clinical signs are shown in Tables 1.8 and 1.9.

Mouth

Eyes Pupillary dilatation occurs in poisoning with anticholinergic drugs, such as the tricyclic antidepressants, antiparkinsonian agents, antihistamines and phenothiazines. It also occurs with agents that stimulate the sympathetic nervous system, such as amphetamines, LSD and theophyllines. Unequal pupils are occasionally seen in poisoning. Strabismus is not uncommon, particularly in carbamazepine poisoning. Papilloedema, however, is rare, except as a

1

16

result of methanol, carbon monoxide or glutethimide poisoning. Agents affecting the central nervous system (CNS), particularly phenytoin, may cause nystagmus.

Fig. 1.5

2

Fig. 1.6

Acneiform lesions around the buccal cavity may suggest solvent abuse, O and the characteristic odour of solvent or alcohol may be present on the breath. Examination of the oral cavity may reveal ulceration due to corrosion by acids or alkalis. The mouth may be abnormally dry because of anticholinergic compounds, and increased salivation may be seen in subjects who have taken agents that stimulate the parasympathetic nervous system, e.g. carbamate insecticides and anticholinesterases.

Cardiorespiratory Chest wheeze may be caused by inhalation of irritant compounds: pulmonary oedema can also be caused by these

TABLE 1.9 Physical signs suggestive of overdose or poisoning: trunk and limbs Sign

Possible cause

Bronchospasm

Irritant gases NSAIDs (rare) Arthropod envenomation Scombroid fish

Pulmonary oedema

Irritant gases Paraquat Opioid analgesics Aspirin

Tachycardia

Anticholinergics Sympathomimetics Cocaine Phencyclidine Theophylline Amphetamines Aspirin

Bradycardia

Cardiac glycosides p-Blockers Verapamil Organophosphates and carbamates

Hypertension

Sympathomimetics Amphetamines Phencyclidine MAOIs Clonidine Cocaine SSRIs (rare)

Needle tracks

Drugs of addiction

Absent bowel sounds

Anticholinergics Tricyclic antidepressants Opioids

Bladder distension

Anticholinergics Tricyclic antidepressants

Hypothermia

Phenothiazines Tricyclic antidepressants MAOIs (rare)

Hyperthermia

MAOIs Sympathomimetics Anticholinergics MAOIs SSRIs (rare) Amphetamines Ecstasy Aspirin

Hyperreflexia

Anticholinergics Sympathomimetics Opioids

Blisters

Barbiturates Carbon monoxide Opioids Tricyclic antidepressants

Extensor plantar responses

Anticholinergics Tricyclic antidepressants

agents, or by ingestion of drugs such as opioids or aspirin. There may be tachycardia with poisoning due to anticholinergics, salicylates or drugs that stimulate the sympathetic nervous system. Verapamil, (3-blockers or drugs that stimulate the parasympathetic nervous system, on the other hand, may cause bradycardia. Cardiac arrhythmias may be associated with hypoxia, but are particularly troublesome in patients who have taken tricyclic antidepressants, drugs stimulating the sympathetic nervous system, phenothiazines or quinine. Many agents can cause hypotension, but hypertension is rarer and usually associated with monoamine oxidase inhibitors or clonidine, which in overdose may act like a-adrenoceptor agonists.

1

Abdomen Absent bowel sounds or bladder enlargement suggest the possibility of anticholinergic poisoning. Rectal examination should be performed and may reveal melaena due to poisoning by iron, non-steroidal anti-inflammatories or other agents that damage the gut mucosa. The rectal temperature should be recorded using a low-reading thermometer. This may reveal hypothermia (e.g. with phenothiazine poisoning) or hyperthermia (in poisoning by salicylates, anticholinergic compounds or drugs that stimulate the sympathetic nervous system).

Skin and nails The hands, arms and legs may show needle tracks indicating drug abuse. Blisters over the bony prominences, commoner on the lower limbs, are seen in patients who have taken a barbiturate, opioid or tricyclic antidepressant overdose, 2 or are suffering from carbon monoxide poisoning (Fig. 1.7). Their presence reflects the combination of severe hypoxia and pressure, and they often occur where the patient has been lying against a hard surface. They should alert the physician to the possibility of hypoxic damage of other organs such as muscle, since there is sometimes associated rhabdomyolysis. The nails may show transverse white (Mee's) lines in poisoning with the heavy metals, arsenic and thallium (Fig. 1.8).

Neuromuscular Examination of the limbs normally reveals hypotonia and hyporeflexia in poisoning, but anticholinergic drugs, some opioids and drugs that stimulate the sympathetic nervous system may cause hypertonia, hyperreflexia and, occasionally, myoclonus. Anticholinergic compounds may also produce an extensor plantar response that is usually bilateral. Convulsions may be caused directly by a variety of compounds, including the anticholinergic agents and drugs stimulating the sympathetic nervous system, opioids, mefenamic acid and isoniazid, or may be secondary to hypoxia or hypoglycaemia associated with overdose. Dystonic reactions are much less commonly seen, and are generally

17

FIG. 1.7 Poison blister Blister on the knee of a 25-year-old male who took an overdose of the opiate drug, dextropropoxyphene. He subsequently developed signs of rhabdomyolysis.

FIG. 1.8 Mee's lines in the nails of a patient with thallium poisoning These dystrophic changes causing a white line distal to the cuticle are also seen in poisoning with arsenic.

related to poisoning with drugs with antidopaminergic activity, such as metoclopromide, the phenothiazines and butyrophenones.

The urea and electrolytes are rarely of diagnostic value. Hypokalaemia can be caused by theophylline and drugs stimulating the sympathetic nervous system, or may indicate potassium loss owing to diuretic or purgative overdose. Hyperkalaemia is sometimes seen in severe digoxin overdose. Reduction in the serum bicarbonate may indicate metabolic acidosis and should be investigated by arterial blood gas analysis. The alcohols and glycols, as well as aspirin, may cause a metabolic acidosis, often with a high anion gap (see p. 1114). The presence of both a high anion and an osmolal gap in the absence of detectable ethanol is suggestive of methanol or ethylene glycol poisoning. Profound hypoglycaemia may indicate overdose with oral hypoglycaemic agents or be secondary to overdose with alcohols, salicylates or paracetamol. Clinical judgment must be used in deciding which test will give useful diagnostic or prognostic information. It is useful to store blood for future reference.

LABORATORY INVESTIGATIONS Routine laboratory investigations are sometimes helpful in identifying the drug ingested.

Urine The urine should be examined both macroscopically and microscopically. A specimen should also be stored for possible future analysis even if this is not indicated initially. Red discoloration may be caused by rifampicin or phenindione, 1and haematuria by drugs affecting haemostasis or clotting, or by drugs causing haemolysis, such as chlorates. Drugs that contain methylene blue (e.g. some commercial analgesics) may colour the urine blue or green; a greyish discoloration may indicate poisoning with phenols or cresols. Microscopic haematuria or albuminuria may be the first sign of renal damage. Crystalluria can occur with primidone and some of the sulphonamides, and calcium oxalate crystals may be present in ethylene glycol poisoning.

Blood Brown discoloration of venous blood may indicate methaemoglobinaemia, which can be caused by oxidizing agents such as dapsone. Pink discoloration of the plasma indicates haemolysis, which may be drug induced.

O Fig. 1.7

18

Specific tests An increasing number of compounds can now be detected specifically in either urine (which tends to be more useful as it concentrates the drugs or its metabolites) or blood. Toxicological analyses are not a substitute for clinical judgment and the physician should consider whether the result would alter his or her management.

PRINCIPLES OF MANAGEMENT OF POISONING General supportive measures to prevent complications are usually more important than specific therapies to treat the particular poison, since the latter are not always available.

SUMMARY 4 Principles of management of poisoning • General supportive management, particularly of airway, breathing and circulation (ABC), is of prime importance • Identification of the cause of poisoning can be made by history and clinical and toxicological examination • Decontamination procedures should then be considered to prevent further adsorption of drug remaining in the gastrointestinal tract • Elimination of some drugs can be enhanced. Procedures can produce complications and should only be considered when clearly indicated • General supportive care. Prevention and treatment of complications remains the mainstay of management

SUPPORTIVE MANAGEMENT

• Increased capillary permeability, resulting in a reduced circulating blood volume. In hypotensive patients a central venous pressure line should be inserted. If the venous pressure is low the blood volume should be increased using a plasma expander, such as Haemaccell. If the hypotension does not respond, intravenous (i.v.) dopamine or dobutamine may restore the blood pressure: low-dose dopamine may prevent subsequent acute renal failure (see Ch. 14). Cardiac arrhythmias, which may be a direct effect of the drug or a result of hypoxia and acidosis, may also contribute to hypotension by reducing cardiac output. It is important to correct the predisposing causes if possible before considering specific antiarrhythmic therapy, as many antiarrhythmics may further reduce cardiac output. The choice of an antiarrhythmic agent is usually based on the arrhythmia rather than the specific toxin.

Maintenance of respiratory function

Prevention of complications

Maintenance of an adequate airway and respiration is the first priority in an unconscious patient, as many early deaths from poisoning are due to airway obstruction. Dentures should be removed, the tongue pulled forward and the mouth and pharynx cleared of saliva and vomitus. If the patient has no cough reflex, a cuffed endotracheal tube is necessary to protect the lungs; in other cases a short oropharyngeal tube should be inserted. The patients is then moved into a semiprone position and, if there is any suspicion of inadequate ventilation, oxygen should be administered. Assisted ventilation may be necessary. In less urgent circumstances arterial blood gas analysis is the best guide to assessing the adequacy of ventilation. Drugs that stimulate respiration (e.g. doxapram) should be used cautiously in poisoned patients as they may provoke convulsions or cardiac arrhythmias. The shock lung syndrome (acute respiratory distress syndrome, ARDS; see p. 697) may appear at about 24 hours. The patient begins to hyperventilate but remains hypoxic, and develops a respiratory and metabolic acidosis. A chest radiography may show changes similar to pulmonary oedema and, in severe cases, the lung fields may be completely obscured. Assisted ventilation using positive endexpiratory pressure (PEEP) may help to support the patient through this severe complication. The treatment of aspiration pneumonia is discussed on page 635.

Good nursing care of the unconscious patient is essential. Skin and muscle compression injuries associated with prolonged immobility and hypoxia are a particular problem. The bladder can sometimes be emptied by firm suprapubic pressure, but if it is markedly distended catheterization may be necessary, particularly when measurement of urine production is required. The level of consciousness should be measured using a scale such as the Glasgow Coma Scale (GCS) (Table 1.10). If the conscious level does not improve after 12 hours, this may indicate neurological complications of the overdose or that there is an alternative cause of the

TABLE 1.10 Assessment of consciousness (Glasgow Coma Scale) Score Eyes open

Best motor response To verbal command To painful stimulus

Maintenance of circulatory function Maintenance of circulatory function is essential. Hypotension and reduced peripheral perfusion may be caused by several mechanisms, including: • Depression of the vasomotor centre; • A direct cardiodepressant action; • Increased venous capacitance, leading to reduced venous return to the heart and a consequent fall in cardiac output;

1

Spontaneously To verbal command To pain No response

4 3 2 1

Obeys Localizes pain Flexion-withdrawal Flexion-abnormal (decorticate rigidity) Extension (decerebrate rigidity) No response

6

Orientated and converses Disorientated and converses Inappropriate words Incomprehensible sounds No response

5 4 3 2 1

5 4 3 2 1

Best verbal response

Total

3-15

19

coma. Cerebral oedema is the most life-threatening complication of overdose, and may be caused by hypoxia, hypoglycaemia, hypercapnia or hypotension. It is particularly likely to occur after cardiorespiratory arrest and may respond to dexamethasone or mannitol. Convulsions may also be secondary to hypoxia or cerebral oedema, but otherwise are generally short-lived and do not always require specific therapy. If necessary, intravenous diazepam may be used, but if the patient is vomiting, temporary intubation and mechanical ventilation may be necessary to prevent aspiration pneumonia. Gastrointestinal bleeding or paralytic ileus may be seen in patients with drug overdose. Insertion of a nasogastric tube and regular emptying of the stomach help to reveal bleeding early and prevent the aspiration pneumonia or acute gastric dilatation sometimes seen with paralytic ileus.

MEASURES TO PREVENT FURTHER DRUG ABSORPTION Emptying the stomach can sometimes reduce further absorption of the drug from the gut. Such measures are only indicated if the amount of poison taken is likely to cause serious symptoms, and are of little value if the material ingested has already left the stomach. They should therefore generally only be considered when less than 1 hour has elapsed since ingestion of the toxin, although they may sometimes still be worthwhile if the time of ingestion is unknown.

Gastric lavage Gastric lavage has been the most widely used method of emptying the stomach in adults. It is contraindicated in patients who have taken corrosive agents, unless the danger of systemic toxicity associated with a corrosive agent is greater than the risk of local damage. It is also usually contraindicated after oral ingestion of petroleum distillates, because of the risk of aspiration pneumonia. In the unconscious patient, a cuffed endotracheal tube must always be inserted first. Lavage should be carried out in the left lateral position in a bed or trolley raised at the foot. A large-bore Jaques stomach tube (external diameter

SUMMARY 5 Prevention of further drug absorption

20

• Activated charcoal prevents further absorption of many drugs, if administered sufficiently early, and is the safest and most effective method if the substance is adsorbed by charcoal • Gastric lavage may prevent absorption of the drug if performed within 1 hour of drug ingestion • Emptying the stomach by lavage should not be attempted if the airway cannot be protected • Emesis induced by syrup of ipecacuanha is no longer recommended in children or adults

TABLE 1.11 Substances not effectively adsorbed to charcoal Boric acid Cyanide Ethanol Ethylene glycol Iron salts Lithium salts Malathion Methanol Petroleum distillates Strong acids and alkalis

c 14mm) is introduced into the stomach and the contents removed. Aliquots of water warmed to body temperature are then introduced and removed until the returned fluid is clear of particles. However, gastric emptying by lavage is often relatively inefficient and may be difficult or even dangerous in some individuals (e.g. small children and patients with tricyclic antidepressant poisoning, who may become hypoxic).

Emesis There is little benefit from inducing vomiting. Although syrup of ipecacuanha produces emesis, its effect is delayed for 20-30 minutes and there is no good evidence that it reduces drug absorption significantly in poisoned patients. In addition it has side-effects (e.g. persistent vomiting) which may increase morbidity and make diagnosis more difficult. Salt and water should never be used to produce emesis, as death may occur from hypernatraemia, particularly in small children.

Oral adsorbents Activated charcoal is increasingly being used as it may reduce the absorption of many drugs. Drugs not adsorbed effectively to charcoal are listed in Table 1.11. Charcoal is most effective for agents with high-toxicity taken at low total dose, as 10 times as much charcoal as drug is required to provide optimum adsorption. The charcoal must be given soon after the overdose (normally within 1 hour) to achieve maximum efficacy. A dose of 50-100 g (lg/kg in children) can be administered orally or, in an adult, by nasogastric tube. It is of particular value in accidental poisoning in children, as these cases normally present early enough for the adsorbent to be effective.

ANTIDOTES There are direct antidotes only to about 5% of poisons, so supportive therapy remains the mainstay of treatment for

1

CASE STUDY 1.1 DOSULEPIN (DOTHIEPIN) OVERDOSE A 50-year-old man was admitted to hospital 2 hours after taking 750 mg dosulepin (dothiepin) (conventionalrelease form) and, possibly, 100 mg diazepam. He was very drowsy (Glasgow Coma Scale 9) and tachycardic (pulse 120 bpm, regular). He had dilated pupils (responding slightly to light), exaggerated limb reflexes and bilateral extensor plantar responses. His respiratory rate was 15 per minute. Bowel sounds were present but quiet, and percussion revealed a distended bladder. His blood pressure was 100/70 mmHg. His ECG confirmed sinus tachycardia but showed marked widening of the QRS complex. Questions 1. What other investigations would you ordef ? 2. What treatment would you consider? <

Discussion

Dosulepin (dothiepin) is a tricyclic antidepressant with marked anticholinergic activity in overdose. The clinical features in this patient are consistent with anticholinergic poisoning due to tricyclic antidepressants, although diazepam (a benzodiazepine tranquillizer) could also be contributing to his reduced conscious state. Tricyclic antidepressants can produce cardiac and respiratory depression, with metabolic and respiratory acidosis. Blood gases should be measured urgently and any hypoxia corrected. His temperature should also be checked to exclude hypothermia (which can be caused by a variety of drugs in overdose, including tricyclic antidepressants). Finally, urea and electrolytes should be measured. The presence of significant ECG abnormalities is an indication for intravenous sodium bicarbonate

the vast majority of patients, including those where an antidote exists. Antidotes work in several different ways: • They may speed the metabolism of a toxin, e.g. thiosulphate speeds the metabolism of cyanide by providing sulphur to form non-toxic thiocyanate, which is then excreted in the urine. • If a compound is only toxic when metabolized, antidotes may reduce the rate of metabolism. Ethanol is used in poisoning by methanol and ethylene glycol to reduce the rate of production of toxic aldehydes and acids. • For receptor-mediated toxins, an antagonist that competes for a receptor may be useful, e.g. the use of (3agonists in the treatment of (3-adrenoceptor blocking drugs overdose. • If the toxin competes with a natural substrate, an antidote may increase the natural substrate's concentration, competitively displacing the poison from the receptor site. Inhibitors of acetylcholinesterase reverse some of the features of anticholinergic poisoning in this way. • Antidotes may combine with the poison to form an inactive compound. Thus, specific antidigoxin antibody (Fab) fragments combine with digoxin to produce an inert complex, which can be excreted more rapidly via the kidney. The availability of a specific antidote does not necessarily justify its routine use. Some antidotes are toxic in their

(50mmol), even in the absence of acidosis. Further doses may be necessary, depending on clinical response. Charcoal is of limited value at this stage and may be dangerous if he cannot protect his airway, so it should not be given. Convulsions are normally controlled by diazepam. Flumazenil (a benzodiazepine antagonist) may reverse the effects of any diazepam this patient may have taken, and thus precipitate convulsions and arrhythmias. It should not be given to patients who have taken tricyclic antidepressants. Because of the large volume of distribution of tricyclic antidepressants, forced diuresis, haemodialysis and haemoperfusion are of no proven value in enhancing their elimination. If this patient were to have a cardiac arrest, there is evidence that prolonged resuscitation might be successful.

own right and should only be used if the severity of the poisoning merits their administration. In general, they are most useful early on in the course of poisoning.

ENHANCEMENT OF DRUG ELIMINATION Methods to enhance drug elimination (Table 1.12) are of limited value in the treatment of overdose and are generally only effective when the volume of distribution of the drug is relatively small. This is because drug removal can only occur from the central (intravascular) compartment, which must therefore contain a reasonable proportion of the total body burden of the compound.

Forced diuresis Forced diuresis has been used to enhance the elimination of drugs (such as aspirin) that are excreted predominantly in an unchanged form by the kidney. It acts by increasing the glomerular filtration rate and by reducing the tubular reabsorption of unchanged drug. This is dependent on renal tubular flow, and on the degree of ionization of the drug in the tubular fluid. For weak acids, alkalinization of the tubular fluid by administration of bicarbonate will result in a greater proportion of the drug existing in the

21

TABLE 1.12 Techniques used to enhance drug elimination Enteral adsorbents Activated charcoal Carbamazepine Dapsone Phenobarbitone Quinine Theophylline

Dialysis Barbitone Ethanol Ethylene glycol Isopropanol Lithium Methanol Phenobarbitone Salicylates Haemoperfusion Barbiturates Carbamazepine Glutethimide Meprobamate Phenytoin

Colestyramine (cholestyramine) Digitoxin Phenprocoumon Warfarin

ionized, poorly lipid-soluble form. It is then unable to diffuse back across the lipid membrane of the tubular cell and be reabsorbed. Conversely, acidification of the urine by administration of ammonium chloride, arginine or lysine hydrochloride will increase the ionization of some basic drugs, thereby reducing reabsorption from the tubular fluid. Owing to the risks of fluid overload and electrolyte imbalance, and its relatively limited efficacy, forced diuresis is falling out of favour and is being largely replaced by other, safer, methods of enhancement of elimination.

Dialysis Peritoneal or haemodialysis is useful to remove drugs with low plasma protein binding, poor lipid solubility and relatively low volume of distribution. It is generally reserved for cases where other methods of removal are contraindicated. This process effectively removes salicylates, phenobarbital and barbitone, methanol, isopropanol, ethylene glycol and lithium. Haemodialysis is normally the more effective of the two approaches.

Haemoperfusion Haemoperfusion involves the passage of arterial blood through an extracorporeal column of activated charcoal or ion exchange resin. This effectively removes a number of drugs where the lipid solubility or plasma protein binding reduces the efficacy of removal by dialysis techniques,

1 22

MCQ 1.12-1.14

but a low volume of distribution of the drug is still required. Haemoperfusion is therefore sometimes considered in severe poisoning with barbiturates (including the shortacting barbiturates not removed effectively by haemodialysis) and non-barbiturate hypnotics such as glutethimide, ethcholorvynol, meprobamate, methaqualone and chloral derivatives (e.g. chloral hydrate). Fortunately, these agents are now used less frequently as hypnotics. Haemoperfusion using resin columns is particularly effective in removing theophylline and, to a lesser extent, salicylates. Complications of the procedure include thrombocytopenia, bleeding and possible infection, so haemoperfusion should only be used in patients with severe poisoning who fail to improve despite supportive management and who have developed, or are at high risk of developing, complications of the overdose. Enteral removal Enteral removal is the most recent approach to the enhancement of drug elimination. It relies on the ability of some non-absorbed agents either to remove drugs that have been excreted in the bile but would have been reabsorbed in the gut, or to remove them by 'dialysis' of the splanchnic circulation. Charcoal is the most widely used enteral adsorbent, and markedly enhances the rate of removal of phenobarbitone, digitoxin, quinine, theophylline and dapsone. It is given as an initial loading dose of 50-100 g, followed by repeated doses of 50 g every 4 hours, sometimes with a single dose of a laxative at the beginning to prevent constipation. Oral colestyramine (an anion-binding resin) increases the rate of removal of digitoxin, warfarin and some other coumarin anticoagulants. This non-invasive procedure usually has no major adverse effects, although an antiemetic may be needed in patients who are vomiting as a result of the toxin (e.g. theophylline). 1 SUPPORT SERVICES It is impossible for a clinician to be familiar with all the details of the recommended current management of all toxins. As in other countries, in the UK and Eire Poisons Information Services have therefore been established in several major cities to serve their region(s) (Fig. 1.9). They also contribute to the editing of a database system (TOXBASE), accessible via the Internet, and soon via the NHS Intranet. Its role is to provide health professionals with up-to-date relevant first-line information on the management of poisoning. It can be found on the Internet at http://www.spib.axl.co.uk. TOXBASE also contains advice on holding centres for antidotes, analytical facilities for toxins, slang terms for drugs of abuse, and references on the most important aspects of toxicology. In cases where further advice is necessary, the nearest Poisons Unit in the UK can be contacted via a single national telephone number: 0870 600 6266, or, in Eire, Dublin 837 9964.

National Poisons Information Service Toxbase: http://www.spib.axl.co.uk UK national number 08706006266

SUMMARY 6 Management of benzodiazepine poisoning

1

• Benzodiazepine poisoning is common but death is rare unless other potentiating agents (e.g. alcohol) have also been taken • Flurazepam, tlunitrazepam and triazolam may cause deeper coma than the other agents • Management is generally supportive • Flumazenil, a specific short-acting benzodiazepine antagonist, has been used in the diagnosis of mixed overdose, but may precipitate benzodiazepine withdrawal in dependent individuals, convulsions in individuals with epilepsy and serious arrhythmias in patients who have also taken tricyclic antidepressants

fatal cases each year in which a benzodiazepine is the only drug identified, but they potentiate the effects of other CNS-depressant drugs such as ethanol and can thus contribute to fatal outcome.

FIG. 1.9 Areas served by the UK and Eire Poisons Information Services

On recovery from an overdose, the patient should undergo thorough psychiatric and, if necessary, social assessment (Ch. 8, p. 224). There appears to be a poor correlation between the severity of the overdose and any underlying psychiatric illness. The GP also plays an important part in supporting the patient once he or she has left hospital. Finally, voluntary services often provide useful emotional and practical support to the patient at risk, and details of these local organizations should be made available to each patient.

MANAGEMENT OF SPECIFIC DRUG OVERDOSE The drugs most commonly involved in acute poisoning are hypnotics and tranquillizers, antidepressants and analgesics.

Diagnosis The major effects are on the CNS and include drowsiness, ataxia, dysarthria and nystagmus. Coma then supervenes, although it is rarely deep unless other agents have also been ingested. Recovery is normally within 24 hours, although flurazepam and flunitrazepam may cause deeper and more prolonged coma, and clonazepam has been associated with cyclical coma. Lorazepam causes only modest CNS depression but can be associated with restlessness and hallucinations, particularly in children. Respiratory depression occurs rarely in benzodiazepine overdose, although it may occur in patients with pre-existing respiratory disease, particularly with flunitrazepam. Cardiovascular effects are also generally mild, although slight hypotension may occur. Flunitrazepam may cause more profound hypotension and a sinus bradycardia. Bullous skin lesions have been reported, but are much rarer than after barbiturate overdose. Management Management is supportive: methods to enhance elimination of the parent compound or its active metabolites are of no value. Most patients will recover with no sequelae. Antagonists of benzodiazepines (e.g. flumazenil) have been used in the diagnosis of mixed overdose, or in avoiding the need for ventilatory support in severe poisoning. However, they may precipitate benzodiazepine withdrawal (including convulsions) in benzodiazepine-dependent individuals, convulsions in those with epilepsy, and serious arrhythmias in patients who have also taken tricyclic antidepressants concomitantly. They are also short-acting, so that benzodiazepine-induced coma can recur. They are therefore seldom used.

HYPNOTICS AND TRANQUILLIZERS Barbiturates Benzodiazepines Benzodiazepines are involved in almost half of the drug overdoses in the UK. There are only a small number of

The incidence of barbiturate poisoning has declined markedly over the past 20 years, but this group of compounds is still associated with high mortality.

23

SUMMARY 7 Management of barbiturate poisoning • Barbiturate poisoning is now relatively rare but has a significant mortality • Gastric lavage should be performed within 1 hour and activated charcoal left in the stomach. The airway must be protected • Bullous lesions should alert to the possibility of underlying muscle damage (rhabdomyolysis) • Repeat-dose activated charcoal enhances the elimination of phenobarbitone already absorbed, but not of other barbiturates • Charcoal haemoperfusion should be considered only in severe poisoning with barbiturates

Diagnosis In overdose the barbiturates produce an impaired level of consciousness and, in severe cases, hypotension and respiratory depression. Approximately 5% of patients have bullous lesions on the extremities, particularly over the extensor surfaces, but these are not specific to barbiturate overdose and can occur with other agents producing profound hypoxia. Barbiturates can be detected in the urine. Although their plasma concentrations can also be measured, these correlate poorly with the severity of the overdose, partly because of other confounding drugs such as alcohol, and partly because of tolerance, particularly in those taking the drug chronically. Plasma concentrations are therefore of limited value in the management of a particular case, although very high concentrations in the presence of deep coma and other complications of overdose may indicate the need to enhance the removal of the drug. Management Gastric lavage should be performed if tablets have been taken less than an hour previously, or if the time of ingestion in an unconscious patient is unknown. General supportive management is all that is required in most cases, with particular attention to respiration and cardiovascular function. However, even with these measures the mortality in deep coma (GCS 3) is about 5% and procedures to enhance drug elimination may be considered in such subjects, particularly if complications of overdose are already present. Repeated doses of activated charcoal given by mouth through a nasogastric tube can reduce the half-life of phenobarbitone from over 100 hours to less than 20 hours. This procedure is relatively non-invasive and may replace the use of haemoperfusion for phenobarbitone poisoning. Activated charcoal has not yet been shown to enhance the elimination of any of the other barbiturates, and charcoal haemoperfusion may be necessary in such cases who are deteriorating despite intensive management or who have developed serious complications (e.g. pneu-

1

24

MCQ1.15

monia, hepatic or renal failure). The blisters should be treated in the same way as burns. 1

Non-barbiturate, non-benzodiazepine hypnotics These were once widely used but have now been largely replaced by benzodiazepines. Isolated overdoses of these agents still occur, however. Specific effects may be seen in addition to CNS depression. Glutethimide may cause fluctuating coma, convulsions and acute respiratory failure. Pulmonary oedema and cerebral oedema may also occur, and papilloedema may be a sign of severe overdose. Charcoal haemoperfusion is indicated in severe overdose, together with management of specific complications. Chloral hydrate and its relatives (e.g. dichloralphenazone) may cause gastrointestinal irritation, haematemesis and subsequent oesophageal stricture. Cardiac arrhythmias are a particular problem with this group of agents but generally respond well to (3-blockers. Haemodialysis may remove the active metabolite, trichlorethanol, but charcoal haemoperfusion is probably more effective and may be helpful in very seriously poisoned patients. Clomethiazole (chlormethiazole) is often used in alcoholic patients, although its value is not proven and it may be dangerous when taken in overdose together with alcohol. Increased salivation is a particular problem which may increase the risk of aspiration pneumonia, and regular oropharyngeal aspiration may be necessary. Meprobamate is still used in the treatment of anxiety, and particularly with muscle spasm. Signs of overdose are muscular incoordination and weakness and CNS depression. In severe cases pulmonary oedema may be seen. Haemodialysis may be useful in severe overdose, but haemoperfusion appears to be more effective.

ANTIDEPRESSANTS Antidepressants, particularly those in the tricyclic group, are still commonly taken in overdose and are particularly dangerous.

SUMMARY 8 Overdose of tricyclic antidepressants • Tricyclic antidepressant poisoning is common and life-threatening • Anticholinergic signs predominate in the early stages but coma may supervene • Direct cardiotoxicity may result in ventricular tachycardia or ventricular fibrillation. The ECG should me monitored until recovery • Arrhythmias and hypotension may respond to correction of acidosis, hypoxia, hypothermia and hypovolaemia. Persistent arrhythmias may respond to sodium bicarbonate, even in the absence of acidosis • Prolonged cardiopulmonary resuscitation has been reported to be successful after cardiac arrest

1

CASE STUDY 1.2 BIPOLAR AFFECTIVE DISORDER A 25-year-old woman with bipolar affective disorder presented 4 hours after taking 9g (18 tablets) of paracetamol. Her parents confirmed that she had taken a package up to the bathroom and locked herself in. When she came out there were almost empty paracetamol packages, which had originally contained 20 g of paracetamol. In the emergency department she was vomiting and complained of some epigastric discomfort. She was taking carbamazepine 600 mg daily for her psychiatric condition. Examination revealed no abnormalities. She was 1.55m (5 feet 2 ins) tall and weighted 50kg (110 pounds). Her plasma paracetamol concentration at 4 hours was 150mg/L (normal treatment concentration at 4 hours is 200mg/L). Questions 1. What other investigations would you ask for? 2. What treatment (if any) 1 would you consider?

Discussion Paracetamol can damage the liver and kidney and her INR, U&Es and LFTs should be measured. In some cases the liver transaminase enzyme concentrations (AST or ALT) may rise (as early as 12 hours) before changes in the INR, and can be a useful early marker of hepatotoxicity. In this woman the y-GT and AST may already be slightly raised because of her carbamazepine therapy (this effect is seen in patients on enzyme-inducing drugs). It might also be useful to check her serum carbamazepine concentration to see whether she might have taken these in overdose. Normally, the potentially toxic dose of paracetamol is 150mg/kg or 10 g, whichever is the smaller. This patient weighs only 50kg, so any dose over 7.5 g should cause concern. In addition, carbamazepine is an enzyme inducer and may increase the risk of toxicity at any given dose. In the UK, individuals receiving enzyme-inducing drugs (carbamazepine, phenobarbitone, phenytoin, primidone, rifampicin or St John's

Tricyclic antidepressants The tricyclic antidepressants act by blocking the uptake of noradrenaline (norepinephrine) by neurons both peripherally and in the CNS. They also have an anticholinergic effect and, in overdose, may have direct actions on the heart. Diagnosis The first signs are usually due to anticholinergic effects. Dryness of the mouth and initial stimulation of the CNS are followed by depressant effects. In severe overdose muscle hypertonia, hyperreflexia and extensor plantar responses are followed by hypotonia and hyporeflexia; generalized convulsions may also occur. Other anticholinergic features are warm dry skin, pupillary dilatation, urinary retention and, in severe cases, paralytic ileus. Depression of respiration may be associated with hypoxia and respiratory and metabolic acidosis. Hypothermia may also occur, and respiratory complications such as aspiration and ARDS (p. 697) can further compromise tissue perfusion.

wort) or at increased risk for other reasons (a regular consumer of alcohol in excess, or badly nourished and thus glutathione depleted) are treated according to the 'high-risk' line, which is half the 'normal' treatment line. The timing of the overdose relative to measurement of plasma paracetamol concentration is crucial in informing such treatment decisions. In this case the parents could confirm when the overdose occurred. However, if there is doubt about the timing of the overdose, consider treating the patient anyway. The treatment of choice in the UK is intravenous N-acetylcysteine for 16 hours. Emptying of the stomach is not indicated after 1 hour. Monitor the INR and U&Es at intervals and when the N-acetylcysteine is discontinued. Subsequent psychiatric assessment is also an important part of the treatment of this patient.

The most life-threatening complication is cardiotoxicity. Tricyclic antidepressants resemble quinidine in their action on the heart, producing a decrease in myocardial contractility, prolongation of the PR interval and widening of the QRS complex. In severe cases the T wave may disappear and the ECG may resemble ventricular tachycardia. Severe ventricular arrhythmias occur in some patients (Fig. 1.10), particularly in the early stages, and the ECG should therefore be monitored preferably even before emptying the stomach.

Management If less than an hour has elapsed, activated charcoal should be given by mouth or nasogastric tube to prevent further absorption of drug in those who have ingested more than 4mg/kg, provided the airway can be protected. A second dose of charcoal (50 g) should be considered after 2 hours in patients with features of toxicity who are able to swallow, and multiple-dose activated charcoal may be useful if a modified-release preparation has been taken. Other methods to enhance elimination of tricyclic antidepressants from the body are of little or no value because

25

FIG. 1.10 Rhythm ECG in tricyclic overdose

of the large volume of distribution, with only a small proportion of the drug therefore in the blood compartment. Convulsions may be related to anoxia and will sometimes respond to adequate oxygenation, although they may require treatment with anticonvulsants such as diazepam. The anticholinesterase agent physostigmine has been used to reverse the CNS effects, but it may precipitate convulsions, ventricular tachyarrhythmias and increased salivary and bronchial secretions. It is therefore rarely used. Arrhythmias and hypotension may respond to correction of acidosis, hypoxia, hypothermia and hypovolaemia. Persistent arrhythmias may respond to bicarbonate, even in the absence of acidosis. Sodium bicarbonate (50 mmol) should be given intravenously to an adult with arrhythmias or significant ECG abnormalities. Further doses may be required depending on clinical response. Antiarrhythmic agents should be used cautiously as they may worsen already impaired myocardial contractility, but positive inotropes (e.g. dobutamine) may be necessary to correct hypotension. Prolonged cardiopulmonary resuscitation has been reported to be successful in patients with tricyclic antidepressant poisoning. 1

Monoamine oxidase inhibitors The monoamine oxidase inhibitors (MAOIs) are much less frequently taken in overdose than tricyclic antidepressants. The onset of symptoms may be delayed for up to 12 hours or more after overdose, so observation for this period at least is necessary in those who have taken a potentially toxic dose. They produce features of increased sympathetic activity, with hyperexcitability and hyperreflexia followed by coma and convulsions. Peripheral muscular effects include myoclonus and rigidity leading to hyperpyrexia and rhabdomyolysis. A sinus tachycardia is often present, but the patient may be either hypertensive or, more rarely, hypotensive. Hyperthermia, or more rarely hypothermia, has been reported. Management The agitation, convulsions and hyperthermia may respond to diazepam, and dantrolene may also be useful in controlling muscle rigidity and hyperthermia. Sympathomimetic agents (e.g. dopamine and dobutamine) should be avoided if possible, as they may lead to a severe hyper-

1MCQ1.16 2MCQ 1.17 26

tensive crisis. Severe hypertension should be treated with an a-blocker such as phentolamine.

Selective serotonin reuptake inhibitors (SSRIs) and other antidepressants These agents appear to be safer in overdose than tricyclic antidepressants and their increasing prescription in depression is reflected in an increased use in self-poisoning. Fluoxetine, fluvoxamine, citalopram, paroxetine and sertraline are the agents available in the UK. There is sometimes a latent period of hours before features occur, and alcohol may potentiate toxicity. Drowsiness, nausea, vomiting (particularly with fluvoxamine), abdominal pain, diarrhoea and sinus tachycardia have been reported. Agitation, tremor, dizziness and sweating are also sometimes seen. Rarely, junctional bradycardia, seizures and hypertension have occurred, and influenza-like symptoms may develop after a day or two. In rare instances a condition known as serotonin syndrome has recently been described. Agitation, myoclonus and sweating, shivering tremor and hyperthermia may occur, followed in some cases by rhabdomyolysis, renal failure, disseminated intravascular coagulation and, rarely, death. This condition is similar to, although normally less severe than, neuroleptic malignant syndrome (NMS), which is sometimes seen after therapeutic doses of some antipsychotic agents. Although both syndromes present with varying degrees of mental status changes and autonomic instability, high fever and muscle rigidity predominate in NMS, whereas gastrointestinal dysfunction and myoclonus are more pronounced in the serotonin syndrome. Venlafaxine potentiates neurotransmitter activity in the central nervous system, probably by inhibiting reuptake of serotonin and noradrenaline (norepinephrine). Although probably less toxic in overdose than tricyclic antidepressants, case reports involving seizures, arrhythmias, the serotonin syndrome (alone or in combination with other drugs), rhabdomyolysis and death have been reported in overdose. Nefazodone inhibits reuptake of serotonin and also selectively blocks serotonin receptors. Nausea, vomiting and drowsiness and, on occasion, convulsions have been reported after overdose. Mirtazapine, a presynaptic a2-antagonist, increases central noradrenergic and serotonergic neurotransmission. It may cause sedation in overdose. Reboxetine is a selective inhibitor of noradrenaline (norepinephrine) reuptake. Overdose may be associated with postural hypotension, anxiety and hypertension.

Management Supportive measures are normally sufficient, although activated charcoal will reduce absorption if given in the first hour. Convulsions should be treated with intravenous diazepam if they are frequent or prolonged. Treatment for all these agents is symptomatic and supportive. Management of the serotonin syndrome is also symptomatic and supportive, although drugs with antiserotonin activity, e.g. cyproheptadine, have been advocated in severe cases.

Lithium carbonate Lithium carbonate is widely used in the prophylaxis of manic depressive psychosis. It has a low therapeutic ratio, and signs of toxicity are most often due to excessive therapeutic doses. Diagnosis Nausea, vomiting and diarrhoea are often the first signs of overdose, followed by more severe signs of toxicity, including dysarthria, ataxia, drowsiness, confusion, coma and convulsions. Renal impairment and nephrogenic diabetes insipidus may occur, and liver damage has also been reported. Cardiotoxicity may be first indicated by ECG findings of non-specific ST segment depression, T-wave inversion, atrioventricular block, and prolongation of the QRS and QT intervals. Symptoms and signs may be delayed, particularly in those who have taken a sustainedrelease preparation, so observation for at least 24 hours, with ECG monitoring and 6-12-hourly plasma lithium measurements, is recommended.

Managemenf Consider gastric lavage for non-sustained release preparations if an adult has ingested more than 4g within 1 hour. Sustained-release preparations (the majority) are unlikely to be able to be removed via a lavage tube, however, and activated charcoal does not adsorb lithium. Rehydration is important, but hypernatraemia (which may indicate nephrogenic diabetes insipidus) may require i.v. dextrose until the serum sodium and osmolality become normal. Hypokalaemia should also be corrected. Toxicity correlates relatively well with the plasma lithium level, although there are differences between those who have taken an acute overdose in the naive state and those on long-term lithium therapy. Haemodialysis should be considered in the former situation in patients with signs of severe toxicity, if the plasma concentration exceeds 7.5 mmol/L, but for those with acute-on-chronic overdose or who have chronic accumulation, plasma concentrations above 4.0 mmol/L in patients with severe toxicity indicate consideration of haemodialysis. Because of the relatively large volume of distribution of lithium, plasma concentrations may increase by up to twofold shortly after the end of the dialysis. Repeated dialyses may therefore be necessary to maintain the serum lithium level at concentrations less likely to cause toxicity. 2

ANALGESICS

1

The dangers of analgesics are not widely appreciated. They are the commonest group of drugs involved in overdose in adults, and in children are associated with serious and sometimes fatal poisoning.

Salicylates Salicylate overdose is responsible for approximately 200 deaths each year in the UK. In addition to being acids themselves, salicylates also increase circulating lactic and pyruvic acid concentrations by interfering with normal carbohydrate metabolism. Stimulation of fat metabolism causes increased production of (3-hydroxybutyric and acetoacetic acids, and increased protein metabolism results in increased circulating amino acids. Impairment of reabsorption of these amino acids, by competition with the salicylate for active tubular reabsorption, produces aminoaciduria and increased fluid loss owing to the effect of this solute load. Fluid loss is worsened by the vomiting associated with salicylate overdose, and by the uncoupling of oxidative phosphorylation, which causes hyperpyrexia and increased sweating. Stimulation of the respiratory centre by salicylates causes a respiratory alkalosis, which partly offsets the metabolic acidosis but causes further bicarbonate depletion, electrolyte loss and dehydration. Diagnosis The clinical features of salicylate overdose are agitation, tinnitus and deafness, increased sweating and hyperventilation; nausea and vomiting and epigastric pain may also be present. Respiratory alkalosis may occur initially but is quickly followed, particularly in children, by a metabolic acidosis associated with increased production of lactate, pyruvate and ketones. There may be hypokalaemia and hypo- or, more rarely, hyperglycaemia. Impaired clotting factor and platelet function are sometimes seen, but bleeding is rare. Confusion and coma may occur before death. Patients older than 70 years or younger than 10 are at the greatest risk from salicylate poisoning, and CNS features, the presence of acidosis, hyperpyrexia, pulmonary oedema and salicylate levels greater than 700mg/L, as well as late presentation, are associated with increased mortality. Management Activated charcoal should be administered at the appropriate dose (50 g for an adult, 10-15 g for a child) in adults or children who have ingested more than 250mg/kg body weight of aspirin within the previous hour. Arterial blood gas measurement (in children, capillary gases or venous blood gases would be a suitable alternative) should be performed, and correction of any acid-base abnormalities and rehydration should be sufficient in mild or moderate cases of salicylate poisoning. Urea and electrolytes, international normalized ratio (INR) and blood glucose and plasma salicylate concentration should be measured, and once

27

SUMMARY 9 Salicylate poisoning • Salicylate poisoning is very serious because of the derangement of acid-base and electrolyte balance (metabolic acidosis, respiratory alkalosis and dehydration) • Coma is rare except as a terminal event, and agitation and restlessness, tinnitus, deafness and hyperventilation are characteristic • The very old and young, those with metabolic acidosis, hyperpyrexia, pulmonary oedema, coma or high salicylate concentrations, and those who present late are at greatest risk of a fatal outcome. • Correction of acid-base abnormalities and dehydration is usually sufficient in mild to moderate poisoning, but haemodialysis may be necessary in severe poisoning

the serum potassium concentration is within the normal range, sodium bicarbonate may be given intravenously to enhance the urinary salicylate excretion (optimum urine pH 7.5-8.5). This entails giving 225 mL 8.4% sodium bicarbonate over 2 hours (or 1.5 L of 1.26%) if the salicylate level is greater than 500mg/L in an adult, or 1 mL/kg 8.4% bicarbonate diluted in 0.5% dextrose or saline at 23 mL/kg/h if the plasma salicylate concentration is greater than 350mg/L in a child. Potassium supplements may be necessary to achieve adequate urinary alkalinization, but if this cannot be attained without increasing the blood pH to greater than 7.5, haemodialysis is treatment of choice. It is indicated for severely poisoned patients (generally with salicylate concentrations greater than 700mg/L, although the threshold may be lower in children under 10 years or the elderly, or those with severe acidosis or impaired renal function). Pulmonary oedema may occur as a result of fluid overload or as a direct effect of the drug on capillary permeability. In the former case the pulmonary artery pressure will be raised, but not in the latter. Ventilation with positive end-expiratory pressure may be helpful. 12

Other non-steroidal anti-inflammatory drugs Salicylates form one group of non-steroidal antiinflammatory drugs (NSAIDs) but there are five other groups that can be classified according to chemical structure. These are: • The arylalkanoic acids, which include phenyl propionic acid derivatives such as ibuprofen, naphthyl propionic acid derivatives such as naproxen, and phenyl acetic acid derivatives such as fenclofenac; • The anthranilic acids, which include the fenamates (mefenamic acid and flufenamic acid);

O Case 1.3 O MCQ1.18

28

QMCQ1.19

• The pyrazalone group of compounds, e.g. azapropazone and phenylbutazone; • The cyclic acetic acids, which include indomethacin, sulindac and tolmetin; • The oxicams, at present represented by piroxicam. Diagnosis The arylalkanoic acids, like all NSAIDs, cause headache, nausea and epigastric discomfort, but coma is rare. Gastrointestinal bleeding may occur and, rarely, bronchospasm. Poisoning with anthranilic acids (e.g. mefenamic acid) is more serious and severe haemorrhagic enterocolitis may occur. CNS effects are also commoner and may include convulsions. Overdose with pyrazolones can produce gastrointestinal and CNS toxicity and hepatic and renal damage. The cyclic acetic acids also produce coma, and piroxicam has been associated with coma and convulsions. Management The stomach should be emptied if appropriate. Early administration of charcoal will reduce the absorption of some of the anthranilic and cyclic acetic acids, and may also be of benefit in poisoning with other groups of NSAIDs. Although there is no definite evidence that histamine2-receptor antagonists (e.g. ranitidine) protect against gastrointestinal haemorrhage, they are often administered prophylactically. Convulsions are generally short-lived but can be treated with diazepam. Blood pressure should be measured regularly and hypotension corrected to prevent subsequent acute renal failure.

Paracetamol Paracetamol is a remarkably safe drug in therapeutic doses but in overdose may cause fatal hepatic necrosis and is responsible for approximately 200 deaths per year in the UK. The toxicity of paracetamol in overdose is related to the metabolism of the drug. Approximately 90 % of the drug is conjugated with glucuronide or sulphate in the liver and gut, with only 10% oxidized by the mixed-function oxidase system to a highly reactive intermediate compound (7V-acetylaminobezoneinonimine, or NABQI), which is normally conjugated with glutathione and subsequently excreted as a mercapturate conjugate. In overdose, more NABQI may be produced than can be detoxified, as glutathione stores in the liver become exhausted, and NABQI instead binds to macromolecules in the liver (and sometimes renal tubular) cells, causing necrosis. Patients on enzyme-inducing drugs (see Table 1.5) may produce more NABQI at any given dose of paracetamol and develop liver damage at lower plasma concentrations of paracetamol, and so they are termed 'high risk' and are given antidotal treatment at half the plasma concentrations recommended for patients in general. Other high-risk patients include those with glutathione depletion (e.g. malnourished individuals, including those with HIV, those with pre-existing liver disease and those who consume excess alcohol).

Diagnosis The clinical features are initially mild and consist of nausea, vomiting and epigastric pain. Metabolic acidosis, hypoglycaemia, hypotension and cardiac arrhythmias have been described, particularly with massive overdoses. Incipient liver damage may first be detected by measuring hepatic transaminase enzymes or the one-stage prothrombin time (INR), which may be abnormal even as early as 6-12 hours after the overdose. Liver damage may occur with as little as 12 g of paracetamol and may present clinically with right subcostal pain and tenderness, recurrence of nausea, vomiting, and then jaundice. Loin pain, haematuria and proteinuria after the first 24 hours strongly suggest incipient renal failure. Acute renal failure has also been reported, usually (but not always) in association with severe liver damage. Renal function and electrolytes should be measured daily. A/lanagemenf Activated charcoal should be considered if more than 150mg/kg or 12 g (whichever is the smaller) in total has been taken within 1 hour. The plasma paracetamol concentration should also be checked after 4 hours (earlier concentrations are uninterpretable) and, if it is greater than that on the treatment line (shown in Fig. 1.11), an antidote should be administered. Antidotal therapy should also be used without waiting for the plasma paracetamol concentration if the overdose has been taken at least 8 hours previously, as the efficacy of the antidotes declines markedly with time.

Summary 10 Acetylcysteine and methionine administration

1

Acetylcysteine Dose: by intravenous infusion, in 5% glucose, initially 150mg/kg in 200 ml over 15 minutes followed by 50mg/kg in 500ml over 4 hours, then 100mg/kg in 1000mL over 16 hours. Methionine Dose: by mouth 2.5g initially, followed by three further doses of 2.5g every 4 hours.

Acetylcysteine (see Summary 10) is a glutathione precursor that is usually given intravenously in the UK because, given orally, it can exacerbate vomiting. Acetylcysteine appears to have some efficacy up to 24 hours or even longer after a large overdose. Oral methionine is also used, but is probably ineffective 8 hours or more after an overdose, and the vomiting associated with paracetamol ingestion makes it difficult to retain in the stomach. Anaphylactoid reactions have been reported in around 8% of individuals after the use of acetylcysteine, often during the initial fast infusion. The reaction is probably due to histamine release, and consists of urticaria and bronchospasm, nausea, vomiting, flushing, angioedema, tachycardia, hypotension, respiratory depression and collapse, and in severe cases renal failure and disseminated intravascular coagulation. Death has occurred in patients who have been inadvertently given an overdose (tenfold greater) of acetylcysteine, so the dose of this antidote should be checked carefully. The infusion should be stopped temporarily and an antihistamine given if necessary, with corticosteroids in addition if the reaction is severe. Nebulized salbutamol may be effective in reversing bronchospasm. Once the reaction has settled, acetylcysteine can be infused at a rate of 50mg/kg over 4 hours; further reactions are extremely rare. Blood sugar should be measured and any abnormality treated. The INR should be measured daily. An INR of 4 or more at 48 hours or >6 at 72 hours indicates the likelihood of severe liver failure, in which case the patient should be commenced on a low-protein diet and oral lactulose to help to remove nitrogenous compounds from the gut. A specialist liver unit should be contacted for advice and consideration of possible transfer for management of liver failure. Management of liver failure also includes prophylaxis against infection and peptic ulceration, and management of coagulation abnormalities. Liver transplantation has been successfully performed in some cases, but has a limited role, and patients should be identified and transferred as soon as possible. Factors associated with poor outcome are metabolic acidosis, hypoprothrombinaemia, renal failure and grade 3-4 encephalopathy. Even if acute hepatic failure does occur, the mortality is less than 50% with good supportive management (p. 854). 0

Opioids FIG. 1.11 Plasma paracetamol concentrations above which antidotal treatment is required

Opioids are used not only as narcotic analgesics but also as constituents of some cough suppressants, antidiarrhoeal

29

Summary 11 Management of opiate poisoning • Opiate poisoning is a common, serious medical emergency • Naloxone reverses the effects of all opiates (except buprenorphine if given in sufficient dose, since it is a competitive antagonist) • The short half-life of naloxone necessitates repeated doses or even constant infusion in treatment of opiates with long half-lives of elimination (e.g. dihydrocodeine and methadone) • The association between intravenous opiate abuse and infection (hepatitis B and HIV should be recognized and appropriate safeguards taken)

products (e.g. diphenoxylate) and veterinary tranquillizers. The one most often involved in poisoning is dextropropoxyphene, which is combined with paracetamol in compound preparation. Opioids are being increasingly misused in society for their effects on mood, being administered intravenously as well as orally. Diagnosis Respiratory depression is the commonest cause of death in opioid overdose. The pupils are often pinpoint. Depression of the CNS eventually leads to coma and convulsions. The muscles may be hypotonic, but muscular twitching and hyperreflexia have been described. Hypotension is common, but hypertension has been described after pentazocine overdose. Non-cardiogenic pulmonary oedema may occur, particularly in methadone overdose. Management If the opiate has been taken orally activated charcoal may be of value in the first hour, or longer if the preparation is sustained release. Naloxone is a competitive antagonist that is effective in reversing the symptoms of opiate overdose with all compounds except buprenorphine (a partial agonist and not completely antagonized). It is best given intravenously (0.8-1.2 mg for an adult, and 0.01 mg/kg body weight for children), and the dose repeated if there is no response within 2 minutes. Those who have taken very large overdoses, particularly of a long-acting agent such as dihydrocodeine or methadone, may need further doses to competitively antagonize the opioid activity, and continuous intravenous infusion of naloxone, together with intensive monitoring, may be necessary. Assisted ventilation with positive end-expiratory pressure may be required if pulmonary oedema occurs. In buprenorphine poisoning naloxone may have only limited benefit and, if doxapram fails to stimulate respiration sufficiently, assisted ventilation may be necessary. Naloxone may precipitate acute opioid withdrawal, with abdominal pain, diarrhoea and piloerection, but the dan-

1MCQ1.20 2MCQ1.21 30

gers of opioid overdose are greater and life-saving treatment should not be withheld.

RECREATIONAL DRUGS In addition to opioids such as diamorphine (heroin), other drugs are being increasingly widely abused in society and may be dangerous in overdose. Amphetamines are sympathomimetic drugs of abuse. Street amphetamine usually contains up to 5% amphetamine along with inert fillers or other stimulants, but purity varies widely. Initial CNS stimulation (agitation and dilated pupils) may be followed by marked hypertonia and hyperreflexia (>39°C) and generalized convulsions. Hyperpyrexia and rhabdomyolysis may also occur. Sinus tachycardia is usual but supraventricular and ventricular arrhythmias, including ventricular tachycardia, may subsequently occur. Hypertension may be a feature and hypokalaemia may sometimes be profound. The stomach should be emptied within 1 hour and activated charcoal given. Sedation with diazepam may be necessary. Sinus tachycardia with adequate cardiac output is best left untreated, but (3-blockers may be needed. cc-Blockers or labetalol are used to manage hypertension. Diazepam is normally effective in managing convulsions, but if they are more frequent, or if vomiting is also a problem, it may be preferable to paralyse and ventilate the patient. Ventricular arrhythmias occurring in a patient who is having convulsions may respond to disopyramide. Ecstasy (3,4-methylenedioxymethamphetamine, or MDMA) is an amphetamine-like compound with a similar spectrum of toxicity. However, there is marked inter- and intraindividual variability in susceptibility, and deaths have occurred after only a single tablet in those who have previously tolerated similar or greater doses. Signs of severe toxicity include delirium, coma, convulsions, and supraventricular and ventricular arrhythmias. Hypertonia and hyperpyrexia may result in rhabdomyolysis, metabolic acidosis and acute renal failure. Hyponatraemia, disseminated intravascular coagulation, hepatocellular necrosis, adult respiratory distress syndrome and cardiovascular collapse have also been described. Death may occur from intracerebral haemorrhage. Treatment is largely symptomatic and supportive, with correction of metabolic acidosis and dantrolene (or paralysis and ventilation) for the hyperpyrexia. 1 Cocaine may be taken by the oral or intravenous route, or by inhalation. Toxicity is characterized by agitation, tachycardia, tachypnoea, sweating, delirium and hallucinations. Hypertension, hyperpyrexia, generalized convulsions, stroke, cardiac ischaemia, myocardial infarction and arrhythmias, metabolic acidosis, rhabdomyolysis and cardiorespiratory arrest may also occur. Treatment is symptomatic and supportive. Hypertension may best respond to a-blockers or combined a- and p-blockade (e.g. with labetalol). The risk of severe toxicity is particularly high

in 'body stuffers' who, on the verge of arrest, swallow the evidence, rather than in 'body packers', who are hired specifically to smuggle drugs and who have time to wrap them carefully (e.g. in a condom). Lysergic acid diethylamide (LSD) in overdose may cause acute panic attacks, delirium, hallucinations, seizures and coma, respiratory arrest and metabolic acidosis. Treatment is symptomatic and supportive, although features of psychosis may last for several days. Panic often responds to diazepam. Phenothiazines are best avoided in treating the psychosis, as they may reduce seizure threshold. y-Hydroxybutyric acid (GHB),or liquid ecstasy, is a 'psychedelic' agent that has recently become popular. It can cause coma, sinus bradycardia and hypotension in overdose, as well as respiratory depression leading to respiratory arrest. Treatment is symptomatic. The bradycardia may respond to atropine (1.2mg intravenously for an adult, 0.02mg/kg for a child). 2

ANTICONVULSANTS The incidence of psychiatric abnormalities is higher in epileptics than in the non-epileptic population. The availability of anticonvulsants to some epileptic patients results in their being taken relatively frequently in overdose. Barbiturates and primidone, which are converted into phenobarbitone and phenylethylmalonamide in the body, are described on page 23.

Phenytoin Phenytoin is one of the most widely prescribed anticonvulsants. Acute overdosage leads to ataxia, dysarthria, horizontal nystagmus, drowsiness and, in severe cases, coma. Respiratory depression is a rare but potentially serious complication, which is generally only seen after massive overdose. The stomach should be emptied within 1-2 hours of ingestion and activated charcoal will help to prevent further absorption. Because phenytoin is highly protein bound, only charcoal haemoperfusion will effectively remove the drug from the circulation. This should, however, be reserved for patients with severe overdose. Rarely, cases of permanent cerebellar damage with persistent ataxia after phenytoin overdose have been described.

Carbamazepine Carbamazepine is widely used for the treatment of generalized and partial seizures. The drug is structurally similar to tricyclic antidepressants and overdose can be associated with anticholinergic effects, including dry mouth, hyperreflexia and generalized convulsions. Nystagmus, strabismus, diplopia, ataxia, hypotension and respiratory depression have been noted, and cardiac conduction disturbances may also occur. The stomach should be emptied if appropriate, and activated charcoal administered. Repeated doses of charcoal speed the elimination of drug in

normal volunteers and may be effective in severe overdose. Effective removal of the drug by haemoperfusion has also been reported.

1

Sodium valproate Sodium valproate poisoning is associated with coma and respiratory depression. The drug is rapidly absorbed, and supportive measures form the mainstay of treatment.

CARDIOVASCULAR DRUGS Anti hy per ten si ves The major adverse effect is hypotension. With clonidine, hypotension is often associated with bradycardia, but severe hypertension caused by the drug's partial a-agonist activity is sometimes a major problem. Bradycardia often responds to atropine, and a-adrenergic blocking drugs have been used to treat the hypertension. (3-Receptor blocking agents also cause bradycardia and hypotension, and may also cause drowsiness, coma and convulsions in overdose. In hypotension atropine is used to increase cardiac output, but large doses are often required (3mg intravenously in an adult and 0.04mg/kg for a child). Glucagon is used if atropine is ineffective. An intravenous bolus of 50-150mg/kg body weight dissolved in dextrose may be effective but may need to be followed by the same dose or an intravenous infusion of 4mg/h until the patient's condition is satisfactory. Glucagon may induce vomiting and the airway needs to be protected. Pi -Agonists such as prenalterol are also effective, although large amounts may have to be given intravenously to competitively antagonize the (3-blockade. pY Agonists (e.g. salbutamol) may be necessary if bronchospasm is present, and the blood sugar should be monitored to detect and treat hypoglycaemia. Convulsions normally respond to intravenous diazepam (0.1-0.3mg/kg body weight). Calcium antagonists (especially verapamil) may also cause severe hypotension and bradycardia and conduction abnormalities and, in severe cases, asystole. If atropine is ineffective in treating the bradycardia, calcium gluconate (10-20 mL of a 10% solution in an adult) may be required. Pacing may be necessary to treat the bradyarrhythmias, and dopamine and glucagon (and, in resistant cases, insulin and dextrose) have been used to give inotropic support.

Digitalis glycosides Poisoning with the digitalis glycosides is rare but potentially very serious. Nausea, vomiting and diarrhoea are early features, but the major life-threatening complications are hypotension and cardiac conduction disturbances, with brady- and tachyarrhythmias. Hyperkalaemia is often present and mirrors roughly the severity of the overdose. This should be treated with glucose and insulin, or a sodium resonium ion exchange resin, or by haemodialysis

31

SUMMARY 12 Digitalis glycoside poisoning • Cardiac glycosides may cause nausea, vomiting, diarrhoea and generalized malaise • Cardiac effects include bradycardia with PR and QRS prolongation. Sinus arrest and varying degrees of AV block may also occur • Hyperkalaemia and metabolic acidosis may occur in severe cases • Multiple doses of charcoal enhance elimination by interrupting enterohepatic recirculation • Prophylactic pacing should be considered if hyperkalaemia or heart block occur • Severe intoxication is often effectively treated with digoxin antibodies, the dose depending on the amount taken and the serum glycoside concentration

if necessary. Metabolic acidosis may also occur. Both digoxin and digitoxin have enterohepatic recirculation processes and repeated doses of activated charcoal (see p. 24 for dosage) will speed elimination. Continuous monitoring of cardiac rhythm is essential: bradycardia sometimes responds to intravenous atropine (1.2mg for an adult or 0.02mg/kg for a child), repeated as necessary until atropinization occurs. More severe bradyarrhythmias may require temporary cardiac pacing, particularly if heart block or hyperkalaemia are present. Ventricular tachyarrhythmias often respond to lignocaine, but phenytoin and intravenous (3-blockers have both been used in resistant cases. Digoxin-specific Fab antibody fragments may be necessary in the treatment of severe digoxin and digitoxin poisoning. The dose of antibody (Digibind) is about 60 times the estimated body load of digoxin or digitoxin (the amount of drug in milligrams ingested multiplied by 0.8) rounded up to the nearest 40 mg. This can be calculated from the amount taken, or the plasma concentration of the drug.

RESPIRATORY DRUGS Theophylline Theophylline is the respiratory drug most often taken in overdose. Nausea, persistent vomiting and epigastric pain may be followed by agitation, hallucinations and convulsions. Hypertonia may be associated with rhabdomyolysis. Supraventricular and ventricular arrhythmias with hypokalaemia (often profound) may follow a sinus tachycardia. Signs may be delayed for up to 12-24 hours in those who have taken a sustained-release preparation. The stomach should be emptied to prevent further absorption and charcoal (50 g) left in the stomach. The metabolic abnormalities should be treated appropriately, 1MCQ1.22 2MCQ1.23 32

but convulsions and hypertonia may respond to diazepam administered intravenously. Supraventricular and ventricular arrhythmias may respond to the use of adrenoceptor blockers (e.g. metoprolol),but the metabolic abnormalities may only respond to agents that also possess ( 2-antagonist activity (e.g. propranolol). However, (3blockers should not be used in asthmatic patients because of the risk of precipitating bronchospasm. Disopyramide or verapamil have been used successfully to treat arrhythmias in such patients. Repeated oral administration of charcoal has recently been shown to reduce markedly the half-life of elimination of theophylline. Persistent vomiting may mean that an antiemetic such ondansetron (8 mg slowly intravenously in an adult) will be necessary to allow the charcoal to be retained. Haemoperfusion is an efficient way of enhancing the removal of theophylline from the circulation. It should be considered in severely poisoned patients, particularly those with major arrhythmias, hypotension or convulsions, and if repeated-dose charcoal is not feasible. 1

Ephedrine, salbutamol and terbutaline Ephedrine, salbutamol and terbutaline are generally safer than theophylline in overdose, but can cause all the same features in severe cases. Treatment is broadly similar to that of theophylline overdose: again, -blockers must be avoided in asthmatic subjects.

DRUGS WITH ANTICHOLINERGIC ACTIVITY In addition to tricyclic antidepressants many drugs have anticholinergic activity, including some used to suppress gastrointestinal motility (e.g. propantheline), some antiparkinsonian drugs (e.g. amantidine and orphenadrine) and some antihistamines (e.g. chlorphenamine (chlorpheniramine) and promethazine). The effects of these drugs are similar to those caused by overdosage with tricyclic antidepressants, but they are less likely to cause arrhythmias. The anticholinesterase physostigmine is no longer recommended in anticholinergic poisoning because of severe adverse effects. Treatment is supportive and symptomatic.

ANTIDIABETIC DRUGS Poisoning with antidiabetic drugs and insulin produce all the features of hypoglycaemia. The management is described in Chapter 19, page 1020. Sulphonylureas (particularly those with long half-lives of elimination (e.g. chlorpropamide) are of particular concern, as hypoglycaemia may be a problem for several days and they are sometimes resistant to glucose and glucagon. Severe hypoglycaemia has therefore been treated with diazoxide, which

inhibits insulin release and raises circulating plasma catecholamines. At the dose used (5 mg/kg body weight/24 h by mouth in two or three divided doses), it rarely produces hypotension. Diazoxide may be the treatment of choice if there is a risk of hypoglycaemia over several days. Metabolic acidosis may occur after sulphonylurea poisoning, but is more common after poisoning with the biguanide metformin. ©

atrophy. Most patients recover some vision, although some are left with permanent visual field defects. The management of quinine poisoning is supportive. Stellate ganglion block has been advocated in the past but has no proven benefit. As quinine has a large volume of distribution and is predominantly metabolized by the liver, methods to enhance elimination are of limited value, but repeated dose of oral charcoal may be beneficial.

1

Chloroquine ANTIBIOTICS The penicillins and cephalosporins are generally safe in overdose. Sulphonamides may cause bone marrow depression and renal damage (and haemolysis in glucose 6-phosphate dehydrogenase-deficient patients). Maintenance of a high urine output and alkalinization of the urine may prevent sulphonamide-induced renal damage, by preventing crystallization of the drug in the renal tubule. Calcium folinate may help to reverse bone marrow depression induced by sulphonamide or trimethoprim. Isoniazid can cause CNS symptoms with coma and convulsions. Pyridoxine can be used to treat or prevent both convulsions and the metabolic abnormalities: it should be given in a dose of 1 g for every gram of isoniazid ingested. Rifampicin overdose may be associated with orange discoloration of the skin (although this colour can be removed by washing). Treatment is usually supportive. Although most patients recover completely, occasional sudden deaths have been reported, presumably from cardiac arrhythmias. The antileprotic drug dapsone (also used in the treatment of dermatitis herpetiformis) is dangerous in overdose, causing methaemoglobinaemia and sometimes convulsions and coma.

ANTIMALARIALS Quinine Quinine, which is also used to treat night cramps, is the antimalarial agent most often taken in overdose and also the one most likely to cause serious toxicity. Children may take tablets belonging to a family member. Early symptoms of overdose are similar to those of aspirin, with nausea, vomiting, bowel disturbances, tinnitus and deafness. Initial respiratory stimulation may be followed by respiratory depression and respiratory arrest. Quinine can cause brady- and tachyarrhythmias, with death from ventricular fibrillation. Its effect on vision is dramatic. Visual acuity is initially impaired, and this may progress to constriction of visual fields or even complete blindness. Retinal examination reveals constriction of the arterioles initially, followed by retinal oedema and sometimes optic

Chloroquine is widely used as prophylaxis against malaria and can be purchased without prescription. Its toxicity is similar to that of quinine, and visual disturbances are also sometimes seen. As little as 5g may be fatal. Ventricular arrhythmias often occur early (within the first 3 hours) and include intraventricular conduction defects with a wide QRS, and prolongation of the QT interval. Ventricular tachycardia and fibrillation are early arrhythmias, and torsade-de-pointes sometimes occurs later. Very large doses of diazepam (2 mg/kg body weight intravenously over 30 minutes) may reduce mortality, but should be considered only when facilities for assisted respiration are available. Chloroquine is also predominantly metabolized by the liver and has a high volume of distribution; thus, none of the techniques for enhancing removal appears to be effective.

ANTICOAGULANTS The major risk of oral anticoagulant poisoning is haemorrhage: this occurs several days after the overdose because of the delayed effect of these agents. In cases of large overdose, and where the INR is raised at 36-48 hours, vitamin K (10 mg twice daily by slow intravenous injection, 250 g/kg body weight for a child) should be administered and may be necessary for up to a week. Repeated oral administration of colestyramine (cholestyramine) (4 g three times daily) may enhance the removal of warfarin and phenprocoumon, and also reduce the length of time for which vitamin K is required. Rodenticides contain only small amounts of oral anticoagulant and rarely cause toxicity.

POISONING WITH GASES AND VOLATILE SOLVENTS CARBON MONOXIDE Carbon monoxide is still the most common cause of fatal gas poisoning in the UK, despite the substitution of natural gas for coal gas in general use. Deaths occur either accidentally (often by inhalation of fumes during fires) or deliberately, from inhalation of car exhaust fumes in a confined space. Methylene chloride, which is widely used in

33

paint removers, is metabolized by the body to carbon monoxide and causes symptoms identical to those of direct carbon monoxide exposure. Diagnosis The early symptoms of chronic or acute exposure are headache, mental agitation, nausea and vomiting. The characteristic cherry-pink colour of the skin is rarely present during life, but the patient is not cyanosed, despite considerable hypoxia because of the production of carboxyhaemoglobin. Cerebral oedema and myocardial ischaemia or infarction may also occur. Management Oxygen should be given in as high a concentration as possible. There is no clear evidence that hyperbaric oxygen affects outcome, and its use should be discussed with the regional poisons unit. Patients who have been unconscious after exposure, who have a carboxyhaemoglobin concentration over 20% or who have neurological or psychiatric features, are most likely to obtain any possible benefit. Hyperbaric oxygen is not always immediately available, and transfer to such a facility should not interrupt the treatment of severely poisoned patients. Oxygen should be administered until the carboxyhaemoglobin concentration in blood falls below 15%: this may take up to 2 days. Mannitol, Ig/kg intravenously over 20 minutes, may be considered if there is evidence of cerebral oedema. Most patients who die of carbon monoxide poisoning do so before they reach hospital; those who do arrive in hospital usually survive. Unfortunately, permanent neuropsychiatric sequelae may be seen, including intellectual and personality deterioration, parkinsonism and, in rare cases, akinetic mutism. These severely affected patients often show areas of low density in the area of the globus pallidus on CT brain scans. 1

IRRITANT VAPOURS AND GASES

Carbon disulphide, hydrogen sulphide, sulphur dioxide, hydrochloric acid and ammonia are all highly irritant to the eyes and respiratory tract. Cough and chest pain may be followed by dyspnoea, bronchospasm, haemoptysis and pulmonary oedema. Management

Treatment is supportive, with bronchodilators for bronchospasm, oxygen for hypoxia and, in severe cases, assisted ventilation (with PEEP if pulmonary oedema is present). Steroids, diuretics and antibiotics have been used, but because the complications are direct effects of the gas, they

SUMMARY 13 Cyanide poisoning • Cyanide is absorbed through the skin, by inhalation or oral absorption • The clinical features are those related to severe hypoxia and metabolic acidosis • Oxygen in high dose should be given and the metabolic acidosis corrected • Dicobalt edetate is the antidote of choice if the diagnosis is certain • Sodium thiosulphate, together with sodium nitrite, is also an effective antidote

are unlikely to be of benefit. The metabolic acidosis may improve with adequate oxygenation, but if this fails sodium bicarbonate may have to be administered. Ophthalmologic advice should be sought to diagnose and treat corneal abrasions.

CYANIDE Cyanide may be inhaled in the form of hydrogen cyanide, sometimes in fumes caused by burning plastic. Symptoms appear within a few minutes, first with headache, agitation, confusion leading to coma, cardiovascular collapse and respiratory arrest. Cyanide may also be taken orally, in which case symptoms may occur much later, particularly if food in the stomach delays gastric emptying. Management Oxygen in high concentrations should be administered and assisted ventilation performed if necessary. Dicobalt edetate is the most effective antidote, but it may cause hypotension and chest pain on intravenous administration (300 mg intravenously over 1 minute), and should therefore only be used if the diagnosis is certain. It acts by forming inert complexes with cyanide, thereby preventing the cyanide-induced inhibition of cellular oxidizing enzymes. Sodium thiosulphate (25 mL of 50% solution intravenously over 10 minutes for an adult; 400mg/kg for a child) acts as a sulphur donor to allow conversion of cyanide to thiocyanate via the enzyme rhodanase. Subsequent administration of sodium nitrite (10 mL of 3% solution intravenously over 5-20 minutes for an adult; 0.13-0.33mL/kg of 3% solution for a child) converts haemoglobin to methaemoglobin, which chelates the cyanide ion. The two treatments act synergistically. They are generally used if there is doubt about the diagnosis, there is no response to dicobalt edetate, or if the latter is not immediately available.

HYDROCARBONS 1

34

MCQ1.24

Poisoning with hydrocarbons is nearly always accidental, generally from industrial or domestic exposure, but also increasingly from solvent abuse.

Aliphatic hydrocarbons The aliphatic hydrocarbons appear to be less toxic than the aromatic group. The shorter-chain aliphatic hydrocarbons are volatile, so that poisoning is usually from inhalation. High concentrations may cause asphyxia, and the butane in lighter fluid has been used to produce euphoria by solvent abusers. Coma may occur, with death from respiratory depression or ventricular fibrillation. Renal and liver damage may develop if the patient survives. Pulmonary and cerebral oedema have also been reported. Inhaled hydrocarbons may also produce myopathies, neuropathies and permanent neuropsychiatric damage, although some of these effects may be related to additives to the hydrocarbon. Ingestion of liquid aliphatic hydrocarbons can produce effects similar to the above, due either to a systemic effect or, more likely, to aspiration of the low surface tension solvents. Pyrexia, cough and hyperpnoea may be followed by basal crackles in the chest, consolidation, and collapse involving predominantly the middle and lower zones of the lung. Treatment of solvent inhalation or ingestion is generally supportive. Gastric lavage or emesis should not be performed after ingestion because of the risk of chemical pneumonitis. There is no evidence that steroids and antibiotics are effective in the pulmonary complications, but in severe cases ventilatory support may enable the patient to recover.

Chlorinated aliphatic hydrocarbons Chlorinated aliphatic hydrocarbons are widely used as solvents in industry. In addition to the effects described with other hydrocarbons, the chlorinated aliphatics are also more likely to produce hepatic and renal damage (particularly carbon tetrachloride and chloroform), haemolytic anaemia and aplastic anaemia. Methylene chloride is metabolized by the body to carbon monoxide and its systemic toxicity is that of this metabolite, although it is itself locally corrosive. It has been suggested that acetylcysteine may inhibit the metabolism of chlorinated hydrocarbons to reactive intermediate metabolites and thereby reduce hepatotoxicity if given early.

Aromatic hydrocarbons Aromatic hydrocarbons are less often involved in poisoning. Benzene is the most toxic of these agents and, in acute overdose, produces effects similar to those of the aliphatic hydrocarbons. Chronic exposure may lead to aplastic anaemia and leukaemias. Toluene and xylene are similar in structure and effects to, but less toxic than, benzene: any marrow toxicity is probably related to benzene contamination. Benzene contamination is probably also responsible for the haematological problems sometimes seen with chronic petrol inhalation.

1

FIG. 1.12 Blood concentrations after oral administration of ethyl alcohol (2 ml/kg) The concentration declines in a zero-order fashion at an average rate of 190mg/L each hour.

OVERDOSE WITH ALCOHOLS ETHANOL (ETHYL ALCOHOL) (See also ch. 24 p. 1435) Ethanol, present in alcoholic drinks, is the alcohol most often taken in overdose. It is also found as a solvent in some cosmetic and antiseptic preparations. Its major toxicity is related to its central nervous depressant effects. The loss of the gag reflex max result in aspiration of stomach contents. Alcohols inhibit gluconeogenesis and may cause profound hypoglycaemia, particularly in children. Blood ethanol concentrations give a rough guide to the severity of overdose, but may be misleading because of tolerance in chronic alcohol users. Severe intoxication with stupor and marked incoordination is generally associated with blood concentrations above 3000 mg/L. The rate of metabolism is shown in Figure 1.12. Management Because alcohol is rapidly absorbed, lavage is unlikely to be of benefit and the compound is poorly adsorbed by activated charcoal. Metabolism is saturated even at low doses. The treatment is usually symptomatic and supportive, with correction of hypotension and hypoglycaemia. Protection of the airway is paramount. Although haemodialysis will remove alcohol efficiently, it is only indicated in severe and complicated cases of poisoning. Fructose accelerates the metabolism of alcohol, but it can potentiate the metabolic acidosis already caused by the alcohol and should not be used. Intravenous dextrose may be necessary to treat hypoglycaemia. The problems of alcoholism are discussed on page 247.

METHANOL Methanol is much more toxic than ethanol and is widely used as a solvent. Methylated spirit is 5% methyl alcohol

35

and 95% ethanol. The mode of toxicity is not clear, but accumulation of formaldehyde and formic acid is partly responsible. As little as 10 mL of methanol may cause serious illness. The initial features are those of mild inebriation, but 8 or more hours later the patient may develop abdominal pain and vomiting, dilated and unreactive pupils, a metabolic acidosis and, sometimes, signs of cerebral oedema. The optic disc is at first hyperaemic and then becomes oedematous; vision is progressively lost. Management The stomach should be emptied within the first hour of methanol ingestion; activated charcoal does not adsorb significant quantities. Intravenous sodium bicarbonate should be used to keep the arterial pH above 7.2, but metabolic acidosis may be resistant to treatment. Diazepam is indicated if convulsions are a problem. Ethanol competitively inhibits the metabolism of methanol to formaldehyde and formic acid, and is used in less serious poisoning. Haemodialysis may be necessary if the blood methanol concentration exceeds 500 mg/L, or if there is clinical deterioration despite other measures. Haemodialysis also removes ethanol, so that ethanol administration during this procedure has to be increased. Visual impairment is often permanent, although prompt treatment reduces its severity.

ISOPROPANOL (ISOPROPYL ALCOHOL)

Isopropanol is found in rubbing agents, sterilizing fluids and disinfectants, window cleaning fluids and some aftershave lotions. Features of poisoning resemble those of ethyl alcohol poisoning, but because some of it is metabolized to acetone, ketonuria and a characteristic smell of acetone on the breath may be observed. Management

The stomach should be emptied within an hour of ingestion, but treatment is otherwise symptomatic. Haemodialysis will remove isopropyl alcohol but is normally only indicated in severe poisoning.

overdose, but ophthalmoplegias, papilloedema and subsequent optic atrophy may also occur and the patient may become comatose and develop convulsions. After 12 hours respiratory and cardiovascular complications are seen, together with a metabolic acidosis, hypocalcaemia and hyperkalaemia. If the patient survives this phase, he or she may develop acute renal failure after approximately 24 hours. Ethylene glycol is not itself toxic but is metabolized by alcohol dehydrogenase to aldehydes, glycolates, oxalates and lactate, which are responsible for the toxic features. Management The stomach should be emptied within the first hour and arterial blood gases and serum calcium measured. Activated charcoal does not adsorb ethylene glycol. Any metabolic acidosis may respond to intravenous bicarbonate, and calcium gluconate may be necessary to treat hypocalcaemia. Ethylene glycol metabolism can be inhibited by administration of alcohol in doses sufficient to attain blood ethanol concentrations between 1000 and 2000 mg/L. In severe cases (ethylene glycol concentration >500mg/L, or clinical condition deteriorating despite other measures) haemodialysis is necessary to remove ethylene glycol metabolites. Ethanol administration may also be necessary for up to 4 days after the overdose to reduce their production. Dialysis may be required for renal failure. Fomepizole has recently been licensed for the treatment of ethylene glycol poisoning. 1 POISONING WITH CORROSIVE AGENTS

Fortunately, poisoning with corrosive agents is rare, and most incidents are accidental rather than deliberate. The majority of agents produce local tissue damage by protein denaturation, but some (e.g. paraquat) also have serious systemic toxicity. The patterns of damage with acids and alkalis are different and are therefore discussed separately.

ACIDS ETHYLENE GLYCOL

Antifreeze solutions contain ethylene glycol, a solvent that is sweet-tasting and therefore not unpleasant to drink. Death may occur after ingesting as little as 100 mL of ethylene glycol which, like alcohols, is metabolized by alcohol dehydrogenase. CNS toxicity is similar to alcohol

1

36

MCQ1.25

Acids produce severe pain on ingestion. They may cause asphyxia from epiglottal oedema and, although oesophageal damage has been described, they are more likely to cause gastric perforation, particularly in the region of the antrum. Late stricture formation is also a possibility. Lavage should not be performed unless the agents also have serious systemic toxicity. Any metabolic acidosis should be treated with intravenous bicarbonate, but weak alkalis should not be given by mouth. Endoscopy is of limited value in the acute phase and may cause perforation; in later stages, however, it may be helpful in identifying stricture formation. Corticosteroids have been used to try to prevent stricture formation, but their value is unclear.

SUMMARY 14 Acid poisoning • Acid ingestion may cause local corrosion, asphyxia due to oedema of the glottis and perforation of the stomach (less commonly of the oesophagus) • Metabolic acidosis and renal failure may also occur • Lavage should not be performed unless the agent produces severe systemic toxicity • Late stricture formation may occur but the value of corticosteroids in preventing this complication is not proven

Phenol (carbolic acid) can cause systemic toxicity as well as local corrosion, although the latter is normally less severe than with inorganic acids. The urine may be dark brown or black and the breath, urine and blood may smell strongly of carbolic. Haemolysis, renal failure, coma and convulsions, myocardial damage and metabolic acidosis may occur, and early cautious gastric lavage is indicated, provided the airway can be protected. Metabolic acidosis corrected with intravenous sodium bicarbonate and urine output and renal function carefully monitored. Cresols (methylphenols) have similar effects and treatment is therefore as for phenol poisoning. Formic acid is used in kettle descalers and destaining solutions. It too can produce a metabolic acidosis, intravascular haemolysis and renal failure, as well as local corrosion. Gastric lavage is therefore indicated early after ingestion, and any metabolic acidosis should be corrected with intravenous sodium bicarbonate. As with phenols, aspiration into the lung may cause haemorrhagic tracheobronchitis and the airway must therefore be protected during lavage.

ALKALIS Corrosion by alkalis (e.g. drain and oven cleaners) is generally more severe than by acids and is more likely to cause serious oesophageal, as well as gastric, damage. Stricture may develop weeks or months after the initial injury. Treatment is similar to that for acid ingestion; lavage should not be performed. Endoscopy by an experienced operator should be considered to assess the degree of injury and to prevent unnecessary hospitalization. The instrument should normally be passed only to the level of the first area of injury. It may also be useful in identifying alkalis in tablet form, which may adhere to the oesophagus (e.g. Clinitest tablets and some denture cleaners).

BLEACHES Most bleach contains sodium hypochlorite (up to 10%), which may produce corrosion in large or concentrated doses, and laryngeal oedema due to chlorine liberation. The stomach should be cautiously emptied within an hour if more than 300 mL has been ingested in an adult (100 mL

in a child), provided there is no evidence of oesophageal corrosion or severe gastritis. An H2-receptor blocking drug such as ranitidine should be given, and corticosteroids may be of value if laryngeal or pulmonary oedema is present. The sodium load resulting from ingestion of large amounts of bleach can cause hypernatraemia and hyperchloraemic acidosis.

1

DISINFECTANTS Although phenols and cresols (found in Jeyes fluid and some other disinfectants) have been largely replaced by dichlorometaxylenol, corrosive effects may occur after large or concentrated doses. Metabolic acidosis, coma, myocardial and renal damage and laryngeal oedema may also occur after severe overdose, and the stomach should therefore be emptied and the metabolic abnormalities corrected. Some disinfectants also contain isopropanol (discussed on p. 36).

POISONING WITH INSECTICIDES, HERBICIDES, FUNGICIDES AND RODENTICIDES INSECTICIDES The organochlorine group of insecticides (e.g. DDT) have now largely been replaced by compounds that are less persistent in the environment. These include the organophosphorus and carbamate insecticides.

Organophosphates The organophosphates act as inhibitors of cholinesterase in humans, causing symptoms of excessive cholinergic activity, with vomiting, abdominal pain, sweating, constricted pupils and hypersalivation. Muscle spasm, diarrhoea, convulsions and coma follow. Organophosphates can be absorbed from the stomach, skin or bronchial mucosa. Plasma cholinesterase activity, measured in several centres throughout the UK, is useful in the diagnosis of both organophosphate and carbamate insecticide poisoning. Severe toxicity is associated with a reduction in plasma cholinesterase activity to around 10% of normal. Treatment consists of atropine to antagonize the cholinergic (muscarinic) effects. The initial dose is 2 nig for adults and 0.02mg/kg for children, by intravenous injection. Large doses (up to 30 mg in 24 hours in an adult) may be necessary until there are signs of atropinization (dry skin, tachycardia, dilated pupils and dry mouth). Within the first 24 hours the use of an oxime (e.g. pralidoxime mesylate or P2S) will help to reactivate cholinesterase. It can be obtained from holding centres in the UK, details of which can be found on TOXBASE. The normal dose in moder-

37

ate or severe poisoning is 30mg/kg by intravenous injection over 5-10 minutes. Benefit should be seen within 30 minutes, but repeated doses at 4-6-hourly intervals may be required. Alternatively, it may be given by intravenous infusion at a rate of 8mg/kg/h. Intravenous diazepam (5-10 mg for an adult; 0.02mg/kg body weight for a child) is helpful in reducing twitching and may improve survival. A clear airway must be maintained and assisted ventilation may be necessary.

the patient presents within an hour of ingestion) with care because of the corrosive effects of the toxin. Charcoal (1 g/kg body weight) is left in the stomach as an adsorbent for paraquat. Oxygen therapy should be avoided if possible as it may enhance the toxicity of these agents. Finally, haemodialysis and haemoperfusion have been tried in an attempt to remove the body burden of drug, but there is no evidence that either has any effect on mortality, particularly when the more concentrated solutions of paraquat (e.g. Gramoxone) are ingested. 2

Carbamate insecticides

Dinitro compounds

Carbamate insecticides produce similar symptoms (CNS signs are less prominent) and atropine is again the treatment of choice. However, because these compounds have a reversible effect on cholinesterase, their duration of action is much shorter and pralidoxime has not shown clear evidence of benefit in poisoning with these agents. O

Dinitro compounds (e.g. dinitro-orthocresol) are generally used in the form of washes for fruit trees. They cause yellow staining of the skin but can be absorbed subcutaneously and produce toxicity by uncoupling oxidative phosphorylation. This results in pyrexia, hyperpnoea and tachycardia, with subsequent sweating, thirst, dehydration and collapse. Treatment consists of preventing severe pyrexia and replacement of electrolytes and fluid. Pentachlorophenol acts similarly to the dinitro compounds but does not produce yellow discoloration of the skin.

HERBICIDES The five major groups of herbicides are: chlorates, triazines, bipyridilium herbicides, dinitro compounds and phenoxyacetic acids.

Chlorates The chlorates are powerful oxidizing agents that can produce haemolysis, methaemoglobinaemia, haemorrhagic gastroenteritis and, in severe cases, renal failure. Treatment consists of emptying the stomach (if appropriate) and methylene blue to reverse methaemoglobinaemia.

Triazines Triazine herbicides (e.g. simazine) appear to be relatively non-toxic, and serious human poisoning has not been reported.

Bipyridilium herbicides (paraquat) The bipyridilium herbicides, particularly paraquat, are extremely toxic. Not only can they cause local corrosion, but systemic absorption leads to renal and liver damage and later progressive pulmonary fibrosis. Diquat, which is less well absorbed from the gut, can cause similar toxicity, although pulmonary fibrosis seems to be slightly less likely to occur. Treatment consists of immediate gastric lavage (if

Phenoxyacetic acids Phenoxyacetic acids are probably the most widely used herbicides. They are relatively non-toxic unless large doses have been ingested; signs of cholinergic hyperactivity are then evident, resulting in coma and occasionally ventricular arrhythmias. Other features in severe cases include hypoglycaemia, metabolic acidosis, convulsions, hypertonia, myoclonus, rhabdomyolysis and renal failure. The stomach should be emptied if appropriate, contaminated clothing removed and the skin washed with soap and water. With 2-4-D and mecoprop urinary alkalinization may enhance their rate of removal, but haemodialysis is more efficient and should be considered in severely poisoned patients.

FUNGICIDES Organic and inorganic mercurials, organotin derivatives and dithiocarbamates are all used as fungicides in industry and horticultural practice. Organotin compounds are poorly absorbed from the gastrointestinal tract and therefore have low toxicity; dithiocarbamates are also relatively non-toxic.

Organic mercurials 1

38

MCQ 1.26, 1.27

2

MCQ1.28

Inhalation or ingestion of organic mercurials can produce severe corrosion and irritation of the respiratory tract. CNS toxicity consists of tremor, memory and psychiatric disturbances, and visual impairment that may lead to blind-

ness. The chelating agents DMSA and DMPS may reduce the systemic toxicity, but dimercaprol is contraindicated.

Inorganic mercurials Inorganic mercurial fungicides (e.g. mercuric chloride) are also predominantly corrosive. Treatment includes gastric lavage, adequate analgesia and treatment of any complications. Mercurous chloride (calomel) and mercuric oxide (the most commonly used agents) are poorly absorbed and, although corrosive, systemic toxicity is unlikely. If albuminuria occurs, however, significant absorption may have taken place and plasma mercury concentration should be measured. Treatment involves the use of the chelating agents. Although dimercaprol is licensed for this use, it has to be given by deep, painful intramuscular injection, and two orally active chelating agents, DMSA (Succimer) and DMPS (Dimaval) have also been shown to be effective. Guidance on administration should be obtained from the regional poisons centre for each individual case.

Metallic mercury Ingestion of metallic mercury is of no major consequence. However, intramuscular injection of metallic mercury may cause systemic toxicity, and intravenous injection has resulted in pulmonary embolus. Inhalation of mercury vapour may produce cough, retrosternal discomfort and breathlessness due to an acute pneumonitis, and flu-like symptoms such as lethargy, fever, aching limbs and arthralgia. High-dose corticosteroids may be useful if signs of lung damage are present. Signs of systemic poisoning may be an indication for dimercaprol, DMPS or DMSA.

cerebral oedema, coma and convulsions and ventricular fibrillation have also been reported. The chelating agents dimercaprol, DMSA and DMPS may reduce the severity of poisoning. Strychnine is sometimes used as a rodenticide, and has been found as an adulterant in some recreational drugs. It is rapidly absorbed by mouth and gastric lavage is contraindicated, as a convulsion may supervene during the procedure or be precipitated by the lavage. The clinical picture is similar to tetanus, with increase in muscle tone in the limbs and facial muscles, with severe spasms, muscle rigidity and exaggerated tendon reflexes. Twitching of muscles precedes the onset of convulsions, which may typically last between 30 seconds and 2 minutes and occur repeatedly. Convulsions may be provoked by any stimulus, including touch, pain and noise. The patient may be conscious and lucid between the convulsions. Cyanosis and death may occur during the convulsions, and most patients do not tolerate more than five convulsive episodes. The increased muscle tone can lead to hyperthermia and muscle damage (rhabdomyolysis) in severe cases. Treatment involves urgently establishing and maintaining a clear airway and ensuring adequate ventilation. This may be difficult if muscle tone is increased or convulsions are occurring, and it may be necessary to paralyse the patient before an endotracheal tube can be inserted and the patient ventilated. The patient should be kept at absolute rest and any stimuli that may precipitate further convulsions should be avoided. Sensory stimuli should be minimized by nursing in a darkened room; convulsions should be treated with diazepam. In severe cases neuromuscular blockers may be needed to paralyse the patient while ventilation is supported mechanically.

RODENTICIDES Many rodenticides contain warfarin or related coumarin anticoagulants and, unless taken in huge amounts, generally cause few symptoms. If the INR does rise, vitamin K should be administered as for warfarin overdose. Alphachloralose is a CNS stimulant that can cause coma, convulsions and rhabdomyolysis in overdose. Treatment consists of managing the complications. Fluoride salts may be associated with severe toxicity, sometimes delayed in onset, including abdominal pain and diarrhoea, with hypotension and peripheral circulatory failure. Hypocalcaemia and hypomagnesaemia cause muscle spasms, weakness, convulsions and cardiac arrhythmias including ventricular fibrillation. Coma and respiratory failure may supervene. Treatment is symptomatic and supportive. Arsenic is sometimes used as a rodenticide as well as a component of some timber (tannalizing) preservatives. Abdominal pain and vomiting precede the onset of profound (sometimes bloody) diarrhoea. Hypotension,

1

POISONOUS PLANTS AND FUNGI Although the number of plants that have been suggested to be poisonous is very great, many of these contain only small amounts of toxins and generally do not cause severe poisoning. They can be classified into those: • • • • • • • • •

producing cholinergic (nicotinic) stimulation producing cholinergic (muscarinic) stimulation with anticholinergic activity with gastrointestinal activity with effects on the heart with convulsant activity with hallucinogenic effects with dermatological effects with toxicity on other organs.

Plants with cholinergic (nicotinic) activity Alkaloids similar to nicotine are found in hemlock (coniine) and laburnum (cytisine). Nausea and vomiting

39

may be followed in severe poisoning by confusion, hallucinations, convulsions, and finally coma and respiratory arrest. Treatment for severe poisoning is as for nicotine. The stomach should be emptied and activated charcoal administered. Monitoring of fluid and electrolytes and adequate replacement therapy, together with measures to assist the respiration and cardiovascular functions, are generally only necessary in severe poisoning (rarely caused by laburnum).

Plants with cholinergic (muscarinic) activity Plants with cholinergic activity include the clitocybe and inocybe fungi. Ingestion produces increased perspiration, abdominal pain and visual disturbances. Treatment is with atropine, 1.2mg intramuscularly, repeated if necessary.

and delphinium may cause bradycardia and muscular weakness.

Plants with convulsant activity Cicuta viirosa (cowbane) and Qenanthe crocata (hemlock water dropwort) contain potent alcohols with effects similar to picrotoxin in antagonizing the CNS inhibitory effects of y-amino butyric acid. In severe cases nausea, vomiting and hypersalivation are followed by convulsions and respiratory failure. Convulsions can be controlled with either short-acting barbiturates or diazepam; respiration may require support until the effects of the toxin wear off. Strychnine is found in the seeds of the tree Strychnos nux vomica, native to India. Features and management are described under rodenticides (page 39).

Plants with hallucinogenic effects Plants with anticholinergic activity Plants with anticholinergic activity include the deadly nightshade (Atropa belladonna}, the thorn apple and henbane. In addition to atropine these plants may contain hyoscine and hyoscyamine. The symptoms are similar to those caused by drugs with anticholinergic activity (e.g. tricyclic antidepressants) and management is supportive.

Plants with gastrointestinal toxicity Oxalates present in some plants can cause local gastrointestinal irritation and painful ulceration of the mouth if chewed by children. These include the household plants dumb cane (Dieffenbachia) and some of the Monstera and philodendron species. Other plants contain irritant resins and saponins (e.g. arum, yew, bryony and Daphne mezereum), which can cause severe vomiting and diarrhoea. The beans of the Abrus species (e.g. Ricinus and Robinia) can cause delayed gastrointestinal symptoms as they contain compounds that can impair RNA synthesis. Solanum species also produce delayed gastrointestinal symptoms: these include potatoes that have been allowed to sprout, and woody and black nightshades. Treatment is supportive, with replacement of fluids and electrolyte loss.

Plants causing cardiovascular disturbances Digitalis glycosides are found not only in the foxglove but also in lily of the valley (Convallarid). Treatment is as for digitalis poisoning. The aconitine present in the monksfoot

1

40

MCQ1.29

In addition to those anticholinergic agents with hallucinogenic effects, there are several other hallucinogens in plants and fungi. These include the psilocin and psilocybin in psilocybes (magic) mushrooms, the tetrahydrocannabinol in cannabis, and the ibotinic acid and muscimol found in Amanita muscaria. Tranquillizers may be necessary if the patient is a risk to himself or to others, and adequate supervision must be given until the hallucinations subside. Table 1.13 lists various plants and their toxicities.

Plants that cause dermatological toxic ity Plants contain substances that can cause either direct chemical irritation (e.g. histamine and oxalic acid) or delayed hypersensitivity reactions (e.g. Primula obconica and the hogweeds, that can produce photoallergic dermatitis and several other species). Treatment is symptomatic.

Plants with toxic ity on other organs The lectins present in the seeds of Ricinus communis cause delayed gastrointestinal tract symptoms. Haemolytic anaemia may also occur, and haemoglobin precipitation in the renal tubules may cause acute renal failure. The principal lectin, ricin, is the most potent toxin known to man. Management is largely symptomatic and supportive. The beans of Ricinus communis are also highly allergenic and may cause anaphylaxis. Severe liver damage can also be caused by the phallotoxins and amatoxins found in Amanita phalloides and some of the other Amanita fungal species. Cooking or drying does not destroy these toxins. Vomiting, abdominal pain and diarrhoea often occur 6-12 hours after ingestion, and albuminuria, haematuria, renal failure and liver failure may occur around 1-3 days later. Treatment consists of emptying the stomach and replacing fluid loss. Specific antidotes are of unproven value, but thioctic acid and high doses of penicillin may be of value in preventing the liver and renal damage.

Blue-green algae are actually cyanobacteria, which grow as blue-green or brown rafts or sediments, usually in stagnant water, and which produce endotoxins. Their disagreeable odour usually prevents human consumption, but deaths have occurred in cattle. In humans local irritation and allergic skin reactions may occur. Consumption of contaminated water by humans may result in acute gastroenteritis, which usually resolves in 1-2 days. Occasionally a hepatitic picture or atypical pneumonia has been reported. Treatment is normally symptomatic and supportive. O

TABU 1.13 Plants and toxicity Plant or fungus

Latin name

Toxic features

Black (garden) nightshade Blue-green algae

Solarium nigrum Microcystis/Anabaena/ Aphanizomenon/ Oscillatoria Bryonia dioica Ricinus communis Clitocybe sp. Cicuta vimsa Daphne mezereum

Gastrointestinal Hepatotoxic Dermatological

Atmpa belladona Amanita phal/oides Dieffenbachia Amanita muscaria Digitalis purpurea Heracleum mantegazzianum Conium maulatum Oenanthe cmcata Hyoscamus niger Inocybe sp. Laburnum anagroides Conva/laria Psilocybes sp. Aconitum nape/lus

Anticholinergic Hepatotoxic Gastrointestinal Hallucinogenic Cardiotoxic Dermatological

Nerium oleander Thevetia peruviana Solanum tuberosum Primula obconica Philodendron Monstera deliciosa Datura stramomium Arum maculatum Solanum dulcamara

Cardiotoxic Cardiotoxic Gastrointestinal Dermatological Gastrointestinal Gastrointestinal Anticholinergic Gastrointestinal Gastrointestinal

Taxus baccata

Gastrointestinal

Bryony (white/red) Castor oil plant Clitocybe fungi Cowbane/water hemlock Daphne/mezereon/spurge olive Deadly nightshade Death cap fungus Dumbeane/Leopard lily Fly agaric Foxglove Giant hogweed Hemlock Hemlock water dropwort Henbane Inocybe fungi Laburnum/Golden rain tree Lily of the valley Magic mushrooms Monkshood/aconite/ wolfsbane Oleander (pink) Oleander (yellow) Potatoes (sprouting) Primula obconica Sweetheart vine Swiss cheese plant Thorn apple Wild arum/cuckoo pint Woody nightshade/ bittersweet Yew

Gastrointestinal Gastrointestinal Cholinergic (muscarinic) Convulsant Gastrointestinal

POISONOUS ANIMALS

1

ADDER ENVENOMATION The adder, Vipera hems, is the only indigenous poisonous snake in western Europe. Only 50% of those bitten develop signs of envenomation, which may occur either immediately, with shock, vomiting and explosive diarrhoea, or several hours later. The initial collapse may be due to activation of the kinin system; the later shock is more often due to hypovolaemia, caused by increased capillary permeability and fluid loss into the swollen limb, and direct cardiotoxicity of the venom (Fig. 1.13). Management Treatment includes reassurance, as shock may be exacerbated by the fear of impending toxicity. The limb should be kept still, but a ligature should not be used unless there is likely to be a delay of more than half an hour between the patient having been bitten and transfer to hospital. If a ligature is used, it should be tight enough to prevent

Cholinergic (nicotinic) Convulsant Anticholinergic Cholinergic (muscarinic) Cholinergic (nicotinic) Cardiotoxic Hallucinogenic Cardiotoxic

FIG. 1.13 Adder bite on the finger of a child A Note the twin puncture wounds and B subsequent oedema involving the hand and arm.

41

venous return but not to obstruct arterial inflow. If there are signs of toxicity the patient should be admitted and observed for at least a day after the bite. Pulse and blood pressure should be recorded, together with urinary output and fluid losses from diarrhoea and vomiting. Local swelling should also be noted, and the white cell count, urea and electrolytes should be measured each day. The use of antivenom should be considered if hypotension persists, ECG signs occur, the white count rises markedly to above 20000/mm3 or, in an adult, severe swelling has extended up the limb within 2 hours after the bite. Anaphylactic reactions to the antivenom occur in 1 % of patients and so antivenom should only be used when the indications described above are present. A history of asthma or allergy is a relative contraindication to its use. Antivenom is given by slow intravenous infusion and is stopped at the first signs of allergic reaction. Adrenaline (epinephrine) solution must be drawn up prior to administration of the antivenom and given intramuscularly if a reaction occurs (intravenously in the case of a severe reaction). Deaths from adder bite are extremely rare but local tissue necrosis may be severe, particularly in adults.

FISH ENVENOMATION

In the UK the only venomous fish regularly found are the Weaver fish (Trachinus vipera), which is found around the coast in the summer months, and the Lion fish (an aquarium species). Both contain heat-labile venom in their spiny dorsal fins, which causes severe local pain. Hypotension, and myocardial and respiratory depression may develop. Management Treatment is to remove any barbs and clean the wound, immersing the limb in water as hot as can be tolerated without discomfort to destroy the toxin. The wound should be examined at a later stage for signs of secondary infection.

VENOMOUS INVERTEBRATES

The Portuguese man of war (Physalia physalis) is the most poisonous jellyfish found around the coasts of the UK. Its sting contains substances capable of releasing histamine and other kinins. Local pain is common, but more serious symptoms include abdominal pain, dyspnoea, hypotension, muscular paralysis and convulsions. Most other jellyfish found around the UK are harmless, but the compass jellyfish (Chrysaora hyoscella), the lion's mane (Cyanea O MCQ1.30

42

capillata) and the sea nettle (Cyanea lamarckii) can cause a sting. Management The wound should be bathed in vinegar solution (4-6% acetic acid), and any tentacles still adherent removed with forceps, a gloved hand or adhesive tape. Treatment is otherwise symptomatic.

ARTHROPOD ENVENOMATION Wasp and bee stings contain amines, kinins, peptides and enzymes that cause local pain, erythema and swelling. Systemic effects are rare unless the stings are numerous or the subject is hypersensitive, in which case even a single sting may kill. Initial symptoms are generalized urticaria, flushing, dizziness, bronchospasm, collapse and coma. Serum sickness may occur after a week or more. Management The sting of the bee should be removed by flicking or scraping, taking care not to squeeze it and inject more venom. Severe pain may respond to local anaesthetic or aspirin. Swelling may respond to antihistamines and corticosteroids. Anaphylaxis should be treated immediately with adrenaline (epinephrine) (0.5-lmL of a 0.1% solution intramuscularly for an adult and 10|ig/kg for a child). The dose may be repeated every 10 minutes if necessary. Stings in the mouth, even in patients who are not hypersensitive, may cause severe respiratory tract obstruction. In allergic individuals, specific venoms have been used in increasing doses to produce desensitization.

POISONOUS SEAFOODS

Seafood may contain bacteria (e.g. Vibrio, Salmonella) or viruses (e.g. hepatitis A); toxicity may also occur as a result of contamination by dinoflagellates or decomposition by bacteria. Antemortem infestation by dinoflagellates of certain subtropical fish (such as groupers and snappers) may occur, but contamination usually involves shellfish, such as mussels, oysters, scallops and clams. The dinoflagellates contain neurotoxins (e.g. saxitoxin) which can cause nausea, vomiting and diarrhoea, followed by circumoral paraesthesiae, ataxia, muscular weakness and, in severe cases, respiratory paralysis. Treatment is symptomatic, with respiratory and cardiovascular support. Postmortem contamination of some oily (dark-meat) fish, such as mackerel, skipjack and tuna, sardines and pilchards, by Proteus bacteria may result in the production of histamine and other unidentified toxins that produce urticaria, flushing, abdominal pain, nausea, vomiting and diarrhoea and, in some cases, bronchospasm. This 'scombroid fish' poisoning occurs around 4 hours after ingestion,

generally lasts only a few hours and is seldom severe. Treatment is symptomatic. Prevention involves keeping the raw fish frozen or on ice to prevent decomposition by bacteria.

POISONING BY METALS IRON Although mortality from iron poisoning has fallen in recent years, it remains one of the most dangerous metals in overdose. Children are particularly sensitive to its effects. Abdominal pain, nausea and vomiting are often followed by gastrointestinal bleeding, encephalopathy, metabolic acidosis, pulmonary oedema, coma, shock, acute renal and liver failure. Doses of elemental iron of more than 150mg/kg body weight produce severe and possibly fatal toxicity. A serum iron level greater than 5mg/L (90 pmol/L) within 4 hours of the overdose is associated with an increased likelihood of severe poisoning. Coma and shock are manifestations of severe toxicity.

Management The stomach should be emptied if more than 60mg/kg has been ingested within the previous hour. Radiography of the abdomen in the first 2 hours may reveal whether or not iron tablets have been taken. Blood should be taken for urgent measurement of serum iron concentration. If coma or shock develop, desferrioxamine should be considered immediately without waiting for the serum iron result. The dose is 15mg/kg body weight/hour for 5 hours. It often causes hypotension if infused more rapidly than this, and can occasionally cause rash and anaphylactic-like reactions. Pulmonary oedema and ARDS have been reported in patients treated with more than 80 mg/kg/day for longer than 24 hours. With prompt treatment the mortality should be less than 5% in severe cases. 1

HEAVY METAL POISONING Heavy metals can cause acute or chronic symptoms,

SUMMARY 15 Iron poisoning • Children are particularly sensitive to the toxicity of iron • Abdominal pain, nausea and vomiting may be followed by gastrointestinal haemorrhage, encephalopathy and renal and hepatic failure • Severe poisoning is normally associated with plasma concentrations around 4 hours after ingestion of more than 5mg/L, with CNS signs and shock • Intravenous desferrioxamine should be given as soon as possible in severe iron poisoning

depending on the dose and duration of exposure. Acute ingestion of most heavy metals will cause gastroenteritis. Generalized convulsions may also be seen, particularly with lead poisoning, and renal failure and cardiac arrhythmias have also been recorded. Inhalation of heavy metal fumes may cause chemical pneumonitis and, in severe cases, pulmonary oedema, and a syndrome known as 'metal fume fever' may occur several hours after the inhalation of some metallic oxides. Chronic effects of heavy metal exposure are similar but milder. Chronic gastrointestinal symptoms and CNS effects occur, particularly after exposure to lead, manganese and mercury. Lead, thallium, bismuth and arsenic can also produce peripheral neuropathy, and gold, lead, cadmium and mercury may cause renal damage. Inhalation of heavy metal fumes can cause pulmonary fibrosis, particularly with beryllium, cadmium, tungsten, titanium and cobalt. An emphysematous picture has been seen with cadmium, and an asthma-like syndrome may occur as a result of sensitivity to chromium, vanadium and platinum. Nickel, chromium, cobalt, platinum and beryllium may also cause skin sensitization and subsequent dermatitis. Taking a full occupational history helps diagnosis. Clinical examination may reveal specific signs of poisoning, such as the blue line on the gums in chronic mercury poisoning or the raindrop pigmentation of the skin in arsenic poisoning. Toxicological examination of the urine, blood, or hair and nail clippings may also help, but this often has to be performed in specialized centres.

1

Management Treatment consists of preventing further exposure to the heavy metal, treatment of the complications of poisoning, and chelation therapy to enhance the elimination of the metal already absorbed. It must be remembered, however, that most chelating agents are toxic and should only be used when their benefits are likely to outweigh the risks of their use. Sodium calcium edetate is the treatment of choice for lead poisoning. It must be given by slow intravenous infusion. Dimercaprol, DMSA and DMPS are effective in severe mercury and arsenic poisoning. Finally, Prussian blue has been used successfully in the treatment of thallium poisoning.

FURTHER READING British National Formulary. British Medical Association and Royal Pharmaceutical Society of Great Britain, published biannually. An excellent reference text and practical guide to prescribing, with useful tables to prescribing in special circumstances and drug interactions. Laurence D R, Bennett P N 1997 Clinical pharmacology, 8th edn. Churchill Livingstone, Edinburgh. Highly readable, with a practical therapeutic flavour. Ritter J M, Lewis L D, Mant T G K 1998 A textbook of clinical pharmacology, 3rd edn. Edward Arnold, London. A practical guide to clinical pharmacology and drug therapy.

43

This page intentionally left blank

2

Heat cramps

Physicaland

Environmental Disorders

Spasms of painful muscular contraction may occur with sudden rises in temperature during strenuous activities such as long-distance running, or during work in very hot environments. Muscles of the legs and arms are usually affected, typically after the exertion or heavy work is over. There is usually hyponatraemia, and severe cramps may be accompanied by a rise in creatine phosphokinase and evidence of rhabdomyolysis. Part of the pathogenesis appears to be due to replacement of fluid and salt loss by water with too low a salt content; the condition is preventable by drinking fluids containing 0.25% saline.

2

Heat exhaustion John Moxham and Robert L Souhami

Disorders due to heat and cold 45 Whole body radiation exposure 47 Disorders due to extremes of barometric pressure 49

Submersion and drowning 51 Electric shock 53 Smoke inhalation 54 Chemical warfare 54

DISORDERS DUE TO HEAT AND COLD Heat produced by metabolism is lost mainly through the skin and, to a lesser extent, through the gut, urine and breath. Inadequate heat loss leads to sweating and vasodilatation. Cold produces cutaneous vasoconstriction and shivering, which results in increased muscle metabolism and heat production. In moving from a cold to a hot climate, acclimatization occurs over a period of 1-2 weeks; this process is not well understood. There is increased sweating and the salt content of sweat diminishes. Aldosterone secretion increases, at least at first, causing a low urinary sodium with loss of potassium. A progressive rise in temperature above 38-39°C causes irritability and mental confusion. At temperatures above 40-41°C vascular, renal and hepatic damage occur.

HEAT EXCESS Three syndromes due to excess heat occur: heat cramps, heat exhaustion and heatstroke.

Heat exhaustion or heat collapse is the commonest syndrome and is especially likely to occur in individuals exposed to a hot environment for a few days. The very young and the elderly (especially those on diuretics) are particularly at risk. The syndrome is due in part to salt and water loss, and in part to a failure of circulatory adaptation to heat. When water loss predominates, there is intense thirst and weakness, agitation and confusion. Water depletion is more likely to occur in acclimatized people (whose sweat contains less salt). The plasma Na+ and Cl~ rise. With salt depletion, there are muscle cramps, headaches, vomiting and myalgia. Haemoconcentration occurs with a rise in haematocrit and blood urea. Usually, salt and water depletion occur together, and the patient has a normal temperature and appears pale and clammy. The disorder is treated by rest and oral fluids. In water depletion (where plasma sodium is elevated) this should be with water. Typically 5-8 L may be given in 24 hours by mouth or as i.v. glucose. When salt depletion is present the oral fluids should contain salt (e.g. 5L of fluid with 20 g salt).

Heatstroke Heatstroke is a serious and sudden disease which may be fatal unless treated urgently. It is caused by a failure of thermoregulation, resulting in a rapid rise in body temperature. In normal individuals it is usually due to exercise or work in a hot, humid environment. When it occurs in sedentary individuals the patient is usually elderly, often with underlying disease such as cardiac failure, diabetes (with autonomic dysfunction), obesity or alcoholism. Atropine-like drugs, p-blockers, phenothiazines and diuretics predispose to the syndrome. Heatstroke can also occur in fit men working in hot environments, and there are some clinical differences in these cases. Typically the patient develops headache, dizziness and faintness, and abdominal pain and delirium may occur. Sweating stops and the body temperature rises rapidly (and may reach 43°C). Hypotension, renal failure and

45

rhabdomyolysis may occur; the latter is especially common in heatstroke following severe exertion. In this form of heatstroke sweating may continue. On examination the skin is hot and usually dry, and the patient exhibits tachypnoea and flaccidity. Coma and shock are grave signs. The blood shows leukocytosis and respiratory alkalosis (but may be acidotic in the exertional form), the urea is raised, and hypokalaemia and hyperphosphataemia occur. Myocardial necrosis may give ST changes and even an infarction pattern on the ECG. Liver function tests are abnormal, and acute renal failure may occur. Disseminated intravascular coagulation is present in mild degree in the elderly, but may be severe in exertional heatstroke. Management Treatment is urgent. The patient must be cooled as quickly as possible. The most effective way is immersion in cold water, the skin being rubbed to increase the efficiency of cooling. An alternative is to blow cold air over dampened skin. Core temperature must be monitored using a rectal thermometer. Hypotension can be treated with small quantities of saline, taking care to avoid circulatory overload. Hypoglycaemia, if present, can be treated with glucose, and severe acidosis by sodium bicarbonate. More serious complications, such as acute renal failure and disseminated intravascular coagulation, are treated in the usual way (Chs 20 and 23). Cerebellar degeneration may occur, and coma with cortical damage is a terminal event. 1

SUMMARY 1 Clinical features of heatstroke Symptoms • Headache • Dizziness/faintness • Abdominal pain Signs • Increased body temperature • Hot, dry skin (sweating may be absent) • Tachypnoea • Flaccidity • Hypotension • Delirium -> shock/coma Results/complications • Hepatic and renal failure • Leukocytosis, alkalosis/acidosis, hypokalaemia • Rhabdomyolysis • Myocardial necrosis • Disseminated intravascular coagulation • Cerebellar degeneration • Coma, cerebral damage

as 'ecstasy'. Muscle contraction begins after the start of anaesthesia, and body temperature rises rapidly. The anaesthetic must be stopped, the patient cooled and ventilated, and i.v. dantrolene sodium given (2-10mg/kg). Acidosis should be treated. 0

HYPOTHERMIA

Neuroleptic malignant syndrome This syndrome is characterized by muscle rigidity giving rise to hyperpyrexia, tachycardia and impaired consciousness. It is an idiosyncratic reaction to drugs that block dopaminergic transmission, such as phenothiazines and butyrophenones. The increase in muscle tone develops over 12-72 hours. Death from respiratory failure may occur. Treatment is by muscle relaxants, including neuromuscular blocking agents, and ventilation if necessary. Recovery may take several days.

Significant hypothermia is an uncommon problem mainly affecting the elderly, in whom the response to cold may be defective (p. 178). Hypothermia contributes to the excess mortality in this population during winter months. It is defined as being present when the deep body temperature falls below 35°C. As the mouth temperature may fluctuate depending on the ambient air temperature, it is best to use a rectal thermometer when monitoring hypothermic patients. A low-reading thermometer is necessary. The rectal temperature is about 0.5°C higher than oral.

Malignant hyperpyrexia

Aetiology

This is an autosomal dominant disorder in which hyperpyrexia is produced by skeletal muscle contraction following anaesthetic agents such as halothane and suxemethonium and local anaesthetics. It can also occur with drugs such as amphetamines and derivatives such

Hypothermia may be due simply to an age-related impaired thermoregulatory response to a cold environment (Ch. 7, p. 178). In the majority of patients, however, there is an associated underlying pathological cause, including any cause of autonomic dysfunction (p. 178). Hypothyroidism and hypopituitarism are both associated with hypothermia, and the risk is also higher in immobile or demented patients, or in the presence of an acute illness such as bronchopneumonia. There is a clear association between phenothiazines and impairment of temperature regulation, and hypothermia may also be precipitated by sedative drugs, such as hypnotics and alcohol. Although

1

46

Case 2.1

2

MCQ2.1

3

MCQ2.2

a mild drop in body temperature is common, clinically important hypothermia is unusual. However even modest falls in body temperature may produce changes in coagulability, which may predispose to myocardial infarction. Immersion hypothermia is produced by shipwreck. Exposure to cold weather occurs in hillwalkers and climbers, especially if they have not eaten or if they drink alcohol.

Clinical features Hypothermic patients typically live alone in poor housing and have impaired mobility. A typical case would be that of an old person who gets out of bed in the middle of the night to go to the lavatory, falls, and then lies all night on the cold floor in thin nightclothes until discovered the next morning by a neighbour. The skin has a greyish look owing to a combination of vasoconstriction and cyanosis; the face may appear puffy, resembling myxoedema. The diagnosis is often first suspected during examination, when it is found that the patient's skin feels cold in normally warm areas, such as the axilla or the abdomen. There may be ataxia and slow mental response. The patient is often confused when the temperature falls below 35°C, and drowsiness and loss of consciousness occur when the temperature falls below 29-33°C. Shivering is absent below this temperature; instead, there is increased skeletal muscle tone with neck stiffness and abdominal rigidity. The heart rate is slowed and the ECG may show a prolonged PR interval and J waves (Fig. 2.1). Respirations are slow and shallow, and there may be hypoxia and hypercapnia. Hypoxia is exacerbated by a shift of the oxygen dissociation curve to the right, leading to impaired release of O2 from haemoglobin. Acute pancreatitis is often found at necropsy, but this is only rarely diagnosed in life. Impaired renal and respiratory function may lead to a raised blood urea or bicarbonate. Blood sugar levels may be raised, but do not indicate diabetes mellitus, and will fall as the temperature returns to normal. Serum asparate aminotransferase and creatine kinase may be raised due to muscle damage.

Management The lower the deep body temperature, the higher the mor-

SUMMARY 2 Clinical features of hypothermia • • • • • • • •

Contusion/impaired consciousness Cold, greyish skin Shivering absent Increased skeletal muscle tone Slow heart rate, prolonged PR interval and J waves Slow, shallow respiration, hypoxia and hypercapnia Raised blood urea or sugar levels Raised serum asparate aminotransferase and creatine kinase

2

tality (about 50% overall). This is because many hypothermic patients have a serious underlying illness. Patients with mild hypothermia (core temperature 32-35°C) should be wrapped in a space blanket and nursed in a side ward where the ambient temperature is 28-30°C. Their temperature should be allowed to rise gradually, at a rate of about 0.5°C per hour, as fast surface rewarming carries the risk of hypotension. For this reason, the patient's pulse and blood pressure must be monitored. If the blood pressure drops, the space blanket should be removed or the room temperature lowered so that the patient is temporarily cooled. Intravenous fluids are hazardous and may produce acute pulmonary oedema. A broad-spectrum antibiotic is usually given parenterally because bronchopneumonia is nearly always present, even if not clinically apparent. The ECG should be monitored, as both bradyarrhythmias and ventricular fibrillation may occur. The patient is at serious risk of developing a pressure sore and must be nursed on a suitable mattress and turned regularly. When the deep body temperature is below 32°C a more aggressive approach to resuscitation is sometimes advocated, although there is little evidence that this reduces mortality. These measures include positivepressure ventilation to correct hypoxia, measurement of central venous pressure, and active rewarming using a radiant heat cradle over the torso or warmed intravenous fluids.

Thyroid hormones should not normally be given to hypothermic patients. If there is good clinical or laboratory evidence of hypothyroidism, then triiodothyronine (5ug) may be given by slow intravenous injection every 12 hours. A particular problem is presented by patients who have become hypothermic as a result of immersion at sea. They may develop ventricular fibrillation on rescue if they exert themselves. They should therefore be kept still and rewarmed slowly. 3

WHOLE BODY RADIATION EXPOSURE

FIG. 2.1 Electrocardiogram showing J wave (arrowed) in a hypothermic patient

There are two broad types of ionizing irradiation: photons (y and X-rays) and particulate (a and |3 particles). Radiation dose is defined in a variety of ways. The energy

47

deposited in a tissue is measured in grays (Gy; IGy = 1 J/kg). Because different types of radiation have different tissue effects, a unit called the Sievert (Sv) is sometimes used, which weights the radiation dose according to the type of radiation. Particulate radiations are absorbed according to their mass and energy, but do not penetrate deeply into tissues (see also Ch. 6, p. 159). Particulate irradiation transfers more energy and is much more ionizing than Y or X-rays.

EXPOSURE Exposure to ionizing irradiation comes about in three ways: • Background irradiation. We are exposed to approximately 2-3 mSv per year from solar and geological irradiation. Cosmic rays are mostly very high-energy protons. Secondary radiations, from the upper atmosphere, are largely y-rays. Geological irradiation is largely due to radon. • Medical exposure. The therapeutic uses of ionizing irradiation are described in Chapter 6. Accidental exposure of patients or staff has been rare since the advent of stringent precautions following the radiation-induced damage to skin and eyes in the early days of therapeutic and diagnostic radiation. Exposure to whole-body radiation forms part of the treatment of leukaemia and lymphoma. The resulting bone marrow suppression is treated by allogeneic (or autologous) haemopoietic stem cell transplantation. • Military and industrial irradiation. The atomic bombs dropped on Hiroshima and Nagasaki led to numerous cases of acute leukaemia and other cancers. Fallout from nuclear tests has increased annual background radiation by 1%. Industrial exposures also occur, the most dramatic example of which has been the ingestion of radium by radium dial painters from 1916 to 1926, resulting in a great increase in risk of bone sarcoma and cancer of the air sinuses of the skull. The accident in the Chernobyl nuclear power station in 1986 resulted in many deaths due to acute radiation exposure, and an increase in background radiation over a wide geographical area.

TISSUE DAMAGE The tissue-damaging effect of ionizing irradiation is probably related mainly to damage to DNA, leading to strand

1 48

MCQ2.3

breakage and impaired reproductive integrity of the cell. At 0.1 Sv total body exposure there is a slight fall in blood lymphocyte count. At ISv there may be mild radiation sickness with slight nausea and a more profound lymphopenia, and a fall in neutrophils and platelets 2-3 weeks after exposure. Following greater degrees of whole-body exposure certain tissues are damaged acutely by relatively low doses: • Bone marrow. This can regenerate after exposure to 10 Gy, but above this dose permanent aplasia may occur. The white count and platelet count begin to fall within 10 days of exposure. • Intestine. Doses of 10 Gy or over cause severe loss of crypt cells, leading to loss of villi and extensive ulceration. The onset of nausea and vomiting is 1-2 hours after exposure, often recovering a few hours later, but above 10 Gy permanent and fatal damage to the gut occurs, with death in 4-14 days. Death is due to severe diarrhoea, consequent salt and water depletion, and severe Gram-negative sepsis. • Skin. Erythema occurs at doses below 10 Gy. At 20 Gy the skin starts to desquamate and ulcerate. • Lung. Above l0 Gy in a single fraction pneumonitis occurs, and is increasingly severe with increase in dose. Long-term changes include glomerular loss and interstitial nephritis, infertility, pulmonary fibrosis, neuronal loss and gliosis, intestinal stricture and hepatic fibrosis. After acute radiation exposure (such as at Chernobyl and Hiroshima) there is an increased risk of leukaemia at 5-10 years. The safe lower limit is not easily quantifiable but is probably below l0c Gy. Solid tumours may also occur, including bone cancer in radium dial painters, skin cancer, and lung and breast cancer. The risk of solid tumours appears to be 1 in 2000 following total body exposure of 1-2 cGy.

Clinical features After acute whole-body exposure, the clinical features depend on the received dose. Following an initial period of nausea and vomiting (sometimes with parotid swelling), the patient may be relatively well for 7-10 days and then develop a syndrome of haematological failure. There is depression of the white count and platelets, with infection, bruising and bleeding. At the same time skin reactions occur, with desquamation, alopecia and ulceration. If the whole-body dose is above 15 Gy the gastrointestinal syndrome may predominate and be the first sign of toxicity. In a patient with severe leukopenia severe diarrhoea and fluid loss are followed by septicaemia from gut bacteria. This is almost always fatal. Higher doses of whole-body radiation (25-60 Gy) cause cardiovascular collapse and shock.

Management Management is supportive. Antiemetics are given in the acute phase, and fluid losses from diarrhoea are replaced. Transfusion of blood and platelets and intensive antibiotic treatment for gut derived organisms are used during the phase of myelosuppression and infection. Haemopoietic growth factors such as GCSF may be of value. Allogeneic bone-marrow transplantation may help the minority of patients for whom there is a donor, and in whom the haematological toxicity is the major life-threatening complication. It is of no value if the major toxicity is in organs such as the lungs or gastrointestinal tract. Accidental exposure to radionuclides is treated according to the isotope. Potassium iodide is given for I31 I ingestion, and soluble phosphate for 32P overdose. Overdose with bone-localizing isotopes (caesium, radium, strontium) is treated with EDTA and large doses of oral calcium. There has been some interest in drugs that limit damage from radiation: unfortunately, these have to be given before or immediately after exposure. The best known is ethyol, which may also protect normal tissues from cytotoxic drug damage. Late effects occur 3-10 years after exposure. There is an increased risk of acute myeloid leukaemia. The risk of solid tumours is about 1 in 2000. Tumours include cancer of skin, breast and lung. This is a small increase compared to the background incidence in the population.

height. Commercial aircraft are maintained at a pressure equivalent to 2500 metres, so that no additional oxygen is needed unless a patient has severe cardiac or respiratory disease. If there is a loss of pressurization, extra oxygen is made available through masks. Acclimatization to altitude is a complex process. On rapid ascent the relative hypoxia stimulates ventilation, but this stimulus diminishes over a few weeks. This hyperventilation diminishes Paco2, with the consequent development of a respiratory alkalosis. The respiratory centre appears to adapt to a lower Paco2. The acute circulatory adaptation consists of a rise in cardiac output, a fall in renal, cardiac and cutaneous blood flow, and a rise in cerebral blood flow. These changes slowly reverse, and after several weeks cardiac output on exercise is lower than at sea level. The red cell mass increases greatly and with it the total oxygen-carrying capacity of the blood. The oxygen dissociation curve does not shift substantially (as the alkalosis shifts it to the right and the hypoxia to the left). There is an increased formation of tissue capillaries, greatly increasing tissue oxygenation. There are four recognized syndromes of altitude sickness: • • • •

2

Acute altitude sickness Acute pulmonary oedema Acute cerebral oedema Chronic altitude sickness.

Prognosis With adequate treatment, most patients will survive acute exposures up to 10 Gy. Above this dose, bone-marrow failure and the intestinal syndrome are usually fatal. 1

DISORDERS DUE TO EXTREMES OF BAROMETRIC PRESSURE

ALTITUDE SICKNESS The effects of reduced barometric pressure result from the low partial pressure of oxygen and reduced oxygencarrying capacity of the blood. Figure 2.2 shows this relationship. The barometric pressure at sea level is 100 kPa, and the partial pressure of oxygen is 21% of this, i.e. 21 kPa. At 5500 metres the barometric pressure is reduced to 50% (50kPa). The partial pressure of oxygen is, however, less than 20% of this (l0kPa) because the proportion of water vapour rises. Because of the sigmoid shape of the oxygen dissociation curve, oxygen saturation is reasonably well maintained until an altitude of 5000 metres. A precipitous fall in saturation then occurs with increasing

FIG. 2.2 Effect of altitude on alveolar Pao2 and oxygen saturation Because of the steep slope of the oxygen dissociation curve, increasing altitude above 5000 metres causes a precipitous fall in saturation. In an unacclimatized person, acute changes occur at: IT1 3000 metres (10000 feet) - slightly impaired memory and judgement, increased heart rate, abdominal cramps and nausea; [B] 3500 metres (12000 feet) - headache, nausea, diminished visual acuity, and possible pulmonary oedema; [C] 5500 metres (18000 feet) - impaired consciousness after several hours in many people; and [D] 6750 metres (22000 feet) - loss of consciousness.

49

Acute altitude (mountain) sickness This syndrome occurs in unacclimatized individuals who ascend rapidly to heights in excess of 2500 metres. People who are unfit, or very young, or those who have respiratory or cardiac disease, are especially prone. The symptoms usually begin at 12-24 hours after ascent. Lightheadedness is followed by lethargy irritability and severe headache, usually occipital in position. Other symptoms include nausea and vomiting, dyspnoea with periodic respiration, and abdominal pain. Physical examination reveals little, although ataxia (shown by walking heel to toe) may be present. The cause of acute altitude sickness is unknown. It is possible that the hypoxia causes mild cerebral and brainstem oedema. This is suggested on MRI scans and by multiple petechiae in the brain in fatal cases. The syndrome is preventable by slow ascent. Individual susceptibility varies greatly and mountainous hikes and climbs should never put pressure on participants to ascend rapidly, especially if they start to feel unwell. When starting an ascent above 3000 metres a reasonable pace is 300m/day, resting for a day every 3 days. Even this may be too fast for some people. Dehydration should be avoided, and a good diet maintained. Treatment is to avoid sedatives, to take mild analgesics for headache, and if the symptoms do not abate, to descend to 1000m to await recovery. Acetazolamide can alleviate the symptoms of acute altitude sickness by acting as a respiratory stimulant, lowering the Paco2 and raising the Pao2. If ascent cannot be made slowly it may be advisable to start acetazolamide (250 mg b.d.) before ascent, and climbers should be warned of the side-effects, particularly paraesthesiae. Acute pulmonary oedema Rapid ascent to heights greater than 3000 metres may provoke the sudden onset of acute pulmonary oedema. Hypoxia is the cause, but the pathogenesis is unclear. The symptoms usually begin 5-36 hours after ascent; acute dyspnoea, haemoptysis and widespread crackles in the lungs occur. Pathologically there is acute oedema and a 'hyaline' membrane in bronchioles. Treatment is with urgent evaluation to a lower altitude, oxygen and diuretics. Nifedipine has been shown to be useful prophylactically in those who have had a previous episode of pulmonary oedema. Acute cerebral and retinal oedema Acute hypoxia increases cerebral blood flow and intracranial pressure, and can cause cerebral oedema. This is asso-

50

1

Fig. 2.1

4

Fig. 2.2

2 Case 2.2 3 MCQ 2.4

ciated with headache and brain dysfunction. The patient may be incoherent, drowsy, ataxic, and progress to coma. There may be papilloedema. Deepening coma leads to death. The treatment is urgent descent to a lower altitude and oxygen. Dexamethasone (i.v.) should be given. Cerebral and pulmonary oedema often accompany each other. Many people ascending to high altitude develop asymptomatic retinal haemorrhage.1 Occasionally this may leave a permanent blind spot. Oedema of the retina is assumed to be the cause. Retinal oedema occurs in the absence of cerebral oedema. Pink optic discs and retinal haemorrhage are not in themselves a reason for urgent descent.

Chronic altitude (mountain) sickness Individuals (usually middle-aged men), who have lived for many years at altitudes above 4000 metres may develop polycythaemia and a chronic insensitivity of the respiratory centre to hypoxia. Such individuals hypoventilate and develop a raised Paco2. There is central cyanosis, with cough, headache and giddiness. Patients are cyanosed, with vascular conjunctivae and finger clubbing. There may be RV strain on ECG as a result of pulmonary hypertension. The CNS symptoms are similar to those of polycythaemia rubra vera and are due to decreased cerebral blood flow (a consequence of the raised haematocrit) and hypoxia. Not all people living at high altitude develop this syndrome, which appears in part to be an excessive response to decreased inspired Pao2. The syndrome can only be treated by permanent descent to a lower altitude.

INCREASE IN BAROMETRIC PRESSURE

An increase in barometric pressure is encountered in diving or in mining.

Breath hold diving During a breathhold dive the body directly encounters the increased pressure (1 atm for every 10 metres descent). The breathhold causes a rise in Paco2 and a fall in Pao2. The main respiratory stimulus is the raised Paco2, so that hyperventilation before diving leads to a lack of stimulus to breathe (owing to CO2 depletion). This may cause fatal loss of consciousness from hypoxia (which does not cause an irresistible urge to breathe and thus terminate the dive). As the diver surfaces, decompression lowers Pao2 and again may cause loss of consciousness.

Lung rupture, air embolism, pneumomediastinum During a breathhold dive the air in the lung is compressed and re-expands to its previous volume on ascent, with no risk of lung rupture. If a diver breathes air below the surface he must allow time for the excess air under pressure to be exhaled or the lung may rupture. Divers are taught not to ascend more rapidly than the gas they exhale. A diver breathing air may hold his/her breath while ascending rapidly, without breathing out. This often happens if the diver has accidentally lost the air supply. During the rapid ascent the lung may rupture. Lung rupture may be into the mediastinum or into the pleural space, causing pneumomediastinum and pneumothorax, respectively. The symptoms are breathlessness, cough and haemoptysis during the ascent. Air embolism may occur and cause dysphasia, paralysis, visual disturbance or convulsions. Mediastinal air may cause dysphagia and surgical emphysema in the neck. Treatment is with 100% oxygen to eliminate inert gas, and recompression.

Deep diving, decompression sickness In deep diving the gas must be delivered at the same pressure as the surrounding water. If the gas is air, the high pressure of N2 causes mental confusion below 50 metres. Helium/oxygen mixtures are better tolerated and descents as low as 700 metres can be undertaken. These mixtures must be used below 50 metres. At greater depths than 700 metres neurological disturbances occur owing to the high gas tensions. Very high oxygen concentrations in the inspired gas (even though they reduce inert gas narcosis and the risk of decompression sickness) are toxic to the lung and CNS. This toxicity is related to the level of inspired Pao2 and the duration of the dive, and must be carefully regulated. When breathing air, the nitrogen dissolved in blood (PaN2) increases and, on rapid decompression, forms bubbles, which come out of solution in small blood vessels. Ischaemia results, causing pain in joints, spinal cord injury, chest pain, cyanosis, confusion, visual defects, paralysis and fits. Minor confusion occurs with lesser degrees of compression injury. Treatment is by recompression, followed by very gradual decompression over days. Some instances of air embolization during diving occur in individuals who shunt blood from the right atrium to the left via a patent foramen ovale. About 25 % of the population have minor degrees of patency of the foramen ovale. Figure 2.3 shows the relationship between depth, duration of dive and the need for controlled ascent to the surface. QQ

REFERENCE Bennett PB. Elliott DH (eds) 1993 The physiology and medicine of diving. WB Saunders, London.

2

FIG. 2.3 Relationship between depth, duration of dive and need for controlled ascent to the surface

SUBMERSION AND DROWNING Drowning is a relatively common cause of accidental death, particularly in children, in whom only motor accidents and cancer are more common causes of death. In adolescents and adults, alcohol and drugs are frequently a contributory factor. Spinal injury is a relatively common problem, particularly after submersion following diving. There are considerable differences in the osmotic pressure of salt water (5% NaCl), fresh water (0% NaCl) and plasma (0.9% NaCl). It has been suggested that when large amounts of salt water are aspirated into the lung, water may be sucked into the alveoli from pulmonary capillary plasma, causing death from pulmonary oedema.4 When fresh water is aspirated, however, large volumes may pass into the pulmonary circulation, causing haemolysis and consequent potassium release, which may lead to cardiac arrest. In fact, in both cases death can result from acute reflex laryngospasm, causing asphyxia. The differences between submersion in fresh and salt water are probably trivial; the major factors determining survival are the duration of submersion and the severity of the consequent hypoxia.

RESUSCITATION FROM NEAR-DROWNING A patient who is nearly drowned needs urgent resuscitation. The basic techniques are the same as for any 'sudden death', but the physiology of drowning raises special issues. In both fresh- and sea-water drowning gastric aspiration must be performed once a cuffed endotracheal tube is in place. Large volumes of water tend to be swallowed during submersion, and vomiting during resuscitation is common.

51

Similarly, once in hospital, ventilation with positive endexpiratory pressure is required in an attempt to reinflate atelectatic and waterlogged lungs. Acute respiratory distress syndrome (ARDS) is a complication. Hypothermia is frequently an associated problem and septicaemia is a common complication. Defibrillation for VF or VT is less effective in severe hypothermia. Patients who are breathing and coughing have a good prognosis (Table 2.1). Sea-water drowning can be associated with cardiac arrest produced by the rapid absorption of Ca2+ and Mg+ into the blood. Treatment by basic life support is frequently effective in sea-water drowning, if begun within 5 minutes. Fresh-water drowning may produce massive intravascular haemolysis, associated with ventricular fibrillation and

TULEJh Mortality following submersion Mortality (%)

Grade

Definition

1

Normal lung auscultation with coughing Abnormal auscultation, with crackles in some regions Auscultation: acute pulmonary oedema, no hypotension Acute pulmonary oedema, plus hypotension Isolated respiratory arrest Cardiopulmonary arrest

2 3 4 5 6

0 0.6 5.2 19.4 44

93

Modified from Szpilrnan

CASE STUDY 2.1 SUBMERSION AND DROWNING A 26-year-old woman went for a swim with her friend at the local pool. After she had been in the water for some time swimmers at the side of the pool noticed her thrashing around beneath the surface. The onlookers and pool staff pulled her out of the water at the shallow end. She was unconscious, not breathing, and had no detectable pulse. One of the onlookers was medically qualified. Mouth-to-mouth resuscitation and chest compression was initiated and the emergency services were called. On arrival at the scene the ambulance crew documented ventricular fibrillation on the ECG, shocked the patient and restored sinus rhythm. The patient was transferred to the local accident and emergency department. On arrival in A&E the patient was breathing and had a satisfactory blood pressure. There were no signs of external trauma. The axillary temperature was 36.0°C. There was vomitus in the mouth and the patient had obstructed breathing. The Glasgow Coma Scale was 3. The patient became greatly

O MCQ 2.5

52

agitated. Arterial blood gases, breathing supplementary high-flow oxygen, showed ^302 27.3 kPa, Faco2 9.2kPa, bicarbonate 16, pH 7.07. The patient was sedated, intubated and ventilated. She was given intravenous antibiotics and steroids. Additional history obtained from her friend revealed that the patient had been depressed for some time but was not, as far as the friend knew, taking any medication. There was no past history of attempted suicide or epilepsy. Questions 1. Wharf is the prognosA for this patient? 2. On afrival it A&E, fifhat ire the most important management issues? 3. A chest X~fay was performed; what might it show? 4. What is ydur interptetatidn of the arterial blodd gas results? For the next few days the patient was managed in the intensive care unit.

The acid-base status returned to normal. A CT brain scan showed no abnormalities. The chest X-ray initially showed bilaterial patchy air space ©pacification, but subsequently cleared (CASE FIG. 2.1.1). Gas exchange gradually improved and the patient was extubated. Moderate hypoxia persisted; the patient was treated with CPAP and transferred to the high-dependency unit. The patient became febrile, was diagnosed as having a secondary chest infection and treated with antibiotics. One week after admission the patient's jPao2 was within normal limits. The patient was alert, mobilized, and discharged 9 days after admission. She had no recollection of the events that led up to her submersion in the swimming pool. Question 5. What would you regard as being the main factor or ,„, factors that ultimately y resulted in a good outcorne for this patient?

CASE STUDY 2.1

2

CONTINUED

CASE FIG. 2.1.1 Chest X-ray. There is bilateral air space shadowing throughout the lungs, affecting the left lung more that the right Discussion Submersion and drowning is a common cause of death in young people. The most important factor determining outcome is the severity and duration of hypoxia. Clearly, cerebral and spinal trauma may also be critical in patients who, for example, dive into shallow water. Severe hypoxia causes cardiac arrhythmias, cardiopulmonary arrest, brain damage and acute lung injury. Ventricular fibrillation and

cardiopulmonary arrest indicate a poor prognosis (Table 2.1). The most important management issues are to sustain and improve tissue oxygen delivery. This patient's ventilation was inadequate and the upper airway was compromised, hence the importance of intubation and mechanical ventilation. Hypothermia is often a severe problem, but not in this particular case. The patient was treated with steroids, ostensibly to reduce lung

hyperkalaemia, which can be difficult to treat. Survival after prolonged submersion during warm weather is unusual, but survival (with normal cerebral function) has been recorded in children, even after 30 minutes' immersion, particularly in low-temperature water. Sudden death due to a combination of vagal slowing of the heart and intense peripheral vasoconstriction may occur in previously fit individuals who dive into cold water. O

FURTHER READING Szpilman D. Near drowning and drowing classification: a proposal to stratify mortality based on the analysis of 1821 cases. Chest 1997:112:660-665.

injury, but there is little evidence to support such therapy. The chest X-ray showed bilateral airspace shadowing consistent with pulmonary oedema (figure). Hypoxia causes increased pulmonary capillary permeability, alveolar flooding, and can lead to established ARDS. Many patients rescued from submersion vomit and some aspirate, and atelectasis may occur. Acute laryngospasm stops large volumes of water filling the lungs and the hypoxic lung injury is the more important lung insult. The arterial blood gas results show a severe combined respiratory and metabolic acidosis (i.e. the Paco2 is raised and the bicarbonate is reduced). This reflects reduced ventilation and inadequate oxygen delivery to all tissues. Hence the crucial importance of restoring the circulation and ventilation. With appropriate therapy the acid-base abnormalities were rapidly corrected. This patient survived and avoided hypoxic brain damage because of the prompt and effective poolside resuscitation (mouth-to-mouth resuscitation, chest compression and cardioversion). Without such rapid intervention the prognosis would have been very poor.

ELECTRIC SHOCK Injury following electric shock depends on the voltage and current. Below 100 V sudden death is less likely. Alternating current appears to be much more dangerous than direct, as it produces muscle spasm and is more likely to stop the heart. The direction of flow is also important, as flow from hand to hand or hand to foot traverses the heart. At high voltage, the body heats up as the current flows. This heating is much more severe in a small cross-sectional area (fingers) than in a large one (trunk), and burns and charring are thus found mostly in the hands. The upper limbs and trunk are less affected, and deep tissues more damaged than the superficial. The skin burns do not correlate with the level of internal injury. Burns occur at the skin, and arcing of the current may ignite clothing, causing further burns.

53

Clinical features With high voltage, cardiac arrest occurs. Cardiac resuscitation is essential, as is a careful assessment for burns which may manifest themselves later. Arrhythmias may recur for up to 2 days after resuscitation. Intestinal perforation and liver and gallbladder injury may occur. There may be cerebral and spinal cord damage, and peripheral nerves may be permanently damaged. Spinal cord injury may have a delayed onset, with paraplegia or a tabes dorsalis-like syndrome. Acute muscle necrosis may result in renal failure.

Treatment After resuscitation the cardiac rhythm must be monitored for 2 days after the last rhythm disturbance. Hyperkalaemia may occur as a result of acute tissue necrosis, and may need immediate treatment with glucose and insulin or haemodialysis. Renal failure due to rhabdomyolysis must be treated. Intravenous fluids should be used to maintain urine flow, which will minimize the danger of renal failure due to massive myoglobinuria. Mannitol or a diuretic may be needed to maintain urine output. A surgical assessment must be made of burns and charring, and peripheral circulation. Debridement, fasciotomy and even amputation may be necessary. An extreme example of high-voltage injury is a lightning strike. Cardiopulmonary resuscitation is essential, as described above. Rhabdomyolysis and burns are treated in the same way as in other electric-shock injury. With effective resuscitation, over half the victims will survive.

SMOKE INHALATION

54

Smoke inhalation causes anoxia, frequently exacerbated by carbon monoxide poisoning and, if plastics are burned, hydrogen cyanide poisoning. Lung and airway injury is mainly chemical rather than thermal. Initial effects are mainly on the airways, including laryngeal oedema and bronchoconstriction; severe pulmonary oedema can develop. Factors predictive of severe injury include fire in an enclosed space, black sputum, burns around the mouth, reduced consciousness, altered voice, and symptoms and signs of respiratory distress. In fires involving plastics, carboxyhaemoglobin levels generally reflect hydrogen cyanide levels. Carboxyhaemoglobin levels should be measured in all cases and levels greater than 10% indicate significant exposure; above 25% indicates severe exposure. High CO levels cause hypoxia; the true arterial oxygen saturation is not demonstrated by pulse oxymetry, and the %CO must be subtracted from the oxymeter reading to give the correct value. Treatment is supportive, including humidified oxygen, bronchodilators and mechanical ventilation in severe cases. Steroids are useful when airways obstruction is severe and persistent. Severe hydrogen

cyanide poisoning requires treatment with sodium nitrite and thiosulphate. In patients surviving smoke inhalation injury the long-term prognosis for pulmonary function is good.

PASSIVE SMOKING Breathing other people's cigarette smoke is a much less serious cause of morbidity and premature death than active smoking, but it does represent one of the most serious causes of indoor air pollution that can affect health. Respiratory symptoms such as wheezing, coughing and the development of respiratory infections are increased in the children of smoking parents. The effect is particularly noticeable in children aged under 1 year. Various international expert committees have concluded that exposure to environmental tobacco smoke increases the incidence of lung cancer in non-smokers, the observed risk being some 30% higher than expected and contributing approximately 300 cases of lung cancer annually in the UK. There is also evidence that environmental tobacco smoke can exacerbate asthma in adults.

CHEMICAL WARFARE Chemical weapons were first used on a large scale during the 1914-18 war, and although a wide range of chemicals were subsequently developed they were not used extensively until the Iran-Iraq war. Their use, but not their possession, is banned by an international treaty. The agents most readily available are nerve gases and mustard gas (see below). Other chemical agents include lewisite, hydrogen cyanide, and riot control agents such as CS gas, phosgene and chlorine. Biological weapons such as anthrax and Botulinus toxin are also potential weapons. When the use of these weapons is threatened, protection of those involved in, or close to, the conflict with respirators and protective clothing is important. Obviously, civilian populations are very vulnerable. Rescue of casualties by medical teams may expose these personnel to risk, so they should also take preventative measures. For all casualties of poisoning by chemical and similar weapons, decontamination is essential, particularly before they are transferred to areas where unprotected staff may be working. Mustard gas (sulphur mustard) is an alkylating agent (as is nitrogen mustard). It is a 'vesicant', causing skin blistering, sloughing of respiratory epithelium, ocular damage and, in a small number of cases, subsequent bone marrow depression. There is no antidote and the treatment is supportive, including careful management of the skin damage, which often takes many months to treat. Overall mortality of mustard gas exposure is only about 2%.

Nerve agents are organophosphorus compounds, closely related to insecticides. They inhibit the action of acetylcholinesterase, leading to parasympathetic overactivity and ultimately neuromuscular blockade. Clinical manifestations are meiosis, bronchoconstriction, hypersecretion and hypersalivation, rapidly progressing to general paralysis (including the respiratory muscles), convulsions and death. These events may occur in minutes, so rapid treatment is essential; this includes the use of atropine and oximes (which reactivate acetylcholinesterase). If there is a risk of nerve gases being used, pretreatment with pyridostigmine, which binds reversibly to acetylcholinesterase and thus 'protects' it from the effects of the nerve gas, is useful. Many armies now equip personnel with individual antidote packs for use in the field.

FURTHER READING

2

Howard J, Tyrer F H 1987 Textbook of occupational medicine. Churchill Livingstone, Edinburgh. A comprehensive and very readable textbook describing environmental and work hazards. Raffle P A B, Lee W R, McCallum R I, Murray R 1987 Hunter's diseases of occupations. Hodder and Stoughton, London. A traditional and comprehensive text covering the entire field of occupational medicine. Skinner D, Driscoll P Earlam R (ed) 1991 ABC of major trauma. BMV Books, London. A concise, practical account of important topics in major trauma. Ward M, Milledge J S, West J B 1994 High altitude medicine and physiology, 2nd edn. Chapman and Hall, London. An excellent book covering all medical aspects of high altitude medicine.

55

This page intentionally left blank

3

link to form the backbone of each strand, and the bases on each strand face each other and pair by hydrogen bonding. The base pairing is specific: adenine always pairs with thymine and guanine with cytosine. The specific base pairing allows DNA molecules to be copied precisely when cells divide. A sequence of bases coding for a gene is usually divided into coding regions (exons) and non-coding regions (introns). DNA is transcribed by synthesizing a singlestrand ribonucleic acid (RNA) molecule, using one of the DNA strands as a template. The resultant messenger RNA (m-RNA) molecule is processed to splice out the noncoding regions and to modify each end, and then migrates into the cytoplasm, where it directs the synthesis of a specific protein by a process known as translation. This process involves the interaction of the m-RNA with both ribosomes in the cytoplasm and a series of molecules known as transfer RNA (t-RNA) (Fig. 3.2). t-RNA is a specialized molecule to which can be attached a specific amino acid. Each t-RNA carries a three-base coding region which pairs specifically with a corresponding sequence of three bases on the m-RNA. The sequence of three bases is known as a codon and is specific for each amino acid. The sequence of codons makes up the genetic code. There are also three codons which determine the end of a protein; these are known as stop codons. A gene may occupy a sequence of one to tens of kilobases (one kilobase = 1000 base pairs), depending on the length of the resultant protein and the number of introns in the gene. A chromosome consists of a continuous molecule of DNA coding for many thousands of genes. In a chromosome, DNA is intricately coiled and combined with proteins, including histones, which results in a compact structure that can be seen with a light microscope. Often, special stains are used which create a pattern of light and dark bands. These band patterns are specific for each chromosome, and reflect their macrostructure. Figure 3.3 demonstrates the relative size of a gene, in relation to a chromosome band and the chromosome itself.

TheGeneticBasis of Disease Robin M Winter

DMA, genes and chromosomes 57 Ways of looking at genes 58

Loci, alleles and segregation 63 Chromosomal disorders 65 Single gene disorders 69

Types of single gene mutation 60

Multifactorial disorders 72

Chromosome studies 61

Management of genetic disorders 74

Cell division 63

In the western world, chronic diseases with a substantial genetic component occur in 5-10% of the adult population. Furthermore, more than one in 50 babies has a recognizable congenital anomaly at birth, many of which will have a genetic cause. Genetic disorders and malformations account for around 30% of paediatric hospital admissions and up to 50% of deaths under the age of 15 years. The relative frequency of the different types of genetic disorders in the UK is shown in Figure 3.1. In certain populations, specific genetic disorders are of such high incidence as to represent significant public health problems, e.g. the thalassaemias in Mediterranean and oriental populations, and sickle cell anaemia and other haemoglobinopathies in Afro-Caribbean populations. The incidence of affected homozygotes with these disorders can be as high as one in 50 births in some areas.

DNA, GENES AND CHROMOSOMES The basic unit of inheritance is the gene. A gene is a length of deoxyribonucleic acid (DNA) that codes for a specific protein. The DNA molecule is a double helix, i.e. it is formed from two interwoven strands. Each strand consists of a chain of nucleotides, and each nucleotide is made up of a deoxyribose sugar and a nitrogenous base. There are four possible nitrogenous bases: adenine (A), thymine (T), guanine (G) and cytosine (C). The deoxyribose sugars

FIG. 3.1 The relative frequencies of different types of genetic disorder in the United Kingdom

57

FIG. 3.2 DNA transcription and translation (After Emery A E H 1984 An introduction to recombinant DNA. John Wiley, London.) See text of details.

FIG. 3.3 The size of a gene relative to a chromosome and chromosome band, illustrated by the b-globin gene This is situated on chromosome 11 in band 11p15.5, which may contain around 4000kb. The gene itself is around 1600 base pairs long and is part of a cluster of similar genes, the (3-like globin gene cluster.

WAYS OF LOOKING AT GENES SOUTHERN BLOTTING There are three basic steps in the process of Southern blotting for the analysis of the sequences of DNA that make up genes (Fig. 3.4):

58

1. Use of restriction enzymes. Restriction enzymes, which originate from bacteria, cut double-stranded DNA at

FIG. 3.4 DNA analysis S DNA is cut at specific points by restriction enzymes into different sized fragments. H These fragments are separated by electrophoresis on an agarose gel and blotted onto nitrocellulose paper (Southern blotting). Specific genes cannot be visualized at this stage, [c] A radioactive DNA probe is used to locate a specific gene. The position of the gene on the gel depends on the size of the DNA fragment in which it is contained.

specific base sequences. The distance between such restriction enzyme recognition sites may range from one to tens of kilobases, so that the genome is cut into fragments of differing sizes. 2. Electrophoresis and Southern blotting. Fragments generated using restriction enzymes can be separated by electrophoresis on an agarose gel. The fragments are then denatured (i.e. double-stranded DNA separates into single-stranded DNA) and a mirror image of the pattern of fragments is made on to nitrocellulose paper by a process known as Southern blotting. 3. Use of DNA probes. At this stage, single-stranded fragments of DNA have been separated according to size, and fixed on to the nitrocellulose filter paper. A gene probe must now be used to identify the individual gene sequences present in specific bands. A gene probe is a radioactively labelled single-stranded length of DNA whose sequence is specific for a particular gene, or length of chromosome. It is sometimes made using mRNA coding for a specific protein, by a process known as reverse transcription. The single stranded probe recognizes complementary sequences of DNA on the Southern blot, and anneals to them. Autoradiography can be used to locate the position of the gene probe, which will mark the position of all similar sequences on the Southern blot. If a gene-specific probe has been used, only a few bands will be seen on the autoradiograph, according to how many restriction enzyme recognition sites there are within the gene itself. Different

3

restriction enzymes will generate different patterns, using the same probe. The position of the various restriction enzyme cutting sites within the gene can be inferred, and a restriction map produced.

POLYMERASE CHAIN REACTION (PCR) The polymerase chain reaction (PCR) is a way of amplifying specific stretches of DNA so that further genetic analysis can be carried out on them. The revolutionary nature of the technique derives from the fact that large amounts of specific DNA sequence can be copied from only a few (or even a single) DNA molecules. At the start of the process the DNA double helix is denatured by heat to give two single strands. Two specific primers are then needed to amplify the specific sequence of interest. The primers are made up of single-stranded DNA with a specific complementary sequence to the sequences flanking the target DNA sequence (Fig. 3.5). DNA polymerase is then used to copy the target DNA, starting from the primer sequence on both of the original DNA strands. At the end of this process of DNA extension two double helices containing the target sequence will have been made. These double helices are then denatured and the process is repeated several times. At each step the number of copies of the target sequence doubles.

POLYMORPHIC MARKERS Restriction fragment length polymorphisms (RFLP) The restriction enzyme cutting sites in and around genes are variable, as a common, harmless base change may create or destroy a cutting site. Thus, in some individuals, different patterns may be seen on Southern blotting using the same restriction enzyme and the same gene probe. This is because the gene in question will be contained in fragments of different lengths according to the presence or absence of specific cutting sites. Such restriction fragment length polymorphisms (RFLPs) are extremely important because they act as genetic markers, allowing abnormal genes to be tracked through families (Fig. 3.6).

Short tandem repeat polymorphisms (STRPs) The first generation of DNA markers involved RFLPs in known genes or DNA sequences. Use of these markers required Southern blotting. More modern maps use short tandem repeat polymorphisms (STRPs), which are detectable by PCR. The genome contains many stretches of repeated di-, tri- or tetranucleotides of variable length. One of the commonest is the CA repeat. At a specific chromosomal location in an individual there may be (for

FIG. 3.5 Polymerase chain reaction

example) a (CA)10 repeat on one chromosome and a (CA)20 repeat on the other. This length difference can be detected and is the basis of a polymorphism that can be used for linkage studies. Tetranucleotide repeat sequences are easier to analyse and are the basis of the latest generation of genetic maps.

59

TABLE 3.1 Molecular mechanisms of single gene mutation, illustrated by the haemoglobinopathies Abnormality of DNA

Effect

Example

Deletion of gene

Diminished or absent protein, depending on the number of genes involved

Most forms of a-thalassaemia

Point mutation (i.e. single base change)

Substitution of a single amino acid in the protein

Sickle cell anaemia, Hb E, C and many other examples Hb McKees Rocks, some forms of (3-thalassaemia Hb Constant Spring (a-chain)

Creation of a stop codon leading to a short protein chain Removal of a stop codon, leading to an elongated protein chain Alteration of splicing signal

FIG. 3.6 Restriction fragment length polymorphism (1) Shows the two alleles E and H of a gene at a specific locus. The alleles are represented by the coloured blocks on each of the two chromosomes. The arrows indicate restriction enzyme recognition sites. In allele A there is an extra site in the middle of the gene, so that when a DNA sample from this individual is cut with the restriction enzyme and a Southern blot made, three bands will be seen (2). The largest band represents allele B and the two smaller bands allele A. (3) Shows a pedigree where alleles A and B are being passed from generation to generation. Allele B causes an autosomal dominant disorder (e.g. neurofibromatosis) in this family so that the RFLP can be used to track the abnormal gene.

Some forms of a- and p-thalassaemia

Insertion or deletion of a number of nucleotides that are not a multiple of three

Frameshift mutation

Hb Cranston, Tak and Wayne Some forms of p-thalassaemia

Fusion of genes

Absence of normal product and presence of a new protein resulting from the fused genes

Lepore haemoglobins involving (3- and 8-globin chains ((3-thalassaemias)

TRINUCLEOTIDE REPEATS Single-nucleotide polymorphisms (SNPs) The most common type of polymorphism in the human genome is a single base pair change known as a singlenucleotide polymorphism (SNP). These SNPs can now be detected by oligonucleotide microarrays (so called 'gene chips'). This will allow for rapid typing of thousands of polymorphic changes in marker loci and known genes that might predispose to disease. A major project is under way to map thousands of SNPs throughout the human genome.

TYPES OF SINGLE GENE MUTATION

60

At the molecular level there are many mechanisms that can give rise to a malfunctioning gene (Table 3.1). The haemoglobinopathies best illustrate the full range of molecular defects that can occur. A mechanism of mutation, only recognized in recent years, involves the expansion of trinucleotide repeats.

An unusual form of mutation consists of expansion of a CAG sequence within a gene, sometimes leading to a sequence of polyglutamine within the coded protein. This disrupts its function. Table 3.2 gives details of eight disorders caused by trinucleotide repeat mutations. For reasons unknown, they all involve abnormal neurological function. Trinucleotide repeat mutations often show the phenomenon of anticipation. This means that the severity of disease appears to increase as the gene is passed down the generations of a pedigree. The clearest example is in myotonic dystrophy, where a grandparent may just have cataracts without overt myotonia, a daughter may have the classic 'adult' form of the disease with myotonia, and her child may have the severe congenital form of the disorder with severe early hypotonia and mental retardation. The explanation for this is that the trinucleotide repeat expands (i.e. the number of repeat sequences increases) each time the gene is copied and passed on to the next generation. In fragile X syndrome premutation carriers can occur. These are males or females who have no clinical or cytogenetic abnormalities but who carry an increased number of trinucleotide repeats in the fragile X gene. When passed on by a female such a premutation can mutate to the full

3

TABLE 3.2 Triplet repeat mutations and human disease Disease

Clinical features

Chromosome location

Huntington's chorea

Choreoathetosis, dementia

4p16

Spinocerebellar ataxia typel

Cerebellar ataxia

6p21

Friedreich's ataxia

Cerebellar ataxia

9q13-21

Machado-Joseph disease

Cerebellar ataxia, extra-pyramidal signs, ophthalmoplegia

14q32

Myotonic dystrophy

Myotonia, frontal balding, cataracts, diabetes, male infertility

19q13

Dentato-rubropallido luysian atrophy

• Chorea, seizures, dementia

12p12

Chromosome 1

Spinobulbar muscular atrophy (Kennedy's syndrome)

Muscular atrophy/fasciculation, gynaecomastia

Xq21

Fragile X mental retardation

Non-specific mental retardation, fragile X chromosome on cytogenetic examination

Xq28

References Mandel J-L 1993 Questions of expansion. Nature Genetics 4: 8-9 Richards R I, Sutherland G R 1992 Dynamic mutations: a new class of mutations causing human disease. Cell 70: 709-712

mutation (where the repeat is expanded) and a fully affected child can be born.

CHROMOSOME STUDIES Chromosomes are usually visualized during mitosis, at metaphase (p. 63). At this stage the DNA is coiled and condensed, and the chromosomes consist of two identical chromatids, joined together at the centromere (Fig. 3.7). The position of the centromere is specific for each chromosome. In some chromosomes (1-3) the centromere is close to the centre (metacentric); in others (13-15, 21-22) it is near one end (acrocentric), and in the remainder the centromere is somewhere in between these two extremes (submetacentric). By convention, all chromosomes are divided by the centromere into a short arm (p) and a long arm (q).

CHROMOSOME BANDING Like other cell structures, chromosomes must be stained in order to visualize them with the light microscope. Using various methods, a pattern of bands can be produced which

FIG. 3.7 Chromosome nomenclature, using chromosome 1 as an example

is specific for each chromosome. A number of staining techniques are commonly used: • G (Giemsa) banding is the most widely used technique. Light and dark bands are produced. • Q (quinacrine) banding requires a fluorescent microscope. The band patterns are similar to G-banding. Certain chromosome structures, e.g. the long arm of the Y, are particularly well stained. • R (reverse) banding produces a negative image of G-banding. • T (telomere) banding preferentially stains the telomeres (ends of the chromosomes). • C (centromere) banding stains centromeric regions.

Band nomenclature The banding pattern is specific for each chromosome, and a convention exists for labelling each band. The long and short arms are divided into regions by particularly prominent bands (landmark bands) (Fig. 3.7). Within each region, individual bands are numbered from the centromere distally, and landmark bands are included in the distal region they demarcate. Thus, in Figure 3.7 band Iq32 means band 2 in region 3 of the long arm (q) of chromosome 1.

FLUORESCENT IN SITU HYBRIDIZATION (FISH) Fluorescent in situ hybridization (FISH) is an important technique for recognizing very small deletions of chromo-

61

Abnormalities in chromosome number The normal diploid number of chromosomes in humans is 46. Individuals with chromosome counts that are not multiples of the normal haploid number (23) are said to be aneuploid. A fetus can receive higher multiples of the haploid number of chromosomes to give 69 (x3) or 92 (x4) chromosomes. Such triploid or tetraploid fetuses usually miscarry early. If individuals carry one extra chromosome, they are said to be trisomic for the chromosome concerned. Very occasionally, an individual can have two extra chromosomes (double trisomy). Monosomy is the situation where one chromosome is missing. In liveborn children it is virtually only seen for the sex chromosomes, giving monosomy X (Turner's syndrome, p. 69).

Translocation, deletion and insertion

FIG. 3.8 Chromosome FISH in a patient with Williams' syndrome One probe (white arrows) identifies the two No 7 chromosomes. The other probe (elastin, black arrow) is missing from one chromosome 7 indicating a microdeletion.

some material. The technique involves preparing a DNA probe, specific for a particular gene or region of the chromosome. This probe is then labelled with a fluorescent tag. A chromosome slide is prepared and the probe is used to label the specific chromosomal region containing the gene under investigation. For a gene on the autosomes, two labelled tags should show up. If there is a deletion on one chromosome, then only one tag will be visible. FISH can detect chromosome deletions that are much smaller than those visible by conventional light microscopic techniques. Another technique involves chromosome painting. In this technique, probes derived from the whole length of a specific chromosome are labelled with a fluorescent dye (Fig. 3.8). If these probes are then used in the same way as the FISH technique, the whole chromosome is labelled with the fluorescent dye. By using different fluorescent dyes and computer image manipulation, each of the 24 chromosomes can be labelled in a different colour. Chromosome painting is useful for detecting subtle or complicated translocations of the chromosomes (see below).

TYPES OF CHROMOSOME ABNORMALITY

62

Chromosome abnormalities can be divided into those involving the number of chromosomes, and those involving breakage or rearrangement.

Breakage of chromosomes can result in loss of material (deletion), exchange of material between two nonidentical chromosomes (reciprocal translocation), inversion of a segment of chromosome (pericentric inversions involve the centromere, paracentric inversions do not), and joining of the two ends of the same chromosome (to give a ring chromosome). Horizontal splitting of the centromere can result in a chromosome with identical long and short arms, an isochromosome. A translocation between the long arms of two acrocentric chromosomes, joined at the centromere, is known as a Robertsonian translocation. These possibilities are illustrated in Figure 3.9.

KARYOTYPE NOMENCLATURE The convention in karyotype nomenclature is to give the number of chromosomes first, followed by the types of sex chromosomes, followed by the types of any additional, missing or abnormal chromosomes. Thus, for example: • • • • •

46,XX - normal female 45,X - Turner's syndrome 47,XY + 21- male with trisomy 21 69,XXY - triploidy, XXY sex chromosome complement 45,X/46,XX - mosaic Turner's syndrome.

For inversions or translocations, the numbers of the chromosomes involved are given in brackets (the chromosome with the lowest number being given first). This is followed by the bands or regions involved in further brackets. Symbols are used to indicate the type of rearrangement involved. Thus, 46,XX,t(9:21) (qll :pll) karyotype means that chromosome 9 has broken at band qll and chromosome 21 has broken at band pll. The small letter t (for translocation) indicates that material has been exchanged between the two chromosomes.

3

FIG. 3.9 Types of chromosome aberration ffl Two normal non-homologous chromosomes. DEI A reciprocal translocation between the two chromosomes. Note that material has been exchanged between the short arms. Q-] A pericentric inversion of the large chromosome. Breaks have occurred in the long and short arms and the middle segment of the chromosome has become inverted. [B] A deletion. The top part of the large chromosome has been lost. DEI An isochromosome. The small chromosome consists of two identical short arms due to misdivision after replication. DEI A ring chromosome. Both ends of the small chromosome have joined to form a ring.

CELL DIVISION MITOSIS Division of somatic cells is known as mitosis. This is divided into four stages (Fig. 3.10). Between cell divisions, in interphase, the chromosomes are extremely elongated and cannot be seen with the light microscope. 1. Prophase. As a preparation for cell division, the chromosomes replicate to give the classic structure with two chromatids, and they begin to condense. 2. Metaphase. The chromosomes line up at the centre of the cell. 3. Anaphase. Each chromosome separates longitudinally at the centromere, and the two chromatids pass to opposite poles of the cell. 4. Telophase. Two new daughter cells are created by the formation of a nuclear envelope around each new set of chromosomes.

FIG. 3.10 Mitosis and meiosis See text for details.

different gametes of an individual. This is brought about by two separate means: 1. Different gametes receive different combinations of maternally and paternally derived chromosomes. 2. There is an exchange of genetic material between the two members of an homologous pair of chromosomes, known as crossing over. This process results in the creation of a chromosome containing part of the maternal and part of the paternal member of an homologous pair, known as a recombinant chromosome. Without crossing over, alleles on the same chromosome would always be inherited en bloc. Crossing over takes place during the first step of meiosis (meiosis I), which is divided into prophase, metaphase, anaphase and telophase, as in mitosis. However, meiosis I differs in that homologous members of a pair of chromosomes line up and exchange genetic material during prophase, whereas at anaphase each member chromosome passes to an opposite pole without division at the centromere. Meiosis I is known as reduction division, because it results in the formation of two daughter cells containing 23 chromosomes, each of which has two chromatids that may not be identical because of crossing over. The second step, meiosis II, is similar to mitosis except that each 'parent' cell now has only 23 chromosomes (each with two chromatids), and each 'daughter' cell contains 23 chromatids at telophase. Fusion of parental gametes at fertilization restores the diploid number of chromosomes.

MEIOSIS During the production of gametes, the normal diploid complement of 46 chromosomes must be reduced to the haploid number of 23 in the gametes by the process known as meiosis. During meiosis there is the independent assortment of maternally and paternally derived genes in the

LOCI, ALLELES AND SEGREGATION The genes determining a particular trait are situated at identical points or loci on each member of a homologous pair of chromosomes. The two genes on each chromosome

63

FIG. 3.11A Inheritance of alleles at autosomal loci

FIG. 3.12A Autosomal recessive inheritance with first cousin parents The boy at the bottom of the pedigree is homozygous for an abnormal gene inherited from the great-grandmother through both parents. The double line indicates consanguineous marriage.

FIG. 3.11B Inheritance of alleles at X-linked loci

at a particular locus may not be identical, and are known as alleles. If two alleles at a locus are identical, the individual is said to be homozygous at that locus; if the alleles are not identical, the individual is said to be heterozygous. The chance of a child inheriting one particular allele from a heterozygous parent is 1 in 2 (50%). For autosomal loci, if both parents are heterozygous, the chance of a child being homozygous for one particular allele is 1 in 4 (25%) (Fig. 3.11A). If a locus is on the X chromosomes, males can only carry one allele (because they only have one X chromosome). They are said to be hemizygous for a particular allele. All daughters of a male hemizygous for an abnormal gene will receive that gene; on the other hand, none of the sons will receive the abnormal gene (because a male must pass on a Y chromosome to his sons, and not an X). Each time a heterozygous female has a child, there is a 50% chance it will be a boy, and if it is a boy, a 50% chance that he will be affected. Likewise there is a 50% chance of a girl and a 50% chance she will be heterozygous (Fig. 3.11B).Q

O MCQ 3.1

64

O MCQ 3.2

FIG. 3.126 Autosomal dominant inheritance Note that the disease is passed on from parent to child and that both males and females are affected. There is also transmission of the disease from father to son.

DOMINANT AND RECESSIVE INHERITANCE Disorders where it is only necessary to carry one abnormal allele in order for an individual to be affected are said to be dominant. In autosomal dominant disorders, the abnormal gene is carried on a non-sex chromosome; in X-linked disorders, the gene is carried on the X chromosome. In recessive inheritance an individual is only affected if he or she does not carry a normal allele at all. For autosomal recessive inheritance, affected individuals will be homozygous for the abnormal allele and both of the parents will be heterozygote carriers. In X-linked recessive disorders males hemizygous for the abnormal allele will be affected, whereas heterozygous females will be unaffected. Typical pedigree patterns are shown in Figures 3.12 and 3.13.2

RECENT ADVANCES THE HUMAN GENOME PROJECT

FIG. 3.13. X-linked recessive inheritance Note that half of the sons of a carrier female are affected and half of the daughters are carriers. All the daughters of an affected male are carriers and all the sons are normal, because a male passes on his one X chromosome to his daughters and his Y chromosome to his sons.

NEW MUTATION, PENETRANCE AND GENE EXPRESSION Abnormal alleles have to arise somewhere in a pedigree by the process of new mutation. Sometimes this can be inferred in a case where an affected individual for an autosomal dominant disorder has normal parents. However, many autosomal dominant disorders also display incomplete penetrance or reduced expression. This means the gene carriers sometimes show no signs, or only a few minor signs, of the condition. The two phenomena are therefore important in the context of genetic counselling. If a person has an autosomal dominant disorder but there is no family history, both parents must be examined very carefully in order to determine whether either of them carries the abnormal gene. If they do not, the chance of their having a further affected child will be small (as this would require a second, unconnected, new mutation). On the other hand, if one parent shows very minor signs of the condition the recurrence risk would be 50% for a further affected child.

GENE MAPS AND LINKAGE The presence of loci close to one another on chromosomes leads to the phenomenon of linkage. The closer together loci are on a chromosome, the lower the probability that a crossover will take place between them during meiosis. I, and the higher the probability that two alleles inherited from the parent will be passed on together to an offspring. Linkage between loci can be studied in suitable families by looking at the joint inheritance of specific traits, and an estimate can be made of the distance separating the loci. These family studies are informative for linkage but do not give information about the chromosomes on which the loci are situated. This problem is known as gene assignment and can be tackled by a number of different methods, such as somatic cell hybrids. Using these methods, a human gene map is being constructed giving information about the

3

There are thought to be around 100000 genes in humans, and the total number of DNA bases in the haploid genome is around 3000000000 (3000 megabases). The ultimate goal of the human genome project is to locate every gene and to understand its function. It may not be absolutely necessary to sequence all of the 3000 megabases, as genes are known to be separated by large tracts of non-coding DNA sequences that may have no function. The first phase of the project was to create dense genetic maps of polymorphic genetic markers so that genetic diseases and traits can be located by linkage analysis. Markers are evenly spaced every megabase or so. At the same time the genome was broken up into overlapping DNA fragments (contigs) to provide a physical map. Gradually the genetic markers, known genes, and also parts of the coding sequences from expressed genes of unknown function (Expressed Sequence Tags - ESTs), are being located on contigs. This means that if researchers map a disease to a marker by linkage studies they are able to locate the marker on a physical fragment of DNA (the contig) and ask if any candidate genes are present on the same piece of DNA. If a suitable candidate is not found, the overlapping DNA fragments on either side can be examined and so on. The first draft of the entire sequence of the human genome was completed in the year 2000 and contains 90% of the entire sequence. The final sequence, with 99.99% accuracy, is expected to be completed by 2003, or perhaps earlier. The entire gene sequence of model organisms is also being examined. The fruitfly (Drosophila), an important organism in studying the mechanisms of embryonic development, has already been sequenced. Progress is being made in sequencing the genome of the mouse, which provides important models for many medical disorders. assignment of various loci and their distance from others on the same chromosome.

ICHROMOSOMALDISORDERS FREQUENCY OF CHROMOSOMAL DISORDERS About seven per 1000 babies are born with detectable chromosomal anomalies. Of these, three will have sex chromosome anomalies, 1.5 autosomal trisomies (mainly Down's syndrome) and the remaining 2.5 will have chromosomal rearrangements. Most of the latter are balanced, i.e. no genetic material is lost or gained. Of the autosomal trisomies, only trisomy 21 (Down's

65

syndrome) cases are likely to survive into adulthood. The sex chromosome anomalies will frequently present with infertility or amenorrhoea. Unbalanced chromosomal rearrangements usually cause severe physical and mental handicap. However, balanced rearrangements have no effect on the carrier, although they predispose to recurrent miscarriage and children with unbalanced chromosome complements. An unbalanced karyotype means that there is duplication or deletion of some genetic material.O

TABLE 3.3 Chromosomal types of Down's syndrome Type of Down's syndrome

Chromosome complement

Comment

Standard trisomy 21

47,XX,+21 or 47,XY,+21

Approximately 95% of Down's syndrome children

Translocation Down's syndrome

e.g. 46,XY,-14,+t(14;21) 46,XX,-21,+t(21q;21q)

In about 2-3% of cases, an extra chromosome 21 is attached to another chromosome. It is essential to examine the chromosomes of the parents, as one may carry a balanced translocation

Mosaic Down's syndrome

e.g. 46,XX/47,XX,+21

Can be demonstrated in about 2-3% of cases

DOWN'S SYNDROME Between one in 700 and one in 1000 liveborn infants have Down's syndrome. The characteristic physical features are well described in paediatric texts and are present at birth, allowing immediate clinical diagnosis in most cases (Fig. 3.14). The head is foreshortened (brachycephaly) and there are usually three fontanelles. The eyes have an upward slant and crescents of skin are present covering the inner canthi (epicanthic folds). Brushfield spots are commonly seen. These are small white spots, situated around the outer third of the iris. The nasal bridge is flat and the tongue protuberant and deeply furrowed. A single transverse palmar crease is present on the hand, together with a short, incurved fifth finger (clinodactyly). A wide gap is present

between the first and second toes, often with a longitudinal plantar crease. Hypotonia and delayed development are universal. Congenital heart defects (particularly A-V canal defects) and duodenal atresia can occur. It is essential that the chromosomal type of a Down's syndrome child is obtained. There are three possibilities (Table 3.3): • In about 95% of cases there is an extra, free-lying chromosome 21. This form of Down's syndrome is associated with increased maternal age. The risk at a maternal age of 37 years is 1 in 250, at 40 years 1 in 100, and at 45 years 1 in 40. • In 2-3% of cases there is a chromosome translocation, usually involving Robertsonian translocation (p. 62). The parents of such individuals must be checked to see whether they carry a balanced translocation. • The remaining cases are mosaic. In these infants the degree of mental retardation can vary according to the proportion of abnormal cells. The recurrence risks for Down's syndrome are set out in Table 3.4.

TRISOMY 18 (EDWARDS' SYNDROME)

FIG. 3.14 Down's syndrome

1

66

Case 3.1

This autosomal trisomy occurs in approximately one in 5000 live births. There is usually low birthweight, hypotonia, a prominent occiput, a small chin, a short sternum and characteristic hand and foot abnormalities. The index and little fingers of the hand overlap the middle two and the nails are small; in the feet the hallux is dorsiflexed and the lateral profile of the foot is convex, with a prominent heel (a so-called 'rocker-bottom foot', caused by a vertical talus). Analysis of the dermatoglyphic patterns of the digits reveals an increased number of arch patterns. Internal malformations are common, especially congenital heart defects (ventricular septal defects, patent ductus arteriosus) and renal abnormalities (horseshoe kidney,

3

CASE STUDY 3.1 GENETIC COUNSELLING DOWN'S SYNDROME A couple is referred for genetic counselling because they have had a child with Down's syndrome. Question 1. What facts do you need to know before being able to provide accurate genetic counselling? It is crucial to know the age of the parents, particularly the mother. It would also be important to know whether there was any other family history of Down's syndrome. It is essential to know the precise karyotype of the baby. In this case, the karyotype was 46,XY,+t(14;21). This is a karyotype of a male with Down's syndrome, caused by a chromosome translocation of a chromosome 21 to a chromosome 14. Question 2. What further tests would you need to carry out before being able to counsel the parents?

Discussion Most cases of Down's syndrome have standard trisomy 21. In other words, there are 47 chromosomes, and the extra chromosome is a free-lying chromosome 21. The chances of having a baby with this form of Down's syndrome increases with the age of the mother. Recurrence risks after a baby with standard trisomy 21 are mainly related to the age of the mother, although they are slightly increased over the general population risk at any maternal age. Two to 3% of babies with Down's syndrome have a chromosome translocation. In this case it is essential to test the karyotype of the parents. In many cases both parents might have a normal karyotype, in which case the recurrence risk would be small. Nevertheless, one parent can carry a balanced chromosome translocation, as in the current case. Where the mother carries a balanced

14;21 chromosome translocation, the risks of a further baby with Down's syndrome are about 10%. In a future pregnancy an amniocentesis or chorionic villus sample (CVS) can be offered to check the chromosomes of the baby. Serum screening tests for Down's syndrome are not appropriate in this situation, as they do not provide a definitive answer as to whether the baby has Down's syndrome. The balanced translocation may also be carried by other family members. In this family, the wife's parents should be tested if possible, to see whether they carry the translocation. If one of them does, then testing should be offered to other family members, such as the wife's siblings. The other normal children of the present couple should be offered testing when they are of an age to give informed consent.

It is essential to examine the chromosome types of the parents, as one of them may be a translocation carrier. In this case, the mother had a 45,XX, +t(14;21) karyotype. This means that she was a balanced carrier for a 14;21 translocation (Case Fig. 3.1.1). Question 3. What risks would you give for a baby with Down's syndrome in a further pregnancy, and what tests could be carried out in a future pregnancy? 4. What further tests must be carried out?

CASE FIG. 3.1.1 A balanced 14;21 chromosome translocation (arrow) Note that there is only one chromosome number 21 and 45 chromosomes in all. Clinically this is a normal female.

hydronephrosis). Aplasia of the radius, facial clefts and exomphalos each occur in about 20% of cases. Prognosis is very poor, with 50% of cases dying before 2 months, 90% before 1 year and 99% before 10 years; long-term survivors are very severely retarded. Most cases are caused by an extra chromosome 18 (rather than a chromosome translo-

cation) and there is an association with increased maternal age. The risk of recurrence after a trisomic case is about 1 in 100 for a baby with a chromosomal trisomy (not necessarily trisomy 18); however, these risks might have to be modified for older mothers (see section on Down's syndrome).

67

TABLE 3.4 Recurrence risks for Down's syndrome Risk to other relatives (excluding offspring of an affected patient)

Type of Down's syndrome

Risk to siblings of affected patient

Standard trisomy 21 Mother <40 years old

1%

Mother >40 years old

Maternal age risk

Translocation (Parents have normal chromosomes)

Maternal age risk

Maternal age risk

10%

Depends on chromosomes of relative Depends on chromosomes of relative Depends on chromosomes of relative

Translocation (One parent carries the translocated chromosome) Mother carries 14:21 translocation Father carries 14:21 translocation Either parent 21:21 translocation

2% 100%

1%

Mosaic

Probably similar to maternal age risk

Maternal age risk

TRISOMY 13 (PATAU'S SYNDROME) This occurs in about one in 8000 live births. There are significant craniofacial abnormalities, with punched-out scalp lesions (aplasia cutis), microphthalmia, and low-set, malformed ears. Many cases have severe bilateral cleft lip and palate and some cases have hypotelorism or cyclopia (a single central eye with a proboscis-like nose), indicating underlying holoprosencephaly of the brain (failure of midline separation of the cerebral hemispheres). In the hands and feet an extra, hypoplastic digit on the ulnar or fibular side is common (postaxial polydactyly). Brain, heart and kidney malformations are very common. Survival past the first year of life is very unusual and retardation is severe. Most cases arise by standard trisomy, but some are the result of a Robertsonian translocation involving chromosome 13. The risk of recurrence after a trisomic case is about 1 in 100 for a baby with a chromosomal trisomy (not necessarily trisomy 13); however, these risks might have to be modified for older mothers (see section on Down's syndrome).

1

68

MCQ 3.3

FIG. 3.15 Turner's syndrome (45,X) Characteristic fades, webbing of the neck, absent breast development and widely spaced nipples.

SEX CHROMOSOME ABNORMALITIES Klinefelter's syndrome Klinefelter's syndrome (47,XXY) occurs in about one in 1000 males. Adult individuals tend to have eunuchoid proportions with female fat distribution and gynaecomastia. The testes are small. Pituitary gonadotrophins are raised and androgen production is low. Affected individuals are sterile, as they produce no sperm. Androgen therapy may help to improve libido and encourage the development of male secondary sexual characteristics. Intelligence is usually normal, although mild to moderate retardation may be seen.

XYY males About one in 1500 males is XYY. Such individuals are physically normal, although they tend to be tall. Although an increased tendency to psychopathic behaviour has been reported, prospective studies have shown that the majority of affected individuals lead normal lives.

Turner's syndrome The incidence of Turner's syndrome (45,X) is one in 5000

births. Affected individuals are short, tend to have a web of skin on the lateral borders of the neck (Fig. 3.15), and may have other physical abnormalities, such as a broad chest with widely spaced nipples and an increased carrying angle at the elbows. Coarctation of the aorta can occur. The vagina and internal genitalia are normally formed, but the ovaries consist of simple streaks of fibrous tissue without ovarian follicles, so that primary amenorrhoea is the rule. Treatment consists of oestrogen replacement around puberty to promote the development of secondary sexual characteristics.

cific oncogenes (see p. 149). In other cases the precise nature of the genetic change is not known. Occasionally, an individual with a specific chromosome abnormality will be predisposed to develop particular types of cancer (see Table 6.3, p. 150). A group of single gene disorders exists that result in an increased tendency to chromosome breakage, owing to abnormalities of DNA repair after damage by agents such as ultraviolet light and radiation. These disorders, most of which are autosomal recessive, also predispose to malignancies (see Table 6.3, p. 150).

3

47,XXX females Females with a chromosome constitution 47,XXX have no physical abnormalities. However, primary and secondary amenorrhoea may occur, and average IQ is 20-30 points lower than in XX females, so that learning difficulties are more common than in the general population. Nevertheless, the majority of affected females (about 70-80%) have normal intelligence. 1

CHROMOSOME DELETIONS AND DUPLICATIONS A large proportion of individuals with chromosome abnormalities have the normal number of chromosomes but a small piece of a specific chromosome is missing or duplicated. Sometimes there may be a combination of a chromosome deletion and duplication. Any part of any chromosome can be involved. Significant chromosome deletion or duplication usually causes a combination of congenital malformations, with developmental delay. It is often difficult to predict the precise clinical picture from knowing the specific chromosome deletion or duplication, but there are a number of specific syndromes caused by particular chromosomal regions being affected. Microdeletion syndromes are also being recognized, where the deleted chromosome material must be detected by FISH (see above, page 61). Any child with a combination of malformations, or an unusual facial appearance, together with developmental delay, should have his or her chromosomes examined.

CHROMOSOME ABERRATIONS AND CANCER Cytogenetic analysis of any malignant cell will frequently reveal a non-specific chromosome abnormality. However, there are certain malignancies where a characteristic abnormality is found; in others, the development of a specific chromosome change can be of prognostic significance. In the case of Burkitt's lymphoma and chronic myeloid leukaemia, the chromosome translocations are known to bring together the loci for immunoglobulin genes and spe-

SINGLE GENE DISORDERS Sir Archibald Garrod first put forward the concept of an 'inborn error of metabolism' in 1908 after studying alkaptonuria, a familial disorder in which patients excrete large amounts of homogentisic acid. This led to the concept that an enzyme controlling a metabolic step can be missing or reduced, leading to failure of production of essential metabolites further down the metabolic chain (e.g. in albinism), or the accumulation of toxic substrates of the enzyme (e.g. in alkaptonuria and phenylketonuria) (Fig. 3.16). Today, McKusick's catalogue of genetic disorders lists over 4000 defects involving every organ system. Only a few important groups of disorders are described here.

AMINOACIDOPATHIES Defects in the metabolism of amino acids include a form of congenital hypothyroidism, albinism, disorders of the urea cycle and organic acid metabolism, as well as the disorders described below.

Phenylketonuria Phenylketonuria is an autosomal recessive disorder affecting about one in 10000 infants at birth. It is usually caused by a defect of the enzyme phenylalanine hydroxylase, which converts phenylalanine to tyrosine. The most common clinical presentation, now rarely seen in developed countries because of screening, is vomiting in early infancy, with eczema, depigmentation of skin and hair, hypertonia, seizures, and mental retardation developing during the first years of life. Fortunately, screening at birth is now widely available, in which a drop of blood, obtained within the first few days of life, is placed on a filter paper and a bacterial inhibition assay used to measure phenylalanine levels (the Guthrie test}. Affected infants can be treated by giving an artificial diet low in phenylalanine until at least 8 years of age. If dietary control is strict the prognosis for mental development is good. Females with phenylketonuria run a very high risk of having abnormal children, unless a strict diet is reintro-

69

FIG. 3.16 Partial metabolic pathway of phenylalanine, tyrosine and related compounds showing position of 'metabolic blocks' caused by enzyme deficiencies DO Deficiency of phenylalanine hydroxylase results in an inability to metabolize phenylalanine, leading to phenylketonuria. E Tyrosinase deficiency leads to an inability to produce melanin, thus causing albinism. [3 Deficiency of homogentisic acid oxidase leads to the accumulation of homogentisic acid, and hence to alkaptonuria. OS Deficiency of any one of several enzymes at this point means that thyroxine cannot be synthesized. The result is familial hypothyroidism.

TABLE 3.5 transport disorders Disorder

Site

Metabolites affected

Clinical features

Inheritance

Intestinal disaccharidase deficiency Hypophosphataemic rickets

Small intestine

Maltose, sucrose, lactose

Autosomal recessive

Renal tubules

Phosphates

Abdominal pain and distension, diarrhoea Rickets (vitamin D-resistant)

Hartnup disease

Renal tubules

Amino acids (mono-aminomonocarboxylic group)

Red, scaly facial rash, ataxia

Autosomal recessive

Cystinuria

Renal tubules, small intestine

Cystine, ornithine, arginine, lysine

Renal stones

Autosomal recessive

Nephrogenic diabetes insipidus

Renal tubules

Insensitivity to ADH

Polyuria, polydipsia

X-linked dominant

duced throughout pregnancy. Their children do not have phenylketonuria themselves, but the abnormal metabolites cross the placenta during pregnancy and affect the developing fetus. Microcephaly, cardiac defects and mental retardation are common. Phenylketonuria is a heterogeneous disorder: the most common form results from a mutation at the phenylalanine hydroxylase locus itself. However, a rarer form caused by a defect in dihydropteridine reductase has been described. This enzyme is essential for the recycling of the cofactor tetrahydrobiopterine.

Homocystinuria Homocystinuria is an autosomal recessive disorder and affects one in 200000 infants. It is due to a defect of cystathionine (3-synthase and leads to the accumulation of homocysteine, methionine and other metabolites. Clinical 1 70

MCQ 3.4

X-linked dominant

manifestations include lens dislocation, mental retardation, osteoporosis, and thrombosis of arteries and veins. Affected individuals tend to be long and thin, making Marfan's syndrome part of the differential diagnosis. Treatment involves a low-methionine diet, and pyridoxine treatment in that proportion of patients who are vitamin B6 responsive.

Cystinosis Cystinosis, an autosomal recessive disorder, should be distinguished from cystinuria (see Table 3.5). There are abnormally high intracellular levels of free cystine, which lead to crystal deposition in the kidneys, conjunctiva, bones and leukocytes. Presentation is in childhood, with photophobia and crystalline deposits in the conjunctiva, and the effects of renal tubular malabsorption (Fanconi's syndrome). Renal failure can occur before the age of 10 years. Alkaptonuria Alkaptonuria is a rare autosomal recessive disorder caused by a defect in the enzyme homogentisic acid oxidase

(Fig. 3.16). Homogentisic acid, a byproduct of the metabolism of tyrosine and phenylalanine, cannot be metabolized and so accumulates in the body tissues and is excreted in increased quantities in the urine. Classically, the urine turns dark on standing and the sweat may be black, but these findings are not invariable. The homogentisic acid and resultant polymers also stain cartilage and other connective tissue; ochronosis is the name given to the generalized blue/grey pigmentation that results and is best seen in the cartilage of the ear and the sclerae. It can also be seen in the tendons of the hands, through the overlying skin, in older patients. The main clinical problem is arthritis, which affects the hips, knees, shoulders and spine. Onset is unusual before middle age. Radiological changes include narrowing and calcification of the intervertebral discs, together with some fusion of the vertebral bodies. Rarer complications include an increased incidence of ischaemic heart disease, calcification of the cardiac valves, and occasionally arterial aneurysms. Treatment is symptomatic.1

TABLE 3.6 Storage disorders Example

Symptoms

Glycogen storage disorders

12 types including: Von Gierke's disease (glucose-6-phosphate deficiency) Pompe disease (acid maltase deficiency)

Hepatomegaly, hypoglycaemia, muscle weakness, cardiac failure (Symptoms vary according to type)

Niemann-Pick disease Gaucher's disease Krabbe's disease Metachromatic leucodystrophy

Hepatosplenomegaly (in some), progressive neurological deterioration

Tay-Sachs disease

Progressive blindness and neurological deterioration

At least 8 types including: Hurler syndrome, Hunter syndrome, MaroteauxLamy syndrome, Morquio's syndrome, Sanfilippo's syndrome

Hepatosplenomegaly, abnormal bones (dysostosis multiplex), corneal clouding, mental deterioration

Sphingolipidoses (p. 1413)

Gangliosidoses (p. 1413)

Mucopolysaccharidoses (p. 1029)

3

The transport of metabolites across cell membranes often involves specific binding sites and transport proteins. Mutations can disrupt these mechanisms and result in a group of disorders characterized by the failure of essential metabolites to enter the appropriate body compartments. Some important disorders are outlined in Table 3.5.

STORAGE DISORDERS Macromolecules in the body are often in a perpetual state of accumulation and breakdown. Thus, if an enzyme needed to break down a macromolecule is deficient, increased accumulation occurs, leading to a variety of symptoms according to the type of molecule and the site involved. Storage material usually accumulates in the lysosomes. Several examples are given in Table 3.6.

CONNECTIVE TISSUE DISORDERS

Type

(p. 1028)

TRANSPORT DISORDERS

TABLE 3.7 Types of collagen and their functions

The main protein of connective tissue is collagen, of which there are at least four different types, each with its own functions (Table 3.7). Defects of collagen synthesis can cause a wide variety of disorders (Table 3.8). Some of these have been characterized at the molecular level, and various collagen probes are available for precise diagnosis. The specific collagen chain abnormality can sometimes be deduced from the clinical features (Table 3.8). Thus, some forms of osteogenesis imperfecta are caused by abnormalities of type I collagen, whereas type III collagen deficiency can result in a form of Ehlers-Danlos syndrome, with thin, atrophic skin and a tendency to spontaneous rupture of

TABLE 3.8 Connective tissue disorders Disorder

Clinical features

Ehlers-Danlos syndrome (At least 8 distinct forms)

Hyperextensible skin, loose joints, 'cigarette paper' scars, easy bruising, rupture of blood vessels, viscera etc.

Osteogenesis imperfecta (At least 4 distinct forms)

Multiple fractures of bones, blue sclerae, hyperextensible skin, easy bruising

Cutis laxa

Loose skin, premature aged appearance, hernias, diverticula and emphysema

(Several different forms)

Type

Function

Marfan's syndrome

I II III IV

Major collagen in skin, tendon, bones, heart valves and scar tissue Major collagen in cartilage Found mainly in blood vessels and intestine; also in skin Found mainly in basement membranes

Tall stature, span greater than height, loose joints, dislocated lens of eyes, aortic and mitral incompetence, dissection and aneurysms of the aorta

Pseudoxanthoma elasticum

Angioid streaks of retina, linear yellow raised skin lesions, vascular disease

71

CASE STUDY 3.2 LENS DISLOCATION IN A TALL BOY A 14-year-old boy was referred to the genetic clinic because dislocation of the lenses of both eyes had been diagnosed after an episode of blurred vision. He was also noted to be tall.

Marfan's syndrome. Other physical features of Marfan's syndrome are a high-arched palate, pectus excavatum 1 or carinatum, arachnodactyly (long thin fingers) 2, 3 skin striae (stretch marks), particularly over the shoulders, loins and buttocks, and loose joints. In Marfan's syndrome the lens of the eye classically dislocates upwards, whereas in homocystinuria it dislocates downwards. An echocardiogram should be carried out to look for mitral or aortic valve incompetence and dilation of the aortic root, which precedes a possible dissecting aneurysm of the aorta. Urine should be sent to look for homocystine, although a methionine loading test might be necessary where the diagnosis of homocystinuria seems likely. In the present case, span was 2 cm longer than height, which was on the 98th centile. There was a high-arched palate and significant arachnodactyly. The heart was normal. The patient's father was said to have been tall and had died suddenly at the age of 42 from a ruptured aortic aneurysm. There was no evidence of homocystine in the urine. The diagnosis of Marfan's syndrome

Questions ' What is the differential diagnosis?

What would you ask about inthe family history? What investigation would you carryout? Dislocation of the lenses associated with tall stature suggests a connective tissue disorder. The differential diagnosis lies between Marfan's syndrome and homocystinuria. In the family history, one would ask about tall stature in other family members, and heart defects or sudden death caused by ruptured aneurysms or other vascular problems. The latter would suggest Marfan's syndrome. One would also ask about parental consanguinity, which might suggest the recessive disorder of homocystinuria. In the clinical examination one would measure both height and span. A stature above the 90th centile with span greater than height suggests

blood vessels and other internal organs. Defects of the connective tissue protein elastin are thought to cause the different forms of cutis laxa. Both collagen and elastin are abnormal in pseudoxanthoma elasticum (Ch. 12, Table 12.39). Marfan's syndrome has been shown to be caused by a defect in fibrillin protein.

MULTIFACTORIAL DISORDERS Many common malformations, some diseases and many normal traits are said to have a multifactorial origin, in that both genetic and environmental factors contribute to their O Fig. 3.1 0 Fig. 3.2 3

72

Q Case 3.2 0 Fig. 3.4 6

Fig. 3.3 Fig. 3.5

therefore seemed very likely (De Paepe et al, 1996). Discussion Marfan's syndrome is an autosomal dominant condition caused by mutations in the fibrillin gene on chromosome 15. The gene can be examined to look for mutations in suspected cases, but this is technically difficult because of the size of the gene, and in many cases diagnosis relies on clinical findings. Apart from dislocation of the lens, other clinical complications include joint laxity and associated scoliosis, in addition to mitral and aortic valve incompetence and aortic root dilation leading to aneurysms. Prophylactic p-blocker therapy can be given to alleviate the risk of aneurysms. Regular follow-up is needed for cardiac, orthopaedic and ophthalmological problems. Affected individuals have a 50% chance of having similarly affected children.

Reference De Paepe A, Devereux RB, Dietz HC, Hennekam RCM, Reed E (1996) Revised diagnostic criteria for Marfan syndrome. Am J Med Genet 62: 417-426.

aetiology. Furthermore, the genetic component is thought to be due to several genes acting together - so-called polygenie inheritance. A good example is the inheritance of height. The genetic component is determined by several genes, and final height correlates between parents and children. Nevertheless, environmental factors (such as nutrition) can play an important role. The contribution of the many factors involved results in a gaussian (normal) distribution of height in the general population.

COMMON MALFORMATIONS There are many clues which suggest that a common malformation has a multifactorial origin: • After one affected family member, the risk to other firstdegree relatives is usually in the range 1-5%, depending on the condition.

TABLE 3.10 Some common disorders of multifactorial origin, with possible geneti c markers

TABLE 3.9 Recurrence risk (%) in some common multifactorial malformations Normal parent

Condition

Affected parent Risk of second Risk of third Risk of affected child affected child affected child

Cleft palate alone 2 Cleft lip ± cleft palate 4 Neural tube defect 4 Club foot 3 Dislocation of hip 6 Congenital heart disease 1-4 (varies with type) Hirschsprung's disease (long segment): Male index case 8 14 Female index case Pyloric stenosis: Male index case 3 7 Female index case

8

7

10 10 10 10

4 4 3 12

5-10

1-10

-

-

_

4 13

• The presence of additional affected family members increases the risk to subsequent offspring. • Where a condition can have varying degrees of severity (e.g. cleft lip, with or without cleft palate), risks are higher for relatives of an individual with a more severe malformation. • If there is a sex difference in incidence, risks are greater for relatives of an affected individual of the less commonly affected sex. • The risk declines rapidly as one passes from first- to second-degree relatives. It is thought that everyone inherits a liability to a particular malformation. The degree of liability can vary in different individuals and families, according to the number of adverse genes and environmental factors present. If the liability exceeds a given threshold, the malformation is manifested. Individuals from families where there is more than one affected member will, on average, have a higher liability, as will relatives of more severely affected individuals. Similarly, where a condition is less common in a particular sex, affected persons of that sex must carry a greater liability, as will their relatives. Some common multifactorial malformations with their recurrence risks are given in Table 3.9.O0O

COMMON DISEASES, GENETIC MARKERS AND DISEASE ASSOCIATION Many common diseases, e.g. the common allergies and some forms of epilepsy, have been proposed as candidates for multifactorial inheritance. As more research is carried out into these conditions, heterogeneity is recognized and subtypes are defined. Subdivision is aided by the identifi-

Disorder

Marker

Insulin-dependent diabetes mellitus

HLA haplotypes, e.g, B8, B15, DW3, DW4

Coeliac disease Ankylosing spondylitis Belter's disease Anterior uveitis Peptic ulceration Alzheimer disease

HLA-A1-B8-DW3

3

HLA-B27

Blood group 0 APoE

cation of genetic markers, which may indicate that a major contribution to the disorder is being made by a gene on a specific chromosome. These genetic markers are particular blood groups, tissue types or restriction fragment length polymorphisms (RFLPs) at specific gene loci which are found to occur more commonly in individuals with the disorder (Table 3.10). A good example of the complex interaction between environmental factors and genetic loci is provided by the various possible causes of atheroma. A proportion of patients with early-onset atherosclerosis show abnormalities of lipoproteins or lipoprotein receptors which are determined by various gene loci. In some patients a single gene defect (e.g. of the cell membrane low-density lipoprotein receptors) can strongly predispose to atheroma. In the majority of cases, however, factors such as diet, smoking and high blood pressure may play the predominant role, with minor abnormalities of the lipoproteins merely providing an extra predisposition. The genes that predispose to common disorders are now being sought using genetic linkage techniques.

Linkage disequilibrium and pleiotropy There are two possible explanations for the association of a genetic marker and a specific disease: linkage disequilibrium and pleiotropy. Linkage disequilibrium If a genetic marker (e.g. an SNP, p. 60) is extremely closely linked to a disease locus, specific marker alleles may not be separated from the disease allele by recombination over the generations and an association will be observed in random individuals. This is known as linkage disequilibrium. It should be noted that linkage disequilibrium is a special case, observed only with very close linkage. In most cases of linkage, recombination over many generations will mean that a particular marker allele will not necessarily be associated with the disease in individuals picked at random. Pleiotropy The allele at the marker locus itself may predispose an individual to develop the disease, owing to some function

73

RECENT ADVANCES IN GENE THERAPY Gene therapy, defined as the replacement of a defective gene product, has been carried out for many years, for example in the treatment of haemophilia A and insulindependent diabetes (see Table 3.12). However, recent attempts at gene therapy have involved replacing a defective gene itself, so that it functions in situ to produce the appropriate protein. This approach involves targeting normal genes to a specific tissue so that they are integrated into the genome and function under appropriate regulatory control. At present the accessibility of the affected organ or organs limits the feasibility of this form of therapy to a few diseases only. Genetic disorders affecting the bone marrow are obvious candidates because marrow can be removed and cultured and manipulated in the laboratory before being retransplanted; immunodeficiency diseases and haemoglobinopathies are therefore among the first disorders where attempts at gene therapy have been made. Other organ systems where gene therapy is being tried include the liver and the lungs. Gene therapy for neurological disorders and muscular dystrophies will be more difficult, but is being explored.

of the gene. Some of the disorders associated with specific HLA types may be examples of a pleiotropic effect. For example, a specific HLA type may predispose to infection with specific viruses, leading to some forms of insulindependent diabetes mellitus.

EEIS2 =E :ZZL IE :Z :y 32 SS22 =23 SCREENING

The treatment of many genetic disorders is often most successful if the disorder is diagnosed at an early, presymptomatic stage. In addition, the prevention of genetic disease by genetic counselling and prenatal diagnosis requires the identification of asymptomatic gene carriers. For these reasons, genetic screening programmes have been developed. Depending on the nature and frequency of the condition, screening can involve the entire population, subgroups of the population (for example particular ethnic groups), or specific families in which a genetic disorder is known to be segregating (Table 3.11).

O MCQ 3.5

74

2

Case 3.3

In order to introduce normal genes into the appropriate cells a vector must be used. Vector systems that are being tried include inactivated retroviruses, adenoviruses and liposomes (lipid vesicles containing the appropriate DNA that fuse with the host cell membrane). The Table below summarizes some of the disorders where gene therapy is being attempted. Diseases where gene therapy has been attempted or considered Genetic disorder

Target organ

Adenosine deaminase deficiency (immunodeficiency) Thalassaemia/sickle cell disease Hypercholesterolaemia Phenylketonuria Haemophilia A Cystic fibrosis Duchenne muscular dystrophy

Bone marrow Bone marrow Liver Liver Liver Lung Muscle

TREATMENT OF GENETIC DISORDERS It is still technically difficult to replace missing or defective genes; although it may be possible to isolate and replicate a gene in the laboratory, it is much more difficult to ensure that a replacement gene reaches the correct tissue, and is properly regulated. Efforts are now being made to introduce normal genes into cells to correct genetic defects (see Recent Advances box). There are also a number of treatment strategies for genetic disorders which aim to ameliorate the phenotypic effects. Many common malformations of multifactorial or chromosomal origin can be corrected surgically, and a variety of treatments have been devised for single gene disorders (Table 3.12).

GENETIC COUNSELLING Most doctors will be faced at some stage with questions about the risk of a genetic disorder in patients or their families. Straightforward situations, such as Down syndrome or neural tube defect, can be dealt with by the nonspecialist but complex problems, involving the diagnosis of rare disorders or the use of carrier detection tests and linkage studies, require a specialist opinion.O Whatever the problem, there are a number of essential steps in genetic counselling: 1. Detailed family history. A pedigree of at least three gen-

TABLE 3,11 Screening for genetic disorders Disease

Group screened

Method

Purpose

Phenylketonuria

Total population of neonates

Guthrie test

Early dietary treatment of affected homozygotes

Hypothyroidism

Total population of neonates

TSH levels from dried blood spot

Early thyroxine therapy

Cystic fibrosis

Individuals of reproductive age

Direct DMA analysis of gene mutations

Prenatal diagnosis where both parents are carriers

Sickle cell anaemia and other haemoglobinopathies

Offspring of heterozygote mothers (total population of neonates in some regions)

Haemoglobin electrophoresis

Prophylaxis and early treatment of complications Prevention by counselling and prenatal diagnosis

Tay-Sachs disease

Individuals of Ashkenazi Jewish extraction

Carrier detection by serum enzyme analysis

Prenatal diagnosis where both parents are heterozygous

Adult-onset polycystic renal disease

Relatives of affected individuals

Renal ultrasound, blood pressure, gene tracking in families using DNA probes

Early treatment of hypertension and renal failure

Multiple endocrine adenomatosis

Relatives of affected individuals

Calcitonin, urinary VMA, gene tracking in families using DNA probes

Early treatment of endocrine tumours

Familial polyposis coli

Relatives of affected individuals

Sigmoidoscopy, gene tracking in families using DNA probes, ophthalmoscopy for associated retinal lesions

Identification of gene carriers, early treatment of malignancy

Familial breast cancer

Females in families with multiple affected members

Mutation analysis of BRCA1 and BRCA2 genes

Identification of gene carriers. Prophylactic mastectomy or medical treatment

erations should be taken and details of racial background and consanguinity (i.e. whether parents have common ancestors) obtained. 2. Accurate diagnosis in affected family members. An accurate diagnosis in affected family members must be made. This should include both a pathological diagnosis (established where possible from original notes and investigations) and an assessment of any genetic subgroup of the disorder. 3. Ancillary tests. Chromosome analysis, DNA marker studies or other carrier detection tests may need to be carried out, according to the nature of the problem. 4. Risk assessment. For single gene disorders risk assessment is sometimes straightforward, providing a correct diagnosis is available. For instance, the recurrence of risk for a couple who have had one child with phenylketonuria is 1 in 4, because phenylketonuria is autosomal recessive. Difficulties arise where a disorder has different inheritance types (e.g. retinitis pigmentosa) and with X-linked disorders, where carrier detection tests may need to be applied, new mutations have to be considered and risks must be modified to allow for normal sons in the pedigree. In addition, disorders with a late age of onset (such as Huntington's chorea) or of variable expression (such as tuberous sclerosis) can present problems that may require a specialist opinion. Where linked DNA probes are used, risk assessment can become complex. If a condition is caused by a chromosomal abnormality or is multi-

factorial, empirical risks derived from follow-up studies must be used. 5. Prenatal diagnosis. The availability of prenatal diagnostic tests (Table 3.13) has to be assessed and any possible risks to the mother or fetus weighed against the risk of fetal abnormality.© 6. Communicating risks. Having marshalled all the facts, sufficient time must be set aside to discuss the risks in detail with the patient. A non-directive approach should be taken, whereby objective risks are given and discussed in detail, in the light of the severity of the condition and the availability and reliability of prenatal diagnosis. These factors may be crucial in the final decision of the parents. It is best to give the risks as a proportion (i.e. 1 in 2, 1 in 100 etc.) and to emphasize that a 1 in 20 risk of a child being affected means that 19 times out of 20 the child will not be affected. In order to put the risks into perspective, it is also worth pointing out that over 1 in 50 children are born with significant abnormalities at birth. In comparison to these background risks, a genetic risk of 1 in 20 is generally regarded as low, and those of greater than 1 in 10 as relatively high. However, the perception of an 'unacceptable' risk will depend on a large number of factors. The severity of the condition and the availability of treatment will obviously affect the willingness of a couple to have further children. For example, a 1 in 20 risk of a child with isolated cleft lip and palate will often be seen as acceptable, whereas the

3

75

TABLE 3.12 Therapy for single gene disorders*

TABLE 3.13 Techniques of prenatal diagnosis

Treatment

Procedure

Timing

Types of disorder detected

Villus biopsy

Transcervical 9-11 weeks Transabdominal 12-1 6 weeks

Cytogenetic abnormalities Enzyme defects Single gene defects (DNA probes)

Maternal ct-fetoprotein

16-1 9 weeks

a-fetoprotein Raised: neural tube defects Lowered: Down's syndrome Lowered: Edwards' syndrome (N.B. These are screening tests; diagnosis must be confirmed by further investigation)

Amniocentesis

16-20 weeks

Cytogenetic abnormalities Enzyme defects Neural tube defects Single gene defects (DNA probes)

Ultrasound

From 16 weeks

Anatomical defects Neural tube/hydrocephalus Body wall defects Limb reduction defects Hydrops Major urinary anomaly Bone dysplasias Cardiac defects Craniofacial anomalies

Fetoscopy/ cordocentesis

16-20 weeks

Fetal blood sample Clotting disorders Haemoglobinopathies Cytogenetic abnormalities Fetal skin biopsy Epidermolysis bullosa Ichthyosis Albinism Fetal liver biopsy Urea cycle enzyme defects Fetal visualization Craniofacial defects Digital defects Genital defects

Replacement therapy Replacement of deficient protein Replacement of deficient vitamin or coenzyme: Vitamin B12 Vitamin D Biotin Replacement of deficient metabolite: Cortisone Thyroxine Organ or cell replacement Kidney Bone marrow Blood transfusion Amelioration therapy Restriction of substrates in diet: Phenylalanine Methionine Galactose Drug therapy to remove harmful metabolites: Penicillamine Cholestyramine Iron chelating agents Avoidance of harmful drugs or dietary factors Preventive therapy Removal of potentially malignant tissues Removal of site of destruction of diseased cells (e.g. spleen)

Disorder

Haemophilias Diabetes

Methylmalonicacidaemia (some types) Vitamin D-resistant rickets Biotinidase deficiency Adrenogenital syndrome Congenital hypothyroidism Polycystic renal disease Immunodeficiencies Osteoporosis Thalassaemias

Phenylketonuria Homocystinuria Galactosaemia Wilson's disease Hypercholesterolaemia Thalassaemias G6PD deficiency Porphyria

Polyposis coli Spherocytosis

* After Emery & Malcolm (1995)

same risk for a child with severe, untreatable mental retardation may seem too high. 7. Sympathy and understanding. If a couple have given birth to an abnormal child, the genetic counselling clinic is often the place where considerable feelings of guilt are expressed. Alternatively, there may be an understandable tendency to blame poor medical care or environmental factors for the abnormalities. Time should be set aside to discuss all these feelings in a sympathetic and understanding manner, as they are part of the normal reaction to the birth of an abnormal child. Guilt and blame are usually the result of lack of information. Once a couple have been fully informed of the nature and causes of a child's problems, they can be helped to accept the fact that many genetic abnormalities are unavoidable.

76

Successful genetic counselling requires a combination of

communication skills, detailed genetic knowledge, and a sound clinical background in both adult and paediatric medicine, if couples are to be helped to make the most appropriate decisions about childbearing. It is this combination that makes the practice of clinical genetics so rewarding.

FURTHER READING Emery A E H, Malcolm S 1995. An introduction to recombinant DNA in medicine, 2nd edn. John Wiley, Chichester. A good introduction to the techniques and uses of DNA analysis in medicine. Rimoin D L, Connor J M, Pyeritz R E 1996. Emery and Rimoin's principles and practice of medical genetics, 3rd edn. Churchill Livingstone, Edinburgh. A comprehensive multiauthor text covering most of clinical genetics.

Harper P S 1998. Practical genetic counselling, 5th edn. Butterworth-Heinemann, Oxford. A comprehensive introduction to the problems of clinical genetics.

The UK Medical Research Council Human Gene Mapping Project Resource Centre (MRC HGMP-RC).

Lander E S, Schork N J 1994. Genetic dissection of complex traits.

This Centre exists to provide specialist resources and

Science 265: 2037-2047. An excellent review of the techniques for finding the genes involved in multifactorial inheritance. McKusick V A 1998. Mendelian inheritance in man, 12th edn. The Johns Hopkins Press, London. A catalogue of all known single gene disorders in man. . Mueller R F, Young I D 1998. Emery's elements of medical genetics, 10th edn. Churchill Livingstone, Edinburgh. An excellent

services

introduction to medical genetics.

Worldwide Web computer addresses A vast amount of free information is available through the Worldwide Web (WWW). A few useful addresses are given below:

a

URL: .. http://www.hgmp.mrc.ac.uk/

OMIM (Online Mendelian Inheritance in Man) This database is a catalogue of human genes and genetic disorders authored and edited by Dr Victor A McKusick and colleagues at Johns Hopkins and elsewhere. This is an excellent place to look for up-to-date clinical and molecular information about genetic disorders. URL: http://www3.ncbi.nlm.nih.gov/omim/

77

This page intentionally left blank

THEIMMUNESYSTEM

4

4

INNATE IMMUNITY

Immunological Disorders

The components of the innate immune system provide a constant, unchanging system of protection for the healthy individual (Table 4.1). These physical and chemical barriers are an essential contribution to good health.

David Edgar and Herb Sewell Physical barriers and phagocytes

The immune system 79

Hypersensitivity 96

Immunodeficiency

Immune-suppression 102

85

INTRODUCTION This chapter describes the components of the immune system and their interactions in the defence against infection. It is only intended to provide the reader with sufficient basic scientific detail to understand the clinical disorders described. Abnormal or inappropriate immune responses cause the clinical conditions of immunodeficiency and allergy, and these are discussed with an emphasis on clinical presentation, investigation and management. The principles of autoimmunity, along with conventional and novel immunosuppressive therapies, are also discussed. Throughout, the chapter highlights how recent advances in understanding the immune response at a molecular level have had an impact on clinical practice. The human immune system has developed to combat infection by a diverse range of potential pathogenic microorganisms: bacteria, viruses, fungi, protozoa and helminths. It is a complex but effective system comprised of multiple components which have both specific and overlapping functions. The immune system is conventionally subclassified into the innate and adaptive components, the essential functions of which are: • The differentiation of self from non-self; • The recognition of foreign proteins by specific molecules (antibodies, T-cell receptors); • The recall of previous exposure to specific organisms and the production of a rapid, more effective response upon re-exposure. The last two are characteristic of adaptive immunity.

The intact skin and mucous membranes constitute a barrier to the entry of microorganisms. These agents must therefore rely on physical damage, such as injuries or insect bites, to gain access to the body. Numerous natural antimicrobial substances also exist that fatally damage microorganisms. These include unsaturated fatty acids that are secreted on to the skin, lysozyme that is secreted in tears, and others, including gastric acid. Interferons are a further family of molecules (oc,P,y) that block viral replication and promote antiviral immune responses. However, some microorganisms have developed specific attachment and penetration mechanisms to bypass these initial defences. An example is rhinoviruses, which penetrate the respiratory mucosa by first attaching to a cell surface adhesion molecule ICAM-1 (intercellular adhesion molecule-1). If microorganisms are successful in breaching the physical defences, they are usually phagocytosed and killed by polymorphonuclear leukocytes (PMN) or macrophages. Some bacteria (e.g. pneumococci) can resist phagocytosis by virtue of their capsules, whereas others (e.g. staphylococci) release toxins that destroy phagocytes. Whereas some microorganisms can survive within phagocytic cells (e.g. mycobacteria), giving rise to chronic infection, most are killed in the phagolysosomes by a combination of acid pH, oxygen metabolites (e.g. hydrogen peroxide) and various cytotoxic proteins. However, not all killing is intracellular, and macrophages or PMN can secrete toxic molecules in order to kill large targets, e.g. worms (helminths). In the case of viruses and tumours, natural killer (NK) cells also play an essential role.

TABLE 4.1 Innate and adaptive immune systems Innate Physical protection (skin, mucous membranes) Chemical protection (lysozyme, gastric acid) Interferons Phagocytes Complement

Adaptive Antibody Lymphocytes

79

complement and phagocytes are recognized as primary immunodeficiency diseases (PID). In addition, individual elements of the innate system are essential in stimulating the adaptive response during antigen presentation. Although classification into innate and adaptive helps our understanding of the immune response, it is important to remember that these elements are interdependent and function as an integrated system.

ADAPTIVE IMMUNITY

FIG. 4.1 Complement activation pathways The central event in complement activation is cleavage of C3 with the subsequent formation of C5 convertase. The generation of cleavage products (indicated by blue lines) augments the inflammatory response by attracting and activating phagocytes and causing activation and degranulation of mast cells. The deposition of C3b on cellular surfaces to promote phagocytosis is known as opsonization. The terminal pathway components (C5-C9) create the membrane attack complex (MAC) which causes osmotic cell lysis.

Complement system A further line of innate defence is the complement system (Fig. 4.1), a cascade of serum proteins which, when activated in succession, mediate three important effects: • Release of small peptides that stimulate inflammation and attract phagocytes; • Deposition of the cleavage component (C3b) on microbial membranes, which promotes phagocytosis via C3b receptors on phagocytic cells; • Assembly of complement components (C5-C9) into lytic complexes which puncture cell membranes, causing osmotic death. This activation 'cascade' is triggered by numerous microbial surfaces, notably the endotoxins of Gram-negative bacteria, by C-reactive protein (CRP) and mannan-binding lectin (MBL). Through interactions with antibody, phagocytes and mast cells, the complement cascade has an important role in integrating the response to infection.

Summary

80

The innate immune system is essential in the defence against infection, and deficiency can result in serious clinical consequences. The vulnerability to infection of patients with severe burns emphasizes the protective value of normal skin. As will be described later, specific defects of

The adaptive immune system differs in two important respects from the innate, with its characteristics of specificity and memory. The adaptive system responds to the environment and 'remembers' specific encounters with foreign antigens. The most important cells involved in the adaptive response are lymphocytes (B and T cells). The antigen recognition molecules on the surfaces of B and T cells (antibody and T-cell receptors) confer the specificity of the adaptive system. These molecules recognize and bind to specific antigens through amino acid sequences that are complementary in structure to the target antigen. Immunological memory is generated after the first antigen exposure (primary response) because antigen-specific lymphocytes proliferate and generate an increased pool of cells. These memory cells, responsive to the specific antigen, remain relatively inactive until the same antigen is encountered again (secondary response). Subsequent stimulation of memory cells results in a secondary immune response that is more rapid and quantitatively greater than the primary response (Fig. 4.2).

Lymphocytes Lymphocytes originate from stem cells in the bone marrow. B lymphocytes complete their maturation in the bone marrow, but T lymphocytes require passage through the thymus, where they complete their maturation and education process. Crucially, it is in the thymus that T cells learn to identify 'self proteins', and any autoaggressive T cells are either allowed to die (by apoptosis) or are rendered non-reactive (tolerant). The bone marrow and thymus are referred to as primary lymphoid organs. The secondary lymphoid organs include lymph nodes, the spleen, Peyer's patches, tonsils and adenoids. These are strategically located to encounter microorganisms entering the body through external tissues, blood, gut, and upper airway, respectively. The structure of secondary lymphoid tissue is exemplified by the lymph node, where the cortex contains mainly B lymphocytes and the paracortex T lymphocytes, whereas antibody-secreting plasma cells are found in the medulla. Peripheral blood normally contains 70-90% T cells, 5-10% B cells and approximately 1-10% natural killer (NK) cells. A characteristic of lymphocytes is their ability to 'traffic' from peripheral blood into the tissues and secondary lymphoid organs. This property allows lym-

4

FIG. 4.2 Primary and secondary antibody responses Primary and secondary antibody responses to antigen. This demonstrates the characteristics of the adaptive immune response, i.e. specificity and memory. Upon re-exposure to antigen the immune system recognizes the antigen due to immunological memory. The resulting response has a shorter lag time (lag 2) and a quantitively greater antibody response.

phocytes to encounter antigen in almost all physical locations. The lymphoid tissues of the gut, genitourinary tract, breast and lungs behave, to a certain extent, autonomously, in that particular B cells (largely IgAsecreting) and T cells recirculate selectively through them. This recirculation is mediated by the expression of specific homing receptors and is referred to as the mucosal immune system. The result is that primary antigen exposure at one mucosal site will facilitate protective immunity at other mucosal sites. B cells

The main function of B cells is antibody production. Some B cells are stimulated directly by polymeric antigens and these responses are termed T-independent (e.g. bacterial capsular polysaccharides). Most B-cell responses are however T-cell dependent. For T-dependent responses, B cells are triggered to proliferate by interaction between their surface Ig and the antigen, but they require additional signals from T-helper cells in order to differentiate into plasma cells and secrete antibody. These signals consist of non-antigen specific cytokines (particularly IL-4 and -6) and cell-cell-mediated signalling. The interaction of these cell surface molecules, e.g. gp39 (CD40 ligand, CD40L) on T cells with CD40 on B cells, is essential in facilitating effective B-cell antibody production. During T-dependent responses some proliferating B lymphocytes are retained in germinal centres instead of differentiating 'terminally' into plasma cells. These are the memory B cells. Antibodies Antibodies are a family of five major classes of molecule which have a primary role in combating microorganisms free in the blood or tissue spaces, primarily viruses in their

FIG. 4.3 Immunological molecules V = variable, D = diversity, J = joining, C = constant, HC = heavy chain, LC = light chain, VH/L = variable region of heavy/light chain, CH1(2,3) = constant region 1 (2,3) of heavy chain, CL = constant region of light chain.

extracellular stage, and many bacteria. The antibody or immunoglobulin (Ig) molecule is ideally suited to its function of promoting the removal of foreign antigens. Its basic structure is seen in antibodies of the IgG class (Fig. 4.3). The N-terminal half is composed of heavy (H) and light (L) chains and is responsible for binding to antigen (Fab portions). The C-terminal half (Fc), which consists of portions of heavy chains, determines the class of antibody and is responsible for the other biological functions of the molecule. The major role of antibody molecules is to bind to microorganisms via the Fab portion, enhancing phagocytosis, as phagocytes possess Fc receptors. This process of coating microorganisms with antibody is known as opsonization. The basic four-chain structure is preserved in all five classes of immunoglobulin, but each class has certain characteristics that subserve its specific functions. • IgG molecules are involved in the activation of complement, attachment to phagocytic cells, passage into the tissues and transport across the placenta. The four IgG subclasses (IgG1, IgG2, IgG3 and IgG4) vary somewhat in these functions. • IgM is a large molecule with a pentameric structure. It is mainly confined to the bloodstream and is very efficient at activating complement and agglutinating foreign material because of its 10 antigen-combining sites per pentameric complex. • IgA is the predominant immunoglobulin class secreted at mucosal sites. It has a dimeric structure and is protected from proteolytic damage by a special polypeptide, the secretory chain (Sc). Sc is produced by the mucosal epithelial cells and acts as a receptor-transporter, binding to and taking the antibody into external sites. IgA can

81

TABLE 4.2 Immunoglobulin structure and function Class & subclass

Structure

Major functions*

Monomer Monomer Monomer Monomer

C&O

Pentamer

c

Dimer

Mucosal protection Mast cell activation Antigen binding

IgG

IgG, IgG2 IgG3 IgG4 IgM IgA

IgE IgD

Monomer Cell surface bound

c C&O 0

*C, complement activation; 0, opsonization.

therefore survive proteolytic attack in secretions of the eyes, lungs, nose, gut and urinary tract. • IgE attaches via its Fc region to receptors on mast cells. When two bound IgE molecules are cross-linked by antigen, intracellular signalling results in the release of mediators from mast cell granules which cause the clinical features of acute allergy. • IgD is found mainly on B-cell membranes and acts as an antigen receptor. Its soluble form is not thought to have a significant physiological role. The properties of the various Ig classes and subclasses are summarized in Table 4.2. Antibody genes Genes on three different chromosomes code for antibody molecules. The genes of heavy and light chains are not continuous in the germline, but consist of separate clusters of gene segments separated by non-informative regions called introns. The clusters are named according to the part of the immunoglobulin molecule that they encode (Fig. 4.3), i.e. V, variable; D, diversity (heavy chains only); J, joining; and C, constant. During the development of a B cell, the germline DNA encoding the heavy-chain genes undergoes somatic rearrangement and recombination such that a complete heavychain variable region gene, V-D-J, is assembled. After transcription, RNA is spliced to link the V-D-J segment with a C segment to encode a complete heavy chain (V-D-J-C). Once this is accomplished, the light-chain genes rearrange to encode a complete light chain (V-J-C). Heavy and light chains then link together to form a complete antibody molecule. If rearrangement is not successful, the developing B cell is aborted. Enzymes (recombinases) encoded by two genes termed RAG1 and

1 82

MCQ4.1

RAG2 largely control the recombination events in lymphocyte progenitors. Defects or mutations associated with these genes can profoundly and adversely affect the development of the immune system. After gene rearrangement, the B cell does not restore the germline configuration and daughter cells will retain this unique rearrangement, ensuring continued antigenic specificity of that clone and its entire subsequent offspring (progeny). This is the molecular basis of immunological memory, thus maintaining immunological specificity in the clonal progeny. The heavy-chain C gene can, however, change to allow class switching (e.g. IgM-IgG) during the immune response, as occurs in moving from a primary to a secondary response. During the life of a B cell, point mutations may occur within the variable regions (somatic mutation), resulting in a change of affinity of the antibody molecule. Increased affinity for antigen confers a selective advantage to any B cell. This phenomenon results in affinity maturation whereby, as an immune response progresses, antibodybinding affinity increases. The huge potential diversity (i.e. the ability of humans to respond to a multitude of different foreign antigens) of antibody binding specificity (up to 108 specificities) is thus achieved by a number of mechanisms: • Association of different heavy chains with either K or X light chains; • Association of V genes with different (D) and J genes; • Genetic joining errors; • Somatic mutation. Genetic differences between antibody molecules are known as isotypic if they occur in the constant region and as idiotypic if they reside in the hypervariable regions (the sites associated with direct binding to antigen; Fig. 4.3). The idiotype of an antibody is the portion defining the uniqueness of that variable region, and may include both the antigen-combining site and the adjacent regions. There are also allelic differences known as allotypes, mainly within the C regions, which enable the antibody molecules of different individuals to be distinguished.1

T ce//s The major effector role of T cells is the elimination of viruses, fungi and protozoa, but they also act as coordinators of the overall adaptive immune response. B and T lymphocytes are impossible to distinguish by light microscopy and therefore more sophisticated techniques are needed (e.g. flow cytometry) which can identify specific cell typespecific surface antigens (CD antigens). All T cells bear the CD3 antigen, which is associated on the cell surface with the T-cell receptor (TCR). Cytotoxic T cells (Tctx) are responsible for killing abnormal host cells (e.g. virally infected, malignantly transformed or transplanted cells) and, in addition to possessing CD3, are also positive for the surface molecule CDS. They destroy target cells by several mechanisms, including cell membrane interactions (such as that of the Fas antigen

4

FIG. 4.4 Interaction of T cell receptor and HLA molecules The T cell is 'restricted' in that CD8+ T cells only recognize antigen when it is presented in association with HLA class I molecules. CD4+ cells only recognize antigen in association with HLA class II molecules.

(CD95) with its ligand) and the release of cytotoxic substances: perforins, granzymes and other cytokines. Helper T cells (Th), in addition to expressing CDS, also express CD4. It is helpful to remember some of these more common CD numbers, as these are important in describing primary immune deficiency states and lymphoproliferative diseases.

FIG. 4.5 Antigen processing and presentation 2 = 2 -microglobulin, TAP = transporter of antigenic peptides, class 1, WV = peptide, =C = HLA class II.

= HLA

TABLE 4.3 Major HLA associations The T-cell receptor (TcR, Fig. 4.4) The T-cell receptor is a two-chain (a and , or y and 6) molecule, rather like a shortened immunoglobulin. However, it has a more complicated combining site than antibody, and recognizes linear fragments of small foreign peptides in combination with molecules of the human leukocyte antigen system (HLA). The TcR-associated CD4 or CD8 molecules determine whether the individual T cell recognizes antigen in association with class I or class II molecules. CD8-expressing T cells only recognize antigen when it is presented with HLA class I molecules, whereas CD4-expressing T cells are only stimulated by antigen in association with HLA class II molecules. The human leukocyte antigen (HLA) systems The HLA system is the major determinant of transplant rejection responses. However, its normal physiological function is mainly to control cell-cell interactions, particularly within an individual's own immune response. The HLA genes lie on the short arm of chromosome 6 and are grouped into two major sets, referred to as class I and class II. Each set contains three genetic loci that have a large number of alleles coding for HLA antigens. The genes for some complement components and cytokines also lie within the same region (the class III region), but their products are quite unrelated to class I and class II molecules. Class I molecules consist of a large oc chain, and a small P chain known as pVmicroglobulin. 2-Microglobulin is encoded on a different chromosome and shows no sequence variability. Class I antigens (i.e. alleles at the A, B and C loci) are expressed on all nucleated cells (two from

Disease

HLA-B

Ankylosing spondylitis Rheumatoid arthritis SLE Celiac disease IDDM* Hashimoto's disease

27

HLA-DR

4 3 8 8

3 3,4 5

*IDDM, insulin-dependent diabetes mellitus.

each locus on diploid cells). Class II HLA molecules are two-chain ( and ) structures both chains are encoded in the HLA locus and are polymorphic. Class II antigens (alleles at the DP, DQ and DR loci) are expressed on a more limited range of cells: classic antigen-presenting cells, B lymphocytes and activated T lymphocytes. Class I and II antigens are central to antigen presentation and the generation of an effective immune response. Knowledge of the HLA alleles expressed by an individual is clearly essential for most organ transplantation. In addition, particular HLA types are also known to be associated with certain diseases (Table 4.3). In some circumstances the HLA association is so strong as to be helpful in diagnosis, e.g. HLA B27 and ankylosing spondylitis. Antigen presentation (Fig. 4.5) Antigen-presenting cells (APC) include macrophages, dendritic cells of the spleen and lymph nodes, Langerhans' cells of the skin, and B cells. Antigenic fragments are pre-

83

FIG. 4.6 CD4 + cell activation after antigen presentation The binding of cell surface receptors to their cell-bound or soluble ligands results in a series of positive (+) or negative (-) signals transmitted to the CD4 T cell nucleus. Secretion of IL-2 activates both the secreting T cell and its near neighbours. sented to T cells in specialized 'antigen-binding grooves' in HLA class I and class II molecules. Class I molecules present peptides which are endogenously synthesized, e.g. viral proteins in an infected cell, whereas class II molecules present peptides from external proteins (e.g. extracellular microorganisms) which have been engulfed and degraded to peptides by the APC. There are distinct intracellular pathways involved for class I and class II presentation. Class II molecules are synthesized within the rough endoplasmic reticulum (RER) and, bound to a second molecule called the invariant chain, are transported to the endosomal compartment. There the invariant chain is replaced by antigenic fragment and the complex is transported to the cell surface. Class I molecules are also synthesized in the RER and are initially anchored on the inner surface of this organelle. Peptide fragments generated in cytoplasmic proteasomes are transported into the RER by specialized transporter proteins TAP-1 and TAP-2 (TAP = transporter of antigenic peptides). The peptide fragments then bind to the class I molecule and this complex associates with pVmicroglobulin before being transported to the cell surface (Fig. 4.5). * In addition to TcR recognition of the HLA/antigenic fragment, the CD4 molecule binds to a non-polymorphic region of the HLA class II molecule. Several other adhesion molecule pairs (LFA1/ICAM-l, CD2/LFA-3) link together, strengthening the cellular adhesion. The interaction of a T-cell surface molecule CD28 with B7 molecules on APCs is believed to be crucial to T-cell activation, as is the secretion by the APC of IL-1.

1 MCQ 4.2

84

Th-cell activation (Fig. 4.6) CD4+ T-cell activation is dependent on a series of 'costimulatory' signals consisting of cell-cell surface binding and soluble cytokines.* The result of these 'positive signals' is the generation of gene transcription factors (such as NFAT - nuclear factor of activated T cells) which 'switch on' genes (e.g. IL-2 gene) within the T-cell nucleus. Once IL-2 is secreted, it acts on both the responding and neighbouring cells. In contrast to CD28, a molecule termed CTLA4 appears to be crucial in donwegulating the T-cell immune response. CTLA4 appears on T cells late in their activation and also binds to B7 molecules (see above), but this binding delivers a negative signal to 'switch off responding T cells. Constructs of CTLA4 have recently been used in clinical trials (phase I and II) to downregulate unwanted T-cell responses in transplantation rejection and in autoimmune disorders. The increase in understanding of the key signalling molecules involved in antigen presentation will undoubtedly result in further attempts at therapeutic manipulation. Regulation of immune responses Both T and B cells are activated by recognition of specific antigens. Effective removal of antigen is therefore an important physiological endpoint for an immune response. However, the immune system has several additional mechanisms that regulate the nature and intensity of the immune response: • Apoptosis (programmed cell death) is now recognized as an important regulatory mechanism for the disposal of immune cells which have completed their function. • Anti-idiotype antibodies are directed against the V region of the Fab portion (idiotype) of other antibody

4

TABLE 4.4 The major cytokines aiid their functions Cytokine

Natural source

Functions

Potential clinical uses

TNF-a and p

Macrophages

Antitumour, antiviral, vascular damage, PMN activation, bone resorption, induces IL-1 etc.

Anti-TNF moAb used for rheumatoid arthritis, inflammatory bowel disease

Interferon-

Macrophages

Antiviral

lnterferonInterferon-y

Fibroblasts T eel Is

Antiviral Antiviral, activates macrophages, increases mycobactericidal activity

IL-2 IL-5 IL-6

T cells T cells, mast cells T cells, fibroblasts, ARC, hepatocytes T cells, monocytes Macrophages

Activates T, B and NK cells Eosinophil growth and activation B-cell growth and differentiation; induces acute-phase response Neutrophil activation and chemotaxis Granulocyte growth and maturation

Hepatitis B&C, papillomatosis, Kaposi's sarcoma, hairy cell leukemia, chronic myeloid leukaemia Multiple sclerosis Multidrug-resistant TB, lepromatous leprosy, visceral leishmaniasis, IFNyR deficiency, chronic granulomatous disease, hyper-lgE syndrome, chronic mucocutaneous candidiasis Melanoma, renal cell carcinoma Anti IL-5 moAb in phase II clinical trials for asthma

IL-8 G-CSF

IL-8 receptor antagonist in preclinical trials for airway disease Cyclical neutropenia, chemotherapy 'rescue'

After activation, however, there is a preferential commitment towards either a Thl or a Th2 pattern (Fig. 4.7). Thl cytokines are IL-2, IL-12 and IFN-y, favouring a cellular immune response, whereas Th2 cytokines, favouring an antibody response, are IL-4, IL-5, IL-6 and IL-10. There is mutual suppression between these two dominant patterns of cytokine secretion such that, once a response has become committed in either direction, this pattern will tend to be maintained. It is therefore currently believed that the 'suppression' of immune response seen in certain situations is the net result of a balance in the type of cytokines secreted in the microenvironment, rather than the function of a single cell type. This is important in understanding the influence of the immune response on the pattern of clinical diseases such as leprosy (see hypersensitivity section).1

FIG. 4.7 T cell differentiation The blue lines indicate secretion of soluble mediators, the black lines indicate suppressive activity. For explanation see text.

molecules and may block or modulate antibody function. • A specific T-suppressor cell subset has not been identified in humans. There is, however, increasing evidence, in both animals and humans, that the pattern of cytokines secreted by CD4+ T cells will preferentially induce either a predominantly cellular or an antibody response. The major cytokines, their actions and potential clinical importance are summarized in Table 4.4. Prior to activation, most CD4+ T cells are 'ThO' (i.e. they have the potential to secrete a wide range of cytokines).

IMMUNO DEFICIENCY

Immunodeficiency is classified asprnnory (where the cause is/was previously unknown) or secondary, where there is a recognized cause for the immune defect, such as cancer, infection, drug treatment, radiotherapy or malnutrition.

PRIMARY IMMUNODEFICIENCY Understanding of primary immunodeficiency has increased rapidly over recent years with the identification of specific molecular lesions in a number of these previously poorly understood conditions. Research into these conditions has increased our understanding of the cell biology of the immune response and led to new diagnostic strategies and experimental forms of gene therapy in humans.

85

Primary immunodeficiency diseases (PID) are subclassified according to the individual immunological components involved, hence there are antibody, T-cell or combined deficiencies, as well as phagocytic and complement component deficiencies. The defect in each condition leads to preferential susceptibility to certain types of organisms. Typically, antibody deficiency causes susceptibility to bacterial infection: T-cell deficiency to viral and fungal infections; and some complement deficiencies to meningococcal infections. These associations are important in consideration of the initial immunological differential diagnosis.

What are the characteristics of infection in PID? If the clinical history of infection can be described as serious, persistent, unusual or recurrent (SPUR), then immunodeficiency should be part of the differential diagnosis. 'Serious' suggests a potentially life-threatening infection (meningitis or septicaemia). 'Persistent' suggests that an infection is particularly resistant to conventional therapy. Infections may be 'unusual' in terms of site (liver/ brain abscess, or osteomyelitis) or organism (Pneumocystis, Aspergillus, Mycobacteria), and these should always raise the possibility of immune deficiency. Finally, 'recurrent' infection is particularly important. The child or adult who repeatedly presents with infection deserves further investigation of his or her immune status. It is difficult to be specific about how many infections constitute a 'recurrent' problem, and this inevitably involves an assessment of both frequency and severity of infection, evidence of organ damage, impact on systemic wellbeing and normal daily activity. The '10 warning signs of immune deficiency' are by no means exhaustive, but do give practical guidance on the clinical features that should prompt investigation for PID (Table 4.5)

When to suspect the diagnosis? Patients with PID may present with a history of infection or with a family history of an affected relative. More com-

TABLE 4.5 Ten warning signs of immune deficiency

86

1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

8 or more new ear infections/year 2 or more serious sinus infections/year 2 months antibiotics without effect 2 or more episodes of pneumonia within a year Failure to thrive Recurrent, deep abscesses Persistent thrush after 1 year of age Need for intravenous antibiotics 2 or more deep-seated infections (sepsis, meningitis, cellulitis) Family history of PID

monly, they may present with the long-term complications of their condition (e.g. bronchiectasis). They may also present with a combination of clinical features and laboratory abnormalities that may not initially alert the non-specialist to the underlying diagnosis (e.g. chronic ill health, autoimmune phenomena). It is very typical for patients with PID to suffer a prolonged delay before the correct diagnosis is made, despite attending many doctors. During this period of delay they suffer significant morbidity and even mortality, and it is therefore essential that there is increased awareness of these conditions to allow earlier diagnosis and treatment. The simple guidelines from the history (Table 4.5) are useful in determining the most appropriate first line of immunological investigation. It is important that investigations are used appropriately and that expert advice is available, both in the selection of tests and in the interpretation of results. The general principles guiding investigation for immune deficiency include determining both that the relevant components of the systems are present, and that they are functionally intact. It is particularly important in the investigation of children that all immunological results are compared to appropriate, age-related normal ranges. Increasingly, specific molecular tests are used to confirm diagnoses, and therefore subtler variant forms of immune deficiency are being recognized.

Primary antibody deficiency diseases X-linked agammaglobulinaemia (XLA/Bruton's type) Clinical features Affected boys usually present between 6 months and 2 years of age. There may be a family history of affected male relatives. The most common complaint is of recurrent bacterial infections (typically Streptococcus pneumoniae or Haemophilus influenzae), usually affecting the respiratory tract, which if untreated lead to bronchiectasis. Boys may present with ear or joint infection and some have a nonspecific arthropathy that responds to the introduction of immunoglobulin replacement therapy. Chronic diarrhoea and a malabsorption syndrome may occur as a result of Giardia lamblia infection. A helpful physical sign is the absence of tonsillar lymphoid tissue or lymphadenopathy. A particular long-term complication in XLA is mycoplasma or ureaplasma arthritis. This is distinct from the more frequent non-specific arthropathy and is difficult both to diagnose and to treat. XLA patients are also susceptible to enterovirus infection (polio, coxsackie or echoviruses). Oral polio vaccine is therefore contraindicated, as it may cause paralysis, and a chronic meningoencephalitis may result from coxsackie or echovirus infection. This latter complication has become rarer since the advent of intravenous immunoglobulin therapy, but it still occurs either as a generalized condition or in a more limited form, causing sensorineural deafness. The prevalence of XLA is thought to be approximately 1 in 100000 in the UK, but

with the advent of molecular testing a milder, previously undiagnosed, form is being identified. Investigation Serum IgG level is usually less than 2g/L, with serum IgA, IgM and IgE being undetectable. Measurement of IgG subclasses is not usually helpful. Isohaemagglutinins are typically absent, as are specific antibodies to previously administered vaccines. Mature B cells are not usually detectable in peripheral blood, but T-cell numbers and function are normal. Neutropenia may be a feature. In the atypical, less severe, cases immunoglobulin levels are more variable and low numbers of circulating B cells are detected (typically in young teenage or adult males). In such cases the diagnosis can be difficult and test immunization responses to protein and polysaccharide antigens (usually tetanus toxoid and pneumovax II) are very helpful. Antigen-specific antibody levels are measured before and 4 weeks after immunization. Results require expert interpretation. The underlying molecular lesion has been characterized as a defect of the intracellular tyrosine kinase 'Bruton's tyrosine kinase' (Btk), coded for at X21.3-22. This kinase is essential for normal B-cell maturation, and B-cell development is therefore arrested at the pre-B stage. Family assessment to identify carriers is possible and should be undertaken, with appropriate genetic counselling. Rare autosomal recessive cases occur. Management of antibody deficiency Replacement therapy with regular intravenous immunoglobulin (ivlg) is now the standard treatment for primary antibody deficiency. This lifelong treatment is extremely effective in controlling infection, improving quality of life and preventing end-organ damage. The usual dose is 400mg/kg, given every 3 weeks. Regular preinfusion serum IgG levels are measured and should be maintained within the physiological range (usually >8g/L). Patients contribute to their own monitoring by completing diary sheets which record their general wellbeing, frequency of infections and antibiotic requirement. Patients must be weighed regularly and appropriate increments in ivlg dose made, because growth (in children) and weight (in adults) often increases significantly after the initial diagnosis and treatment. Various ivlg preparations differ in their con-tent. Once a patient is established on a particular product, it must not be changed without clear clinical indications. The main potential side-effect of treatment is reaction to the infusions. This may be related to infusion rate, intercurrent infection, or anti-IgA antibodies in IgA-deficient patients. Transmission of viral infection is now a very small risk because of careful viral screening and new antiviral steps in the production of ivlg from pooled donor sera. Recent advances include self-administration of ivlg at home and the development of subcutaneous administration (sclg), which is of particular value in the very young. Although immunoglobulin replacement is the mainstay of

treatment for antibody deficiency, there are other aspects of therapy. When respiratory infection occurs, patients should be treated over an extended period (usually 10 days) with high-dose antibiotics, guided by antibiotic sensitivity. Patients should be encouraged to undertake regular breathing exercises to assist with expectoration of sputum, and in those with established lung damage postural drainage should be undertaken daily. Regular assessment of pulmonary function is advised; however, 'routine' use of X-ray and CT examination (e.g. at annual review) is not advised in patients with PID. Many of these conditions predispose to a range of malignant diseases, and 'non-essential' exposure to radiation should therefore be avoided.

4

Common variable immune deficiency (CVID)/ acquired hypogammaglobulinaemla Clinical features This condition usually presents in adults with established lung disease or recurrent pyogenic infections. It is very typical for there to be a diagnostic delay of many years, during which respiratory function has deteriorated. Patients with CVID are at increased risk of autoimmune phenomena, including thrombocytopenia, haemolytic anaemia, malabsorption syndromes and organ-specific autoimmune disease. They have a particular tendency to granuloma formation and may present with lymphadenopathy or hepatosplenomegaly. The finding of typical granulomata on imaging may be helpful in establishing the diagnosis. These are usually non-caseating and steroid responsive, but treatment is not necessary if lesions are asymptomatic. A further characteristic complication of CVID is nodular lymphoid hyperplasia (NLH), in which lymphoid tissue of the gut becomes hyperplastic and produces characteristic filling defects on small bowel radiology series. NLH is thought to be a prelymphomatous condition, but CVID patients without NLH also have an increased risk of malignancy, particularly of lymphoma. CVID prevalence is estimated as 2-4 :100000 in the UK, but is probably underdiagnosed. IgG subclass and specific antibody deficiencies are thought to be related conditions, and some will progress to more typical CVID. Investigation Total serum IgG level is usually less than 3g/L; the levels of IgA and IgM are more variable. The diagnosis can therefore be difficult, and test immunization as described in investigation of XLA is required. In CVID the responses are poor or absent. Measurement of IgG subclasses is not usually helpful. Peripheral blood lymphocyte subset analysis may indicate a reduction in the proportion of CD4+ T cells that also carry the CD45RA antigen. Lymphocyteproliferative responses may be reduced. Unlike XLA, a single genetic defect has not been described for CVID, but B cells from these patients do not develop normally into plasma cells. It is thought that there is an abnormality of T-cell regulation of B cells in CVID.

87

Management CVID patients require immunoglobulin replacement therapy as for XLA, with the possible addition of steroids for symptomatic granulomatous disease. CVID patients who are deficient in IgA should always be prescribed an ivlg preparation that is known to be low in IgA in order to reduce the risk of infusion reactions (see below). IgA deficiency Clinical features IgA deficiency is the most common primary immune deficiency in the UK, with a prevalence of approximately 1: 600. Most people with IgA challenge are identified by chance and are asymptomatic. The challenge is therefore to identify those in whom there is a significant immune deficiency. The history is crucial because IgA deficiency is known to be associated with an increased risk of pulmonary/sinus infections, allergy, celiac disease, other autoimmune diseases and malignancy. One must therefore look carefully for any suggestive clinical features in patients presenting with incidentally identified IgA deficiency. Furthermore, IgA deficiency can be a feature of some combined immune deficiencies (e.g. ataxia-telangectasia (page 93) and Wiskott-Aldrich syndrome (page 92)), and there may be clinical features suggestive of these diseases; IgA-deficient patients may form anti-IgA antibodies, which create a high risk of transfusion reactions to IgAcontaining blood products. If immunoglobulin replacement therapy is required, a product containing the lowest possible concentration of IgA must be used. Investigation IgA deficiency is defined by a laboratory measurement of serum IgA <0.05 g/L. The automated methods used in most routine laboratories are not sufficiently sensitive to detect these levels. Therefore, a routine report of 'low' level of serum IgA does not necessarily indicate a diagnosis of IgA deficiency. Once IgA deficiency is established, however, the patient must be fully evaluated to determine whether there is a clinically significant immune deficiency. A number of these patients will have associated IgG subclass deficiency and/or a failure of antibody responses to test immunization. If a combined immune deficiency syndrome is suspected, more extensive investigations are required (see ataxia-telangectasia (p. 93) and Wiskott-Aldrich syndrome (p. 93)).

88

Management It is only the minority of IgA-deficient patients who have a significant history of infection, and usually other laboratory evidence of immune deficiency (IgG subclass deficiency, failure of response to test immunization, combined immune deficiency syndromes), who require treatment. Options include prolonged high-dose antibiotics for infectious episodes, prophylactic antibiotics, and immunoglobulin replacement therapy. Very few patients without IgG subclass deficiency or failure of immunization responses require immunoglobulin replacement, which suggests that

isolated IgA deficiency, in the absence of such defects, is a relatively benign condition. IgG subclass deficiency Clinical features IgG subclass deficiency is regarded as a part of a spectrum of conditions, including CVID and IgA deficiency. In common with IgA deficiency, patients may be asymptomatic or give a typical history of recurrent bacterial infection. The simple identification of a slightly reduced level of an IgG subclass is not sufficient in itself to warrant a diagnosis of significant primary immune deficiency. IgG subclass deficiency is usually suspected when the clinical history is suggestive of antibody deficiency, but total IgG (and other immunoglobulin classes) are normal and other primary antibody deficiencies have been excluded. Full immunological assessment is essential before treatment decisions are made. Investigation IgG subclass development occurs slowly during early childhood, and it is therefore important to relate results to the correct age-related ranges. Total serum immunoglobulin levels are usually normal, or may be high if there is a compensatory rise in the level of the non-deficient subclass (usually IgG^. On investigation, T-cell numbers and function (measured by in vitro proliferation assays) are usually normal. A failure of specific antibody response to test immunization would support a diagnosis of clinically significant IgG subclass deficiency. Management Not all patients will require ivlg and a careful assessment of the clinical history and investigative results is required. As discussed under IgA deficiency, several therapeutic strategies may be introduced and monitored for evidence of clinical response. Specific antibody deficiency Clinical features These patients are particularly susceptible to pneumococcal infection, but their total serum IgG and IgG subclass levels are normal. The clinical features are otherwise typical of other antibody deficiency syndromes, except that the diagnostic delay is likely to be longer because of the subtlety of the defect. Investigation These patients fail to mount a specific antibody response to polysaccharide antigens (e.g. Pneumovax II) but respond normally to protein antigen (e.g. tetanus toxoid). By definition, they will have normal total serum immunoglobulin and IgG subclasses. Management As for IgG subclass deficiency, the appropriate therapeutic option for each individual patient is based on a combined assessment of the clinical severity of the defect and the investigative results. The typical clinical features

described will usually indicate a requirement for ivlg; however, equivocal responses to immunization do occur in some individuals who do not require specific therapy. Transient hypogammaglobulinaemia of infancy Clinical features As passively acquired maternal immunoglobulin disappears and the infant begins to produce its own immunoglobulin, there is normally a physiological low level of serum antibody levels between 3 and 6 months of life. In some infants this 'physiological nadir' is profound, prolonged, and associated with bacterial infections. There are no other specific features: this is a diagnosis of exclusion and the clinical priority is to ensure that the child does not have a primary immune deficiency, such as severe combined immune deficiency (SCID, p. 93), X-linked agammaglobulinaemia (XLA, p. 86) or very early-onset common variable immune deficiency (CVID, p. 87). Investigation Serum IgG and IgA levels will usually be below the agerelated normal ranges, but IgM levels may be normal and isohaemagglutinins should be detectable. Differentiation from primary antibody deficiency may be difficult at this stage, but it is essential that a primary immune deficiency be excluded. Close clinical and laboratory monitoring of such infants is required for evidence of emerging antibody production. A diagnosis of transient hypogammaglobulinaemia can only be made retrospectively when antibody levels have returned to normal and the baby is clinically well. Management Decisions regarding treatment depend on the severity of the clinical history. If there are only occasional minor infections, no treatment is indicated. If frequent infection is causing significant morbidity the range of therapeutic options for antibody deficiency should be considered. If immunoglobulin replacement is deemed necessary, despite diagnostic uncertainty, it is essential to observe the patient closely for signs of recovering endogenous antibody production. This may be detected by a rise in IgM or IgA levels in preinfusion testing (intravenous immunoglobulin (ivlg) products do not contain IgM or IgA). Planned withdrawal of ivlg is also possible to allow repeat immunological assessment, including test immunization. Withdrawal of ivlg must, however, be for a prolonged period before immunization, as the half-life of IgG in the circulation is approximately 3 weeks. Immune deficiency with associated hyper-IgM Aetiology This is a disorder of T-cell function which prevents IgMproducing B cells from differentiating into IgG-producing plasma cells. The molecular lesion is of the T-cell surface protein gp39 (CD40 ligand/CD40L) and the gene is located at Xq26-27. The interaction between B cells (expressing CD40) and T cells (expressing CD40L) is crucial in trig-

gering IgM-producing immunoblasts to differentiate into IgG-secreting plasma cells.

4

Clinical features As this is an X-linked condition (although autosomal recessive cases are reported), affected boys present with recurrent pyogenic infections related to their hypogammaglobulinaemia and a commonly associated neutropenia. A typical presentation is with Pneumocystis carinii pneumonia (PCP) in infancy, which is an unusual infection in antibody deficiency and the susceptibility is probably explained by an associated underlying T-cell defect. There is also increased susceptibility to cryptosporidial infection, with ascending cholangitis and chronic liver impairment. There is an increased incidence of malignancy, particularly lymphoma and a range of autoimmune diseases. Investigation Serum levels of IgG are usually less than 3g/L, but IgM levels may be up to lOg/L (normally <3g/L). Neutropenia is common, but T-cell numbers and response to mitogens are normal. The failure of CD40L expression on stimulated T-cells in vitro can help confirm the diagnosis, but mutated forms of the CD40L gene may not be identified without more sophisticated genetic analysis. Management Commencement of ivlg therapy usually results in normalization of both serum IgM and neutrophil counts. The particular susceptibility of these patients to PCP is an indication for primary/secondary prophylaxis with oral cotrimoxazole. Allogeneic bone marrow transplantation is the definitive treatment in childhood. For those (usually older) patients with established chronic liver disease (probably caused by chronic cryptosporidial infection), combined liver and bone marrow transplantation is now considered. Experience to date is too limited to give definitive statements regarding prognosis.

Complement deficiencies The complement system is represented in Figure 4.1. The clinical features of deficiency depend on which part of the activation cascade is interrupted. The major points will be described. Individual complement component deficiency Clinical features Individual complement component deficiencies are rare, and in most cases are inherited as autosomal recessive conditions. Classic pathway component deficiency (Cl, C2 and C4) may present with recurrent bacterial infection or with immune complex-mediated disorders ('lupus-like syndromes'). C2-deficient patients are especially prone to pneumococcal infection. C3 deficiency is very rare; it is associated with recurrent bacterial infection. Deficiencies of properdin (X-linked) or the terminal pathway components C5, C6, C7, C8 and C9 are associated with recurrent neisserial (meningococcal or gonococcal) infection. Any

89

CASE STUDY 4.1 RECURRENT INFECTION An 11-year-old girl was referred because of a history of recurrent infection. She had been fully immunized as an infant. At the age of 4 years she had recurrent otitis media and underwent surgical insertion of tympanic membrane vents (grommets). These had required reinsertion the following year. She was well until the age of 9 years, when she presented to hospital with chest pain and pyrexia. Left basal pneumonia was diagnosed and she was treated as an inpatient. Four months later she presented with recurrent symptoms. Chest X-ray again confirmed left basal consolidation. After antibiotic treatment she was discharged and returned to school. However, her parents noted that she suffered from frequent unexplained pyrexias and persisting chest pain. Her school performance deteriorated, she complained of fatigue and withdrew from sporting activities. She had no history of allergy or family history of recurrent infection. Investigations Hb 11.5g/dL, WCC 9.2 x 109/L, IgG 4.59 g/L, IgA < 0.07 g/L, IgM 0.28 g/L. Questions 1. Give two likely explanations for her recurrent symptoms, 2. Do her preliminary investigations establish a diagnosis? 3. What further general investigations are indicated? 4. What further immunological investigations are required?

She was referred to an immunology clinic, where the details of history were confirmed and further investigation undertaken. The history was suggestive of recurrent bacterial infection and, with the suggestion of IgA deficiency, an antibody deficiency syndrome should be considered as the most likely diagnosis.

90

Further investigation indicated IgA < 0.05 g/L IgG subclasses: IgGj 4.5 g/L (4.5-9.4); IgG2 0.4g/L (1.6-5.1); IgG3 0.5 g/L (0.3-1.1); IgG4 0.07g/L (0.01-1.2) Classic complement pathway (CH50): normal Nitroblue tetrazolium (NET): normal Lymphocyte phenotype: CD4+ T-cell numbers low. CD45RA expression very low on CD4 T cells Immunization responses: failure to increase specific antibody levels after immunization with tetanus toxoid and Pneumovax II CT scan of chest: no evidence of structural lung damage. A diagnosis of common variable immune deficiency was reached and treatment commenced with prophylactic ivlg (400mg/kg three times weekly) commenced. The girl made a rapid response, with a cessation of her febrile episodes, no further chest infections and a significant improvement in energy levels. Her school performance improved and she re-engaged in her competitive sporting activities. Her weight correspondingly increased, to the extent that her ivlg dose had to be carefully monitored and increased appropriately. After 9 months on hospital-based therapy she opted for a home treatment programme. She and her mother were trained in all aspects of home infusion therapy by an immunology nurse specialist, and home therapy, under the guidance and monitoring of the Regional Immunology Centre, has been successfully established. Discussion The history of recurrent infection in this case was probably not remarkable up to the age of 9 years. It is common for children to present with otitis media, but tympanic membrane vents are usually indicated for chronic secretory otitis media ('glue ear'). There may therefore be

some doubt as to the exact diagnosis at the age of 4 years. The episode of pneumonia at the age of 9 years was significant, particularly because of the recurrence 4 months after initial treatment, and the persisting pleuritic pain and recurrent fevers thereafter. It was therefore appropriate to consider a primary immune deficiency and the initial investigations by the paediatrician were abnormal. The total serum IgG was slightly below the age-related normal range, and although IgA was reported as being, 'low', the level quoted was not low enough to confirm IgA deficiency. The additional investigations undertaken at the immunology clinic confirmed that the total serum IgA was <0.05 g/L and IgG2 subclass was low. The measurement of specific antibody response to immunization confirmed a significant defect in antibody production. Tetanus toxoid is a protein antigen (T dependent), whereas Pneumovax II is polysaccharide (T independent). Thus, these two antigens test the functional integrity of both the Tdependent and the T-independent responses. A healthy response should result in a minimum fourfold (usually much greater) rise in specific antibody level. Because there were low levels of IgG and IgA, low IgG2 and a failure of normal immunization response, the diagnosis was common variable immune deficiency (CVID). The lymphocyte phenotype is not diagnostic, but the relative reduction in CD4 T cells (and the CD45RA subset in particular) is typical of CVID. The clinical presentation described is reasonably typical, with the exception that the diagnosis was made before structural lung damage had occurred. Presentation in the second and third decades of life is common, but can be at any age. Usually there is a delay of many years from the onset of infection to

4

CASE STUDY 4.1 CONTINUED diagnosis, and during this period structural lung damage occurs. Ivlg treatment is proven to prevent infection and its complications, and in this case a specialist immunology service established home therapy for

the child. Home therapy is increasingly seen as a desirable option for patients with primary immune deficiency, as it reduces the need for hospital visits that interfere with normal school or work activities.

patient presenting with a recurrence of meningococcal infection should therefore be investigated for complement deficiency. Investigation Investigation of the complement system is indicated in any patient with symptoms of recurrent bacterial (especially neisserial) infection, 'lupus-like' disorders or angioedema. Total C3 and C4 levels are readily measured, but the most useful screening tests for deficiency are the assays that test the functional integrity of the classic and alternative pathways (known as CH50 and AP50, respectively). If an abnormality is detected, further characterization is necessary to identify the individual component that is defective. Management This is dependent on the clinical presentation. For patients with recurrent bacterial infection, prophylactic antibiotics may be indicated. For those susceptible to pneumococcal and or neisserial infection, immunization with pneumococcal capsular polysaccharide and meningococcal vaccines is often advised. In some cases intramuscular penicillin is kept at home for early treatment of suspected meningococcal disease, and it is advisable for patients to wear a medical warning bracelet. For patients with prominent joint symptoms, further rheumatological assessment may be advised. C1-inhibitor deficiency Defects of the control protein C1-inhibitor (C1-INH) are inherited in an autosomal dominant manner and cause hereditary angioedema. There is a failure of the normal regulation of the classic pathway that results in uncontrolled activation of complement and kinin pathways in response to even minor trauma. Clinical features Patients present with recurrent deep tissue swelling following apparently minor trauma (typically dental treatment). This may affect any part of the body, but is of most concern when the airway is compromised. Symptoms usually begin in early adolescence, although late presentation is increasingly recognized. An interesting feature is that attacks may begin, or become more severe, after commencement of the oral contraceptive pill, hormone replacement therapy, or during pregnancy. Although inherited as an autosomal dominant trait, patients may not give a significant family history as approx-

Integrating treatment into family life also empowers the patient and emphasizes that these conditions, like many other medical conditions, can be prospectively managed, preventing illness and maintaining good health.

imately 20% of cases are caused by new mutations in the C1-inhibitor gene. As with other immune deficiencies, there is often a period of diagnostic delay during which patients will have suffered symptoms, often of a lifethreatening nature. It is typical for children with this condition to suffer recurrent abdominal pain (caused by intestinal wall swelling), rather than swelling at other sites. In one series, 30% of patients had an undiagnosed relative who had died of laryngeal oedema, 34% had had an unnecessary laparotomy and 10% had had endotracheal intubation or tracheotomy prior to diagnosis. Investigation During an acute attack there is a characteristic pattern of serum complement levels which is caused by inefficient and excessive activation of the classic complement pathway. Serum C4 and C2 are low but C3 levels are usually normal. This is a very valuable pointer to the diagnosis hereditary angioedema, and measurement of C3 and C4 in any patient with angioedema will help differentiate this from allergic angioedema. In 85% of cases there is a low level of C1-inhibitor level (usually less than 50% of the lower limit of normal) but in 15% of cases the enzyme level is normal. This latter group have 'functional Clinhibitor deficiency', the enzyme being present but nonfunctional. Management Prophylactic treatment with androgenic steroids (danazol 200-400 mg daily, stanozolol 2.5-10 mg/day - alternate-day dosage may be possible) or antifibrinolytics (tranexamic acid up to 4g/day) is usually effective. Antifibrinolytics are the preferred alternative for female patients because of the androgens' virilizing side-effects. If laryngeal oedema occurs (e.g. after dental extraction) it is life-threatening and requires emergency intravenous administration of Cl-INH concentrate (1000-1500 units i.v). Patients are advised to wear a medical warning bracelet. C1-inhibitor concentrate should also be given prior to surgical procedures, and routine dental work should only be undertaken in hospital, where laryngeal oedema can be dealt with effectively. Cellular immune deficiency Because of the central role of T cells in coordinating the immune response, isolated T-cell deficiency is rare. The following will first consider those conditions in which the T-cell defect is the major feature, and then those in which a combined defect is more prominent.

91

DiGeorge's syndrome This is a rare syndrome associated with a developmental abnormality of the third and fourth pharyngeal pouches. The full syndrome consists of: • Abnormal facies with low-set ears, 'fish-shaped' mouth, hypertelorism, notched ear pinnae, micrognathia and a downward slant of the eyes;1 • Hypoparathyroidism (often presenting with neonatal hypocalcaemic tetany); • Congenital heart disease (particularly aortic arch defects: truncus arteriosus, interrupted arch or Fallot's tetralogy); • Cellular immune deficiency. Not all of these clinical features are present in every case, and the immune function usually improves with time. DiGeorge's syndrome is now recognized as one of a group of related disorders that share a common genetic lesion. This is usually a deletion at 22qll, and the group of conditions is referred to as the CATCH 22 syndrome (cardiac anomalies, abnormal facies, thymic hypoplasia, cleft palate and hypocalcaemia). Other clinical syndromes in this group include velocardiofacial syndrome, Kallman's syndrome and Schprintzen's syndrome. Not all the CATCH 22-associated conditions include immune deficiency. Investigation Diagnosis is based on the typical morphological features, associated with recurrent viral or fungal infections such as oral and perineal candidiasis and viral pneumonia. Chest X-ray may show an absent thymic shadow with an abnormal cardiac outline (Fig. 4.8). It is usual to find a reduced number of circulating T cells, but the degree of immunological deficiency is variable and is probably only significant in approximately 20% of cases identified by other criteria. Lymphocyte proliferation in response to mitogens (PHA and PMA) may be reduced. Antibody deficiency is rare. It is now possible to detect the chromosome 22qll lesion by fluorescent in-situ hybridization (FISH). However, as the range of clinical associations with 22q11 lesions is wide and the degree of immunodeficiency variable, the simple demonstration of a 22qll deletion does not necessarily imply that there is clinically important immune deficiency. Management Awareness of the features and early recognition of DiGeorge's syndrome is important, particularly as these infants may be subjected to neonatal cardiac surgery. Calcium levels must be monitored and supplemented as necessary. All blood transfusions must use X-irradiated (25 Gy/unit) blood only, and live immunizations are con-

FIG. 4.8 Di George's syndrome

Chest X-ray of a one-week-old female infant with hypocalcaemia and Fallot's tetralogy. Note the abnormal enlarged cardiac outline and absent thymic shadow consistent with her diagnosis of Di George's syndrome.

traindicated (MMR, oral polio vaccine). Antibiotic prophylaxis to prevent PCP should be considered. In our unit this is instituted if the absolute CD4 count is below 0.4 x 109/L or if there is reduced T-cell proliferation to phytohaemagglutinin (a T-cell mitogen). In cases with severe cellular immune deficiency, bone marrow transplantation is suggested. The place of fetal thymus grafting is still experimental, although there have been some recent encouraging results showing reconstitution of T cells. Other cellular deficiencies Isolated T-cell deficiencies occur but are very rare. The diagnostic approach involves recognizing a suggestive history and undertaking both quantitative and functional assessments of lymphocytes. Chronic mucocutaneous candidiasis (CMC) affects both males and females and presents with chronic candidal infection of skin and mucous membranes.2There may be an associated endocrine deficiency which may precede the immune deficiency, most commonly hypoparathyroidism or Addison's disease. There are normal T-cell responses to non-specific mitogens but a specific failure of proliferative response to Candida. NK cell deficiency is also reported; patients have an increased susceptibility to herpesviruses. Functional studies of NK cells are only performed in specialist centres.

Combined immune deficiencies Wiskott-Aldrich syndrome

92

1

Fig. 4.1

2

4

Figs 4.6, 4.7

Figs 4.2, 4.3, 4.4

3

Fig. 4.5

5

6

MCQ 4.3

Figs 4.8,4.9

Clinical features This X-linked disorder presents in young boys with a typical triad of eczema, thrombocytopenia and recurrent

infection.3 Immune deficiency may be mild. There is a high frequency of malignant lymphoma in the second and third decades of life, and severe autoimmune vasculitis or glomerulonephritis may occur. If untreated, death from one of these causes is usual in the second or third decade. Investigation Platelets are small and have abnormal expression of surface glycoproteins (CD43 and gpIb). There are defects affecting the cytoskeletal proteins, including actin, and associated defects of T-cell proliferation. The molecular defect is mutation of the WASP gene at Xp11.22. Characteristically there is a failure of response to polysaccharide antigens and progressive lymphopenia. Management These patients require careful clinical and laboratory assessment to establish optimum management. Increasingly, affected males are offered bone marrow transplantation in early childhood in order to reduce the risk of long-term complications and increase life expectancy. Ataxia-telangectasia (AT) Clinical features This progressive neurological and immunological disorder presents in early childhood with neurological manifestations, including cerebellar ataxia, nystagmus and oculomotor dyspraxia. Telangiectases are typically seen on the elbows, conjunctivae and ear lobes. 4 There is a progressive, combined immune defect that is variable in its immunological features. Patients ultimately develop lymphoid leukaemia/lymphoma in the second and third decades and this, combined with progressive combined immune deficiency, is the usual cause of death. Investigation The fundamental lesion is defective DNA repair leading to increased radiosensitivity (such as is seen in other disorders, e.g. Bloom's syndrome, xeroderma pigmentosa, Nijmegen breakage syndrome). Chromosomal breaks, inversions and translocations are commonly seen at the T-cell receptor, and immunoglobulin gene complexes on chromosomes 7 and 14, respectively. The serum afetoprotein is often raised. Heterozygotes for the AT gene mutations have an increased risk of developing a number of malignancies (including breast cancer), and family members need to be aware of this. Patients with AT gradually develop immune deficiency and, after comprehensive immunological investigation at diagnosis, this is confirmed and monitored by regular review of the clinical and laboratory features. Management Careful, continuing immunological assessment is essential in these patients, and treatment is dependent on the degree of immune deficiency. Patients commonly require ivlg by late childhood. There is no corrective treatment available for the underlying chromosomal instability.

Severe combined immune deficiency (SCID) Clinical features This group of conditions has a number of causes but the clinical features are similar in most cases. By definition, both humoral and cellular responses are impaired or absent. SCID typically presents early (in the first weeks and months of life) with a history of failure to thrive (FTT), unexplained diarrhoea, and recurrent bacterial, viral or fungal infection. Unexplained lymphopenia (<2.8 x 109/L) is often present and should always prompt investigation. Total lymphocyte counts may, however, be apparently normal in B-cell-positive variants, T-cell activation deficiencies and graft-versus-host disease (GVHD). GVHD may be related to maternofetal transfusion, where normally functioning maternal lymphocytes attack the immunodeficient child. If present, GVHD occurs in the first month of life with jaundice, typical skin rash 5and diarrhoea. It may also occur as a complication of blood transfusion, and all transfusions to potentially immunodeficient individuals should be irradiated (25Gy/unit) to eradicate immunocompetent lymphocytes, the potential inducers of GVHD.

4

Investigation When SCID is suspected, immunological advice should be sought urgently, with regard to both investigation and immediate management. In general, serum immunoglobulins are low and there is a low lymphocyte count, with diminished proliferative response to mitogens. Lymphocyte subset analysis is essential and a range of other investigations is needed to exclude SCID. The molecular basis of many variants of SCID is now recognized: each has a particular pattern of abnormalities and some of the more common are summarized in Table 4.6. Molecular characterization allows greater precision in diagnosis and carrier detection. Management Early diagnosis, prevention of infection and rapid bone marrow transplantation (BMT) are all crucial to ensure ultimate survival. Delay in making the diagnosis allows complications to occur (such as disseminated CMV infection), which reduce the chance of survival. BMT can be curative and, in the UK, is currently performed at two supraregional centres. With increasing characterization of the molecular basis of immune deficiency, it is hoped that gene therapy will become a practical possibility in this group of disorders.6

Phagocytic disorders Phagocytes may be defective in numbers or function. Table 4.7 summarizes the main disorders, the commonest of which are described below. Chronic granulomatous disease (CGD) Clinical features CGD typically presents in childhood with recurrent, deep abscesses caused by catalase-positive organisms (Staphy-

93

TABLE 4.6 The major features of some combined immunodeficiency syndromes Syndrome

Major immunological features

Molecular defect

Inheritance/associated features

X-linked SCID

Immunoglobulin low; T cells low, B cells normal/increased Immunoglobulin, T and B cells all low Immunoglobulin, T and B cells all low Immunoglobulin low. Progressive fall in T and B cell numbers IgG low, IgM high, T cell numbers normal Immunoglobulin normal/low, B cells normal, reduced CD4+ cells Immunoglobulin normal/low, B cells normal, reduced CD8+ cells.

IL-2 receptor y chain

(XL)

Mutation in RAG1/2 genes Stem cell defect Accumulation of lymphotoxic metabolites Mutation in CD40L gene Mutations in MHC II regulator genes TAP-2 gene mutation

(AR)

T-B- SCID Reticular dysgenesis Adenosine deaminase deficiency X-linked hyper-IgM MHC class II deficiency TAP-2 deficiency

(AR): Anaemia, granulocytopenia, thrombocytopenia (AR): Progressive onset, diagnosis may be delayed into childhood: flared ribs (XL) Neutropenia, thrombocytopenia, haemolytic anaemia, (AR) (AR) MHC class I deficiency

(XL), X-linked; (AR), autosomal recessive.

TABLE 4.7 Examples of phagocytic disorders Phagocytic defect

Clinical disorder

Numbers Adhesion

Cyclical neutropaenia Leukocyte adhesion defect (LAD-1) Schwachmann's syndrome

Chemotaxis

Intracellular killing

Chronic granulomatous disease G6PD deficiency IL-12R & IFNyR deficiency

Inheritance/associated features (AR) See text (AR) See text (AR) Anaemia, thrombocytopenia, pancreatic insufficiency, chondrodysplasia (XL/AR) See text (XL) Anaemia (AR) Susceptibility to mycobacteria and salmonella. See text

lococcus aureus, Aspergillus, Serratia, Nocardia).1 Presentation may be with osteomyelitis, unexplained granulomatous lesions, malabsorption or recurrent pneumonia. New cases have been diagnosed in adulthood. Inheritance may be X-linked or autosomal recessive. Investigation The diagnosis rests on the demonstration of a failure of the neutrophil respiratory burst. The most common screening test of neutrophil function is the nitroblue tetrazolium test (NBT) test, in which normal neutrophils reduce the yellow dye formazan into blue intracellular crystals. 2 This test

1 94

Fig. 4.10

0 Fig. 4.11 3

Fig. 4.12

may not be abnormal in all cases, and more sensitive screening techniques have been developed. The molecular defect is a deficiency of the 91 kDa chain of cytochrome b (X-linked CGD) or the 22 kDa chain of cytochrome b, or of p47 or p67 cytoplasmic factors (autosomal recessive CGD). These proteins are all involved in the energydependent process of intracellular killing. The definitive test is to demonstrate a failure of neutrophil-mediated bacterial killing. Management Antimicrobial prophylaxis (usually with co-trimoxazole) is the mainstay of therapy. Prophylactic interferon-y (IFNy) is not useful prophylactically, but may be a useful adjunct to conventional antibiotic and surgical management of abscesses. Allogeneic bone marrow transplantation, before the development of complications of infection, is now the treatment of choice for CGD if a suitable donor can be found. Cyclical neutropenia Clinical features This condition, which may present in childhood or adult life, is characterized by recurrent abscesses, usually occurring at intervals of approximately 3 weeks, although the periodicity can vary greatly. The neutrophil count is typically <1 x 109/L at times of infection, and the diagnosis should be suspected if there is clear periodicity to the history. Alternatively, the neutrophil count may be incidentally noted as low on a blood count during acute infection. Investigation Neutrophil counts should be measured on alternate days for a 4-week period and carefully analysed for any evidence of a cyclical pattern. A cyclical pattern may also be detected for other elements, including platelets, and other leukocytes.

Management This is dictated by the clinical severity of the condition. Prophylactic antibiotics may be adequate during times of low neutrophil numbers. In more severe cases, treatment with granulocyte colony-stimulating factor is considered. Leukocyte adhesion deficiency (type 1)

heterozygous for the genetic lesion, but this is clinically expressed as a dominant susceptibility. This is in contrast to most autosomal genetic disorders, in which the normal allele dominates, so that heterozygotes only show a minor clinical abnormality. The prevalence of these conditions is not yet established.

Clinical features This is typically a disorder of infancy and childhood. Chronic/recurrent skin ulcers and periodontitis are common. There is characteristically little pus formation and an associated peripheral blood neutrophilia. There is likely to be a neonatal history of delayed separation of the cord (>10 days).

Investigation Cell surface expression of IFNyR and IL-12R is detected by flow cytometry. Specialist functional assays of lymphocyte signalling are necessary to confirm and characterize the defects. Histologically, complete IFNyR deficiency is characterized by a failure to form granulomata, whereas partial deficiency is associated with good granuloma formation.

Investigation There is defective expression of the CD18 (B chain) of the adhesion molecules LFA-1, Mac-1 and CR4. The absent expression is demonstrated on both lymphocytes and phagocytes by flow cytometry. A second variant, LAD-2, is associated with defective expression of CD15.

Management Once the diagnosis is established, appropriate antimycobacterial therapy must be commenced. Experience with adjunctive IFN-y treatment in the partial forms is increasing and discussion with specialist centres is advised.

Management Antibiotics are the mainstay of therapy. Drug choice should be guided by antibiotic sensitivities. Prophylactic therapy may be necessary, depending on the frequency of infection. Bone marrow transplantation is necessary for severely affected individuals. /nferferon-y receptor (IFNyR) deficiency Clinical features Defects in the IFNyR (and IL-12 receptor) systems have recently been identified as important susceptibility factors for mycobacterial infection. These conditions have been identified in both children and adults. They typically develop infection with poorly pathogenic mycobacteria, salmonella, or the live vaccine Bacille Calmette-Guerin (BCG). There are three types of IFNyR defect identified. Complete deficiency determines a phenotype that is characterized by severe and fatal susceptibility to either BCG or disseminated non-tuberculous mycobacterial (NTM) infection. Partial IFNyR-1 deficiency is associated with a less severe clinical defect that may allow disseminated BCG infection with tuberculoid granulomata or clinical tuberculosis. Both complete and partial IFN-y deficiencies are inherited as autosomal recessive traits. A functional defect of the IFNyR has also recently been described which is characterized clinically by susceptibility to nontuberculous mycobacteria (NTM) and variable granuloma formation.3 A notable feature of the cases described to date is that a significant number are initially misdiagnosed as suffering from Langerhans' cell histiocytosis (LCH), because of the association of fevers, lymphadenopathy, granulomatous ulcers and lytic bone lesions. The condition is caused by a premature stop codon in the intracellular signalling domain of IFNyR that reduces the mycobactericidal properties of macrophages. Inheritance of this last type is 'dominant negative'. The affected individuals are

a

SECONDARY IMMUNODEFICIENCY Secondary immunodeficiency is much more common than primary immune deficiency in clinical practice. It may occur as a result of systemic disease, including diabetes mellitus, chronic hepatic and renal failure and systemic malignancy. Autoimmune disorders may be associated with immune deficiency as a consequence of both disease and treatment. Specific infections are also associated with secondary immune deficiency. The best known of these is human immunodeficiency virus (HIV), which leads to the acquired immune deficiency syndrome (AIDS; Chapter 11). HIV causes a progressive loss of CD4+ T cells and thus disrupts both antibody and cellular responses. Several other infections are associated with immune suppression. Measles predisposes to bacterial infection partly through its disruption of phagocytic function. Epstein-Barr virus may occasionally cause hypogammaglobulinaemia. Immunosuppressive drugs are widely used in the treatment of cancer and autoimmune disease and are an important cause of secondary immunodeficiency. When using these agents it is essential to monitor treatment carefully and maintain the lowest level of immunosuppression compatible with effective treatment. Malnutrition, infection and other iatrogenic causes, including radiotherapy, are also important causes of secondary immune deficiency. The mechanisms of secondary immunosuppression are usually multifactorial: specific intervention is not usually indicated, but susceptibility to infection must be recognized in patients who are at risk. Secondary antibody deficiency may occur in lymphoproliferative disorders such as multiple myeloma and chronic lymphocytic leukaemia, or in protein-losing states such as nephrotic syndrome or intestinal lymphangectasia. In the

95

lymphoproliferative disorders, where there is a failure of antibody synthesis, there is often a history of recurrent bacterial infection, and in these cases ivIg has been demonstrated to be beneficial. Interestingly, in the protein-losing states, recurrent infection is not usually a significant clinical problem, and antibody replacement therapy is the exception rather than the rule. Splenectomy is an important cause of secondary immune deficiency. These patients have a lifelong risk of suddenonset overwhelming sepsis caused by encapsulated organisms. This is related to the important role of the spleen in the immune response to polysaccharide antigens. Patients in whom Splenectomy is planned should be immunized with pneumococcal, meningococcal and Haemophilus influenzae type b vaccines. Annual flu immunization is recommended, and travel to malarial endemic areas should be avoided if possible. Patient education regarding the risks of splenectomy is very important, and they should be provided with warning cards or bracelets and advised to take penicillin early if they develop cough and fever. Prophylactic penicillin V (250 mg b.d.) should be given for the rest of the patient's life. Penicillin-allergic individuals should take erythromycin (250mg b.d.). Long-term follow-up is recommended both to monitor the need for repeat immunization and to reinforce medical advice. There are also a number of inherited syndromes, many of which are rare, associated with secondary immune deficiency.

HYPERSENSITIVITY

IMMUNOLOGICAL MECHANISMS In contrast to immunodeficiency syndromes, where the immune system is underactive, clinical disease may result from inappropriate overactivity. Overactivity of the immune system is referred to as hypersensitivity. The immunological mechanisms that cause hypersentivity were described and classified by Gell and Coombs (Table 4.8). Four types of hypersensitivity are described, each involving different immunological components. In some cases the hypersensitive response is directed at a foreign antigen (e.g. reactions to allergens, blood transfusions, transplanted organs or infectious agents) and in others it is directed at known 'self-antigens' (e.g. in autoimmune diseases). Finally, some hypersensitivity reactions appear to be directed at unknown antigens (e.g. Wegener's granulomatosis). The term hypersensitivity applies to all four types of reaction in the Gell and Coombs classification, whereas

1

96

Fig 4.13

2

Case 4.1

TABLE 4.8 Classification of hypersensitivity reactions (Gell and Coombs classification) Type of reaction

Effector mechanism

Clinical disorders

I

IgE, mast cells

Allergic rhinitis, urticaria, angiodema

II

IgG directed at cell surface antigens

Transfusion reactions, acute graft rejection, Graves' disease, myasthenia gravis, Goodpasture's syndrome

III

Immune complexes

Cryoglobulinaemia, SLE, post streptococcal glomerulonephritis

IV

CD4+ lymphocytes,

Delayed type hypersensitivity, contact eczema, granulomatous reactions

the term allergy is used in a more restricted sense to describe type I hypersensitivity reactions. In this section, we will use the Gell and Coombs classification to illustrate and explain the important clinical conditions associated with hypersensitivity reactions. In addition to allergy and autoimmunity, the hypersensitivity mechanisms involved in the response to infectious agents can influence the clinical manifestations of disease. Furthermore, the manipulation of the immune response (e.g. in mycobacterial disease) offers new possibilities for therapeutic intervention.

Type I hypersentivity Type I hypersensitivity reactions (the major mechanism in allergy) are the result of the immune system reacting to otherwise apparently innocuous antigen (allergen). The mechanism involves: • Formation of IgE antibody specific to the inducing antigen or 'allergen'; • Binding of this antibody (via Fc receptors) to mast cells and basophils; • Cross-linking of at least two cell-bound IgE molecules; • Degranulation of mast cells, with release of preformed mediators and the generation of late-phase mediators.

Type II hypersentivity Type II hypersensitivity involves the reaction of antibody with cell surface-bound antigen. This reaction then causes downstream activation of the complement system, and attraction and activation of phagocytes and mast cells. The effect is to focus inflammatory damage at the site of antibody binding or, in the case of receptor-binding antibodies, to modulate the function of the target organ. This mechanism is especially important in organ-specific

autoimmune disease (Graves' disease, myasthenia gravis, autoimmune haemolytic anaemia, Goodpasture's syndrome), but also in blood transfusion reactions and hyper acute graft rejection.

Type III hypersensitivity Type III reactions involve the formation of soluble immune complexes that circulate in the blood and cause damage at many sites. Formation of immune complexes is part of the normal immune response, and these are normally disposed of via the complement and reticuloendothelial systems. Complexes remain in the circulation if they are produced persistently or are inadequately disposed of. Such pathological complexes become deposited, typically in the renal glomerulus, in joints and the skin, causing inflammatory reactions at these sites. The experimental model for immune complex disease is serum sickness, in which nephritis, arthritis and rash typically occur 7-10 days after the injection of foreign protein. The best examples in human disease include chronic infection (e.g. hepatitis, malaria, streptococci), systemic lupus erythematosus (SLE), inhalation of antigens (e.g. farmers lung) and cryoglobulinaemia. 1 2

Type IV hypersensitivity This is also known as delayed-type hypersensitivity (type IV hypersensitivity) because in experimental systems signs develop 12 or more hours after antigen exposure. Importantly, the major immunological effectors of type IV reactions are T cells, instead of antibody, which mediates the reaction in types I-III hypersensitivity.

CLINICAL DISEASE ASSOCIATED WITH HYPERSENSITIVITY REACTIONS Allergy Clinical allergy is becoming increasingly prevalent in western societies: it is estimated that up to 20% of the UK population is now affected by allergy. The symptoms depend mainly on the nature of the allergen and the site at which mast cell degranulation occurs. The common clinical syndromes are of seasonal or perennial rhinitis, allergic asthma, urticaria, oral allergy syndrome, angioedema and anaphylactic reactions. The term atopy is used to describe a genetic susceptibility, expressed in the clinical triad of eczema, allergic rhinitis and asthma, which are commonly associated in many patients. Diagnosis The diagnosis of allergy is often thought to be difficult, but in the majority of cases a thorough clinical history combined with relatively simple investigations will prove adequate. The history is the most important component, and

one looks for typically allergic symptoms occurring consistently after exposure to the same agent, and improvement of symptoms during periods of avoidance. Where doubt remains, further investigation by skin-prick testing and/or the detection of allergen-specific IgE (e.g. by radioallergosorbent test - RAST) is necessary. Skin-prick testing is preferred, as it is a rapid, cheap and painless form of in vivo challenge. There are, however, limitations to its use, as it should only be undertaken by trained individuals in an environment with full resuscitative facilities available. Specific IgE testing using the RAST test has the advantage of easy accessibility. Although results correlate well with skin testing for most allergens, they are not directly comparable for every allergen. RAST testing should not be used as a screening test for allergy. Its use should be guided by careful clinical assessment.

4

Allergic syndromes Rhinitis Clinical features Seasonal or perennial rhinitis usually presents little diagnostic difficulty or need for specific investigation. The symptoms of a red, itchy, runny nose with associated swelling and itchy watery eyes are easily recognized. A history of seasonality and/or key triggering factors will usually indicate the offending allergen. Investigation This is required when the provoking allergen is not clear from the history, or when another cause is suspected (Table 4.9). Management The important elements in management are the avoidance of, or reduction in exposure to, the allergens and the use of nasal steroid inhalers, with or without oral antihistamines. The latter are particularly useful in young children, who may find the use of inhalers difficult. For perennial symptoms caused by house dust mite patients need to be advised on how to reduce exposure by modifications to floor coverings, bedding, and cleaning and laundering procedures. There is increasing evidence that effective control of nasal

TABLE 4.9 Differential diagnosis of rhinitis Allergic (aeroallergens) Drug induced (decongestant sprays, cocaine, antihypertensive) Vasomotor Infectious Irritant Vasculitis Hormonal (pregnancy) CSF rhinorrhoea

97

allergy in asthmatic patients can improve their overall airway disease. Desensitization is returning to clinical practice for grass and house dust mite allergy, but only specialist units should undertake it.

Urticaria/angioedema Clinical features Urticarial lesions are raised red itchy weals (often described as 'like nettle stings') caused by epidermal mast cell activation. Angioedema is non-itchy, non-dependent swelling which occurs in the deep dermis. Activation of the kinin system is important in causing angioedema. It is common for both symptoms to occur in the same patient. However, when angioedema occurs in the absence of urticaria, hereditary or acquired deficiencies of Cl-esterase inhibitor (Cl-INH) should be considered. A number of specific triggers for both urticaria and angioedema may be identified (Table 4.10), but in the majority (89-90%) of cases referred to clinics no specific allergic cause is identified. If symptoms persist for more than 6 weeks, this is defined as chronic idiopathic urticaria/angioedema. The natural history is quite variable, with many patients experiencing prolonged remissions between relapses. Investigation A thorough clinical history is essential to determine the pattern of symptoms and any associated trigger factors. Screening tests for allergy are not recommended; however, the wide variety of disorders associated with combined urticaria/angioedema (Table 4.10) may justify further investigation, including thyroid function tests, full blood count, CRP and complement levels. Skin biopsy is helpful in confirming urticaria pigmentosa and urticarial vasculitis. Management If an associated medical condition is identified (e.g. hypothyroidism, infection), the urticaria should respond to management of that disorder. In other patients, if attacks are only infrequent oral short-acting antihistamines (e.g. acrivastine 8mg t.i.d.) will be adequate. In most cases, however, daily non-sedating antihistamines are necessary (e.g. fexofenadine, loratadine and cetirizine). If necessary a sedating antihistamine can be added at night (e.g. hydroxizine hydrochloride). Experience with oral disodium chromoglycate has been disappointing, with the possible exception of urticaria pigmentosa, in which it has an important role in management. H2 antagonists (cimetidine and ranitidine) are often tried for refractory cases, as is doxepin (an antidepressant with potent antihistamine effects). Urticarial vasculitis often responds to NSAID therapy. The relapsing, remitting nature of urticaria makes it difficult to be certain of the efficacy of some of these interventions,

1

98

Fig. 4.14

TABLE 4.10 Causes of urticaria/angioedema Idiopathic Allergic (foods, drugs, hymenoptera venom, contact) Physical (pressure, heat, solar, aquagenic, vibratory) Cold induced (autosomal dominant, cryoprotein) Cholinergic (exercise or heat induced) Adrenergic (stress induced) Vasculitis (hypocomplementaemic urticarial vasculitis) Autoimmune (SLE, antibodies to lgE/FceR1/C1q) Infection (hepatitis B, Helicobacter pylori, EBV, Lyme's disease, cutaneous larva migrans, larva currens) Insect bites (papular urticaria) Drug induced (aspirin, ACE inhibitors, opiates, muscle relaxants, and radiocontrast media) Hypothyroidism Urticaria pigmentosa Systemic mastocytosis Angiodema/deafness, urticaria (Muckle-Wells syndrome) Malignancy

and it is essential that once regular therapy is commenced the dose is periodically reduced or withdrawn to determine whether continued treatment is necessary. Food allergy IgE-mediated allergy to food has been described since the 1920s but remains a source of controversy. This is probably because there are exaggerated claims made for both the prevalence and the effects of food allergy, and the diagnostic tests that are available have not been reliable. There is also confusion between food allergy (an IgE-mediated phenomenon) and food intolerance, which may have many causes. For patients in whom food allergy is suspected a double-blind placebo controlled food challenge (DBPCFC) is the only reliable investigation, but this is available in only a few specialist centres at present. Selfreporting of food allergy is common. In one large study the self-reported prevalence was 20%, but actual prevalence was only 1-2% when confirmed by DBPCFC. Food allergy is often invoked as a potential cause for complex multisystem disorders or obscure symptoms. This is unlikely to be the case, and in the absence of clear evidence of allergy patients should be encourage to eat a normal diet. IgE-mediated food allergy may cause symptoms including urticaria, angioedema, rhinitis, asthma, oral allergy syndrome, anaphylactic reactions and some gastrointestinal symptoms (chiefly acute vomiting and/or diarrhoea). Common food allergens include milk, eggs and seafood. Affected children typically become tolerant to these foods by mid-late childhood. Currently, the most important food allergy is to peanut - a relatively unknown condition until the late 1980s, the prevalence of which has increased considerably. Among the UK preschool population it is now estimated to be as high as 1:50-100. Nut allergy is different in many important respects from other food allergies:

• It appears to develop in very young children. • Reactions characteristically cause angioedema, often affecting the larynx, with airway restriction. • Anaphylactic reactions may result from exposure to even trace amounts of nut. • Indirect exposure may cause life-threatening reactions in highly sensitive individuals. • Nut allergy appears to be a lifelong problem in the majority of cases. Nut allergy is a major challenge facing the medical profession and others who look after children. Successful management requires the complete avoidance of all nut-containing foods, and teaching family, friends and school staff the symptoms and signs of allergic reactions. An emergency medication kit should be available which includes both oral antihistamines and injectable adrenaline (epinephrine) (for life-threatening reactions), and patients are encouraged to wear a medical warning bracelet. Nut allergy has now overtaken bee/wasp sting allergy as the most common cause of anaphylactic reactions occurring outside hospital.

Development of autoimmunity The occurrence of hypersensitivity reactions directed against self-antigens implies a breakdown in the body's normal tolerance mechanisms. A number of factors are implicated in the development of autoimmune reactions, including infection, drugs and chemical exposure and HLA type. The influence of the hormonal state is undoubted and is exemplified by the increased autoreactivity seen in women and the effect of hormonal change (pregnancy and lactation, hormone replacement therapy) on the clinical manifestations of disease. Several cellular and molecular mechanisms are recognized as contributing to autoreactivity. These include 'cross-reactivity', where a pathogen is sufficiently similar to self-antigen that it induces an autoimmune response, e.g. group A B-haemolytic streptococcus cross-reacts with cardiac muscle, causing acute rheumatic fever: 'Molecular mimicry', wherein a short genetic sequence of a pathogen is identical to a selfantigen, has long been postulated as a mechanism for autoimmunity. Mathematical modelling, however, suggests that molecular mimicry is likely to be so common as to be the rule rather than the exception, and thus unlikely to be important in disease aetiology. 'Provision of foreign T-cell epitopes' implies that foreign antigens containing T-cell epitopes bind to host proteins and stimulate an autoimmune response. This may be one of the mechanisms through which drug and chemical exposure causes autoimmune reactions. The important HLA disease associations have been summarized in Table 4.3. It should be remembered that T cells learn to differentiate self from non-self antigens during their maturation in the thymus; thus the occurrence of autoreactive antibodies or lymphocytes represents a breakdown in the normal mechanisms of selftolerance. One of the crucial mechanisms for eradication of self-reactive T cells is through programmed cell death

or apoptosis. Patients deficient in the Fas antigen, an essential signalling molecule in normal apoptosis, develop characteristic clinical features dominated by lymphoproliferation and autoimmunity,1 emphasizing the importance of central (thymic) deletion of autoreactive cells.

4

Autoimmune diseases: classification and diagnosis Human autoimmune diseases are classified according to whether predominantly a single organ is affected (organ specific) or whether there is multiple organ involvement (non-organ specific). As with any classification system, some conditions fall between these two extremes. Table 4.11 lists some common autoimmune diseases in these two categories. Autoantibody detection is frequently requested in the investigation of suspected autoimmune disease. Autoantibodies may be primary and directly cause disease (e.g. antiglomerular basement membrane antibodies in Goodpasture's syndrome, antiacetylcholine receptor antibodies in myasthenia gravis and anti-TSH receptor antibodies in Graves' disease), but more commonly they are secondary and are formed as part of the disease process itself (e.g. antinuclear, anti-dsDNA, antismooth muscle, antimitochondrial antibodies). Detection of a secondary autoantibody does not necessarily indicate that the patient has an autoimmune disease. 'False positive' autoantibodies are associated with, for example, infection, female gender, increased age and surgical procedures. Determining the significance of autoantibody results requires careful clinical assessment. It is also important to know the exact pattern of autoantibody detected, its titre (or strength) and

TABLE 4.11 Autoimmune diseases and their associated autoantibodies Disease

Autoantibodies directed against

Hashimoto's disease Graves' disease Pernicious anaemia

Thyroid peroxidase TSH receptor Intrinsic factor Gastric parietal cells Steroid-secreting cells of adrenal cortex Islet cell Smooth muscle Acetylcholine receptor Voltage-gated calcium channel

Addison's disease Type 1 diabetes mellitus Chronic active hepatitis Myasthenia gravis Lambert-Eaton myasthenic syndrome Pemphigus Goodpasture's syndrome Sjb'gren's syndrome Primary biliary cirrhosis Systemic lupus erythematosus (SLE) Rheumatoid arthritis

Desmosomes Basement membrane Nuclear antigens (Ro/La) Mitochondria Nuclear antigens (ds DNA, Ro/La, Sm) Phospholipid IgG (rheumatoid factor-IgM class)

99

CASE STUDY 4.2 ACUTE ONSET OF SWELLING A 15-year-old girl presented to her local accident and emergency department with an acute onset of swelling that affected her face. Her eyes were closed, she had swelling of the lips, but neither tongue swelling nor difficulty in breathing (Case Fig. 4.2.1). She reported eating a chocolate bar that contained peanuts 4 hours earlier, and a provisional diagnosis of nut allergy was made. She was immediately treated with i.v. chlorpheniramine and hydrocortisone. Adrenaline was not administered but she was admitted for observation and remained in hospital for 4 days, until her symptoms had completely resolved. She was discharged home with an adrenaline (epinephrine) selfadministration kit and told to read the package insert to explain its use.

Questions 1. How common is peanut allergy in the UK? 2'. Was her presentation typical of nut allergy? 3. If she did have peanut allergy was her treatment optimal?

Further history Her GP referred her for further investigation as he felt that she might have more than just nut allergy. A detailed history revealed that she had suffered recurrent swelling of her hands since the age of 4 years. She also had been investigated for intermittent abdominal pain and painful knees. She had eczema affecting her face and wrists, mild asthma and hay fever. She had previously had positive skin-prick tests to house dust mite and dog and cat dander. She was unable to wear cheap jewellery as this caused a rash. On examination she was obese, she 100

Further investigation The disparity between the low C4 and normal C3 levels prompted measurement of Cl-esterase inhibitor levels. These were normal (Clesterase inhibitor 0.52 g/L (0.28-0.5)); however, functional assessment of Cl-esterase inhibitor activity was 0% (70-140), indicating a functionally defective molecule and confirming a diagnosis of Cl inhibitor deficiency (type II).

CASE FIG. 4.2.1 Facial angioedema that is preventing the patient from opening her eyes or speaking had eczematous patches at her right wrist, marked dermographism, and respiratory examination confirmed scattered inspiratory wheeze. Investigations indicated: total IgE 731 kU/L, specific IgE to house dust mite (strongly positive), grass pollen (moderately positive), peanut (negative). Autoantibody screen negative, rheumatoid factor negative, C3 1.38g/L (0.7-1.7), C4 < 0.1 g/L (0.13-0.43), CRP 1.5mg/L (<6mg/L).

Questions 4. How does this further history affect the differential diagnosis? 5. Do the investigations rule out nut allergy? 6. What other condition(s) should be considered? 7. Do any of these have significantly different therapeutic implications?

Discussion This case demonstrates an important differential diagnosis that must be considered in anyone presenting with angioedema. Although the common causes of urticaria/angioedema are as listed in Table 4.10, Cl-esterase inhibitor deficiency (Cl-INH deficiency) typically causes angioedema but not urticaria. This presentation is typical in a number of respects of the experience with ClINH-deficient patients. She was symptomatic as a child, with some hand swelling and prominent abdominal pain. In her case this was diagnosed as abdominal migraine, but the more likely cause was Cl-INH deficiency-related abdominal pain, which is thought to be caused by angioedema affecting the bowel wall. The development of facial swelling as a teenager is also typical. Attacks commonly become more frequent and severe after puberty (and during pregnancy), the explanation for this being the relative changes in hormonal balance affecting Cl-inhibitor synthesis from the normal allele. Interestingly, however, the patient did have an atopic background with eczema and asthma, as well as a history of nickel sensitivity. On examination she also had dermographism. Thus she was demonstrating conditions characterized by types 1 and 4 hypersensitivity reactions, and this undoubtedly contributed to the initial

4

CASE STUDY 4.2 CONTINUED clinical diagnosis in the A&E department of probable allergic angioedema. The importance of the complement investigations in angioedema cannot be overestimated. Cl-INH deficiency, cryoglobulinaemia or genetic deficiency of C4 usually explains the finding of very low C4 despite normal C3. Thus measurement of complement levels is an extremely useful investigation, and a normal C4 during an attack of angioedema virtually excludes the diagnosis of Cl-INH deficiency.

The treatment of acute lifethreatening angioedema in this condition is very different from that required for allergic angioedema. Life-threatening attacks commonly occur after dental work that precipitates laryngeal angioedema and respiratory obstruction. Emergency treatment consists of 1000-1500 units Cl-INH concentrate intravenously, with close clinical observation in hospital. Prophylactic therapy is also considered for patients with frequent attacks. In this

whether it is of the IgG or the IgM isotype. For example, a significant antinuclear antibody (ANA) would be of high titre (>1:80) and IgG class. If the ANA pattern was not homogeneous (e.g. speckled or nucleolar) this might indicate specific subsets of connective tissue disorders (mixed connective tissue disease, systemic lupus erythematosus, scleroderma). Once an antinuclear antibody is detected, specialist immunology laboratories will further characterize it according to which nuclear antigen is targeted (Table 4.11). Again, this is helpful in the subcategorization of connective tissue disorders and the recognition of susceptibility to specific complications (e.g. anti-Ro antibodies and neonatal congenital heart block).

Diseases associated with type IV hypersensitivity Contact eczema The most common clinical example is nickel hypersensitivity. Affected individuals develop an eczematous reaction to nickel-containing jewellery, watches, buttons etc. that are in prolonged contact with the skin. The reaction is usually confined to the area of contact. Although sometimes referred to as 'nickel allergy', IgE mechanisms are not involved in the pathogenesis of this condition. Nickel is absorbed directly through the skin and binds to selfproteins (haptenization).This stimulates CD4+ T cells, which are recruited to the area, and there is associated oedema of the epidermis. This reaction occurs over 24-48 hours and is artificially recreated during 'patch testing', which is widely used to investigate type IV hypersensitivity. Granulomatous hypersensitivity Granulomatous reactions represent the full development of type IV hypersensitivity reactions, in which antigen persists and a focus of inflammation is established. The tuberculin reaction is a modified, self-limiting form which occurs, for example, with the Mantoux and Lepromin tests.

case the patient was commenced on cyclokapron 500 mg q.i.d., in preference to androgenic steroids, because of the potential virilizing effects of the latter. She was also provided with a warning bracelet that gave details both of her diagnosis and the necessary treatment in case of an emergency. Family studies confirmed that her biological parents were not affected and that this was one of the significant minority of cases caused by a new mutation.

These involve the intradermal injection of M. tuberculosis and M. leprae extracts. The local skin reaction is characterized by an area of induration and swelling which is mediated by sensitized lymphocytes. CD4+ lymphocytes predominate and accumulate, along with macrophages around blood vessels, with a maximum infiltration at about 48 hours. This reaction is focused at the site of foreign antigen in the dermis and is self-limiting, as antigen is progressively degraded and removed. Where antigen persists, however, granulomatous features can develop. Granulomas are collections of macrophages, some of which form giant cells; others appear as epithelioid cells. A cuff of small lymphocytes surrounds these. Granulomas may develop in response to foreign antigen (Mycobacterium infections) or as part of a presumed autoimmune response (Wegener's granulomatosis, sarcoidosis or CVID). The clinical effects of granulomas depend on the underlying cause, and the number, size and location of the lesions. They may cause significant destruction and morbidity, e.g.TB in the lungs, or they may be an incidental finding, e.g. in CVID in the liver. In mycobacterial infection it is recognized that the immunoregulatory Thl/Th2 response influences the degree of granuloma formation and indeed the individual's clinical manifestation of infection. If the response is predominantly of the Thl phenotype, with activated T cells producing IFN-y and IL-2, there is marked granuloma formation and, in the case of leprosy, the clinical pattern is tuberculoid. Tuberculoid leprosy is characterized clinically by localized granulomatous lesions which may be asymptomatic and which contain few viable organisms. In lepromatous leprosy there are widespread skin lesions containing numerous bacilli but few lymphocytes. These patients have a predominant Th2 phenotypic response to the organism (Th2 cells secrete IL-4, IL-5 and IL-10) and no clear granuloma formation. Consequently, M. leprae are able to proliferate and disseminate more freely in lepromatous leprosy, and the patient suffers the systemic effects

101

of the infection. Host response in the form of hypersensitivity may therefore be advantageous under certain circumstances, and modification of the response by the administration of exogenous cytokines (e.g. interferon-y) can change the pattern of disease. 1

IMMUNOSUPPRESSION The rapid development of organ transplantation has been dependent on the availability of effective immunosuppressant drugs. An ideal immunosuppressant would be antigen specific and free of side-effects, but none of the currently available drugs meets these criteria. The clinical use and relative merits of the major immunosuppressant drugs are outlined below, as are some newly emerging immunomodulatory strategies. Corticosteroids These drugs have been widely used in the treatment of autoimmune diseases, malignancy and graft rejection for many years. The immunosuppressant action is difficult to separate from other potent anti-inflammatory actions of steroids. In humans it is mediated mainly through an inhibitory action on monocytes and macrophages, with reduced phagocytic function and diminished IL-1 secretion. Although in animals there is a marked lymphocytotoxic effect, in humans this is only significant against abnormal lymphocytes, for example in lymphoproliferative disorders, and when used at very high doses. Corticosteroids are potent immunosuppressants but their major disadvantage is multiple side-effects; these include weight gain, hirsutism, hypertension, diabetes mellitus, recurrent infection, growth retardation, osteoporosis, and reduced response to physiological stress. The appropriate dose of corticosteroid is determined by the individual clinical situation. For the long-term control of autoimmune disorders an oral dose of <10mg/day prednisolone would be appropriate. In the acute management of autoimmune haemolytic anaemia, acute glomerulonephritis or immune thrombocytopenic purpura, l-2mg/kg/day is indicated. In very severe life-threatening situations, such as systemic vasculitis, pulsed high-dose regimens of l0mg/kg are used (maximum 1000mg).

CYTOTOXIC DRUGS (see Chapter 6) This group of drugs kills cells capable of self-replication and they were initially introduced in the treatment of cancer. Their introduction into the management of trans1

102

MCQ 4.4

2

Case 4.2

3

MCQ 4.5

plantation and autoimmunity enabled reductions in the doses of corticosteroids required (in this situation they are often referred to as 'steroid-sparing agents'). The major complication of long-term use is myelosuppression and immunosuppression, and a long-term increase in susceptibility to malignancies. Azathioprine Azathioprine is an analogue of the cytotoxic drug 6mercaptopurine (both are thiopurines) and was originally introduced in the 1950s for renal transplant recipients. It now has a major use as a steroid-sparing agent in the management of many autoimmune disorders. Thiopurines block the synthesis of inosinic acid, the precursor of adenylic and guanylic acids, and thus impair DNA synthesis. Their major cytotoxic effect is on developing T and NK cells, and thus there is a delay before clinical onset of action. The most important side-effect is bone marrow suppression. Weekly monitoring of full blood count (FBC) is advised during induction, with 3-monthly FBC during maintenance therapy. The patient must be warned of potential susceptibility to infection, and any unexplained illness or clinical suggestion of infection requires thorough investigation and, potentially, withdrawal of treatment. The usual maintenance dose range is l-3mg/kg daily. 2 Cyclophosphamide This is one of the most widely used drugs in cancer chemotherapy and is now frequently used in combination with corticosteroids in the management of severe autoimmune conditions, especially systemic vasculitis. As an alkylating agent it acts by cross-linking DNA strands, thereby preventing normal mitosis. It is particularly toxic to B cells. Side-effects include a dose-dependent neutropenia and lymphopenia, alopecia and infertility (especially in males) and, infrequently, haemorrhagic cystitis. Cyclophosphamide is typically used in combination with prednisolone in either oral or intravenous protocols to control severe systemic vasculitis. Dosage is determined by disease severity and according to renal function, bone marrow function and the patient's age. Mycophenolate mofetil Mycophenolate mofetil is used in combination with corticosteroids and ciclosporin in the prophylaxis of renal transplant rejection. Mycophenolic acid (MPA) is the active metabolite and is rapidly produced by hydrolysis after ingestion of the drug. MPA selectively inhibits inosine monophosphate dehydrogenase (IMPDH), which prevents lymphocyte purine biosynthesis. The inhibition of IMPDH causes a depletion of guanine nucleotides, inhibition of DNA synthesis, and prevents clonal expansion of both T and B cells. Close monitoring is necessary with weekly full blood counts (FBC) for the first month of treatment, 2-weekly for the next month, and monthly for the first year of treatment. Its use is currently limited to specialist transplantation units.

Methotrexate This drug is used in maintenance treatment of childhood leukaemias and is increasingly used in the control of a number of inflammatory diseases, including rheumatoid arthritis, polymyositis, psoriasis, Reiter's syndrome and graft-versus-host disease following bone marrow transplantation. It also appears to have a role in steroiddependent asthma, although the exact mode of action in this condition is unclear. Methotrexate inhibits dihydrofolate reductase, thus preventing the conversion of folic acid to its active form, tetrahydrofolate. This prevents normal thymidine synthesis and cell division. Side-effects include bone marrow suppression, megaloblastic anaemia, pneumonitis and mucositis. Long-term drug toxicity is increased in hepatic impairment. Acute side-effects are partially reversible by the administration of folinic acid.

NON-CYTOTOXIC IMMUNOSUPPRESSION Ciclosporin This relatively recently introduced drug has revolutionized the management of post-transplantation patients and is also finding a place in the management of some autoimmune disorders. A fungal metabolite with a cyclic undecapeptide structure, it acts to selectively inhibit T-cell activation. The drug binds to cytosolic proteins, called cyclophilins, which prevent calcium-dependent activation of the IL-2 gene. IL-2 transcription is a crucial event in T-cell activation and thus cyclosporine is a relatively selective immunosuppressant. Ciclosporin does not cause myelotoxicity but does cause a dose-dependent reduction in renal function that must be carefully monitored and differentiated from renal transplant failure or rejection. Monitoring of therapeutic plasma levels is essential. Other side-effects include hirsutism, gingival hyperplasia and seizures. In common with other immunosuppressants there is an increased long-term risk of lymphoma and skin neoplasia. Tacrolimus (FK506) This macrolide is not chemically related to ciclosporin but has a similar mode of action. Its use is currently limited to hepatic and renal transplant patients, although there is evidence to suggest it will have significant clinical impact in autoimmunity. Tacrolimus binds to the cytosolic FK506binding protein (FKBP), preventing IL-2 gene activation. It has a similar range of side-effects to ciclosporin. 3

NEWER IMMUNOSUPPRESSIVE STRATEGIES In addition to the pharmacological immunosuppressive drugs, a number of new immunological therapies are emerging in clinical practice.

Plasma exchange has been available for many years and continues to be of value in the acute management of some severe autoimmune conditions, including Goodpasture's syndrome, myasthenia gravis and rhesus disease, in which autoantibodies are demonstrated to be pathogenic. Plasma exchange allows the removal of these autoantibodies acutely but has no proven role in the long-term management of these conditions. High-dose ivIg therapy (l-2g/kg) is of proven benefit in autoimmune thrombocytopenia and Kawasaki's and Guillain-Barre syndromes. Its mode of action is incompletely understood, but may involve blockade of Fc receptors, disturbance of anti-idiotypic networks, or suppression of macrophage and/or T-cell activation. This therapy has specific side-effects, including aseptic meningitis, acute haemolysis (when ivIg is uncross-matched) and acute deterioration in renal function. It should therefore only be used for conditions in which clinical benefit is proven by properly controlled studies. Monoclonal antibody (moAb) therapy has shown encouraging results in a number of areas of clinical practice. CAMPATH-1 is a humanized mouse anti-CDw52 moAb, an antigen which is present on all leukocytes. It has proved to be of particular value in the treatment of acute graft rejection and graft-versus-host disease following bone marrow transplantation. More specifically, anti-CD25 moAb (basiliximab), which prevents T-cell activation, has been introduced for the prophylaxis of acute renal rejection. Antitumour necrosis factor (anti-TNF) moAb has recently shown promising results in the control of disabling rheumatoid arthritis (RA) and in severe inflammatory bowel disease, especially Crohn's disease. TNF is believed to have a central role in the pathogenesis of these diseases. Future clinical developments in this area may follow from current preclinical studies. 'Biological immunosuppressive agents', including antibodies directed against cell surface molecules of T, B and antigen-presenting cells, are being studied in models of transplantation and autoimmunity. These models may indicate new modalities for controlling undesirable immune responses through modulation of the function of the cell surface signalling molecules. Antibodies directed against gp39 (CD40 ligand) and CD28 have been demonstrated to overcome renal allograft rejection in a primate system. Human clinical trials (phases I and II) of these agents are currently under way. Anti-LFA-1 moAb has also been used to control transplant rejection. The cell surface molecule CTLA-4, which is known to downregulate T-cell responses, also has the potential for therapeutic manipulation. CTLA-4 linked to immunoglobulin molecules (to prolong its half-life in the circulation) has been successfully used in clinical trials to treat psoriasis. In respiratory medicine, IL-8 receptor antagonists are being evaluated in airway disease, and antiIL-5 moAb is under investigation in the treatment of asthma. These experiences of 'biological immunosuppression' suggest that this will be a major area of therapeutic development in the coming years, with novel drug devel-

4

103

opment for a wide range of disorders associated with inappropriate or excessive immune activation.

FURTHER READING Chapel H M 1994 Consensus on diagnosis and management of primary antibody deficiencies. British Medical Journal 308:581-585. Chapel H, Heaney M, Misbah S, Snowden N 1999 Essentials of clinical immunology, 4th edn. Blackwell Science, Oxford. Conley M E, Notarangelo L D, Etzioni A 1999 Diagnostic criteria for primary immunodeficiencies. Clinical Immunology 93(3):190-197.

104

Janeway C A, Travers P 1997 Immunobiology - the immune system in health and disease, 3rd edn. Churchill Livingstone, Edinburgh. Primary immunodeficiency diseases. Report of an IUIS scientific group. 1999. Clinical and Experimental Immunology 118 (suppl l):l-28. Reeves W G, Todd I 1996 Lecture notes on immunology, 3rd edn. Blackwell Science, Oxford. Rosen F S, Cooper M D, Wedgewood R J P 1995 The primary immunodeficiencies. New England Journal of Medicine 333(7):431-440. Stites D P, Terr A I, Parslow T G 1997 Medical immunology, 9th edn. Appleton and Lange, Connecticut.

5

Nutrition in Clinical Medicine Michael E J Lean

Food and its use 105

Overweight and obesity 128

Assessment of nutritional status 115

Nutrition support 132

Dietary intake assessment 117

Nutrition as a cofactor in disease and in therapy 136

Nutrition and feeding practices for growth 119

Food sensitivity and intolerance 140

Malnutrition 120

Nutrition in health promotion: current diet and proposals for change 141

Vitamin deficiencies and excesses 122 Trace elements 127

FOOD AND ITS USE FOOD CHOICE AND FOOD COMPOSITION There is no doubt that diet has profound influence on human health, at all ages and affecting all medical subspecialties. Defining and influencing nutritional status is, however, very complicated. Although we customarily summarize the diets of individuals and populations in terms of nutrient exposure, there are many steps in establishing this exposure, each subject to substantial variability, both within individuals over time and between population subgroups. The general scheme (Table 5.1) may seem mostly self-evident, but indicates the scope for error in research and the need for clarity of thinking. The source of variability which may affect the evaluation of the diet of an individual or a population extends well beyond those that can be established either by retrospective questionnaires (depending on memory) or by prospective food records (depending on representativeness). Most of these influences are beyond the control of individuals, and so may tend to confound efforts to change diets.

As well as variations within the population, many of the components in Table 5.1 are also subject to fluctuations over time. Meal and snack patterns have changed dramatically over the last 10-20 years, with 'grazing' now being the norm in young people. Likewise, a generation has now emerged without experience of traditional meals or home cooking, using almost exclusively pre-prepared foods eaten in 'non-meal' situations - e.g. in front of the television, or on public transport. Increasingly homes do not have conventional kitchens, and people rely on microwaves and freezers. Although daily food shopping is still usual in Mediterranean countries, where fresh foods predominate, the steady trend in UK is towards weekly shopping, and this has profound effects on the need for long-life foods and preservatives. Retailers have progressively selected smaller ranges of foods, for shelf-life, appearance and taste, but disregarding nutrient composition. The dominance of major multiples (supermarkets) has significantly reduced biodiversity and driven farmers to produce highly stylized products. The 'added-value' food processing industry is now the major profit source, such that the most heavily advertised and promoted foods are those with the poorest nutritional value. Individual food choice is very complicated, but essentially demands a mix of the known and safe with a little of something new. Over recent decades there have been progressive trends to increase the sweetness and saltiness of foods. This has created a taste expectation, such that home-prepared foods can taste unpleasant if they contain less (i.e. normal amounts) sugar or salt. On a more positive note, the quest for 'health' by consumers is becoming a major driver for food choice and for reformulation of products. Globalization and sophisticated trading and transport have introduced unfamiliar foods, but also a new layer of uniformity which makes change difficult. The influence of season has all but disappeared, to the extent that many people are suspicious of seasonal foods - even locally grown ones, such as soft fruit. The nutrient composition of foods, catalogued in food tables and databases, is drawn up from analyses of a few examples of each food. One or two (e.g. potatoes) take account of seasonal variations in composition, but not the other factors outlined above. These tables provide only a broad guide as to the nutrient composition of the foods consumed by an individual or in a particular survey. Dietary analysis is imprecise, with scope for serious error. For some foods, similar varieties can contain up to 100-fold differences in specific nutrients, e.g. flavonols in lettuces. One way to resolve this is to assess diets, and to construct dietary advice, in terms of food groups consisting of broadly interchangeable foods - bread/cereal/pasta/rice; fruit and vegetables; spreading fats and oils; meat/cheese/ eggs; sweets, cakes and confectionery. This type of grouping has some limitations, but serves to remind us that diets, and foods, are consumed in meals. The meal is the smallest unit of nutrition, and it is the 'eating episode' which provides a focus for individual planning.

a

105

TABLE 5.1 Influences on diet composition and nutrient exposure with food choice and variability Diet composition

Source of variability

Meal and snack pattern Meal structure Cooking methods Home food storage methods Shopping and catering patterns Food choice of individuals Purchasing policies of retailers Farming methods Seasonal

Age, ethnicity, culture Culture, occupation, family Family structure, housing Housing, finance Socioeconomic, ethnic Age, gender, socioeconomic Profit-driven, convenience Profit-driven, subsidies Latitude, transport Analytical rigour

Nutrient composition of food groups Nutrient composition of specific food varieties

Plant breeding, health concerns

NUTRITION AND NUTRIENTS Nutrition is the process by which food is eaten and utilized by the body. More broadly, the principles of nutrition nutrient supply, storage, requirements, consumption and metabolism - can be applied at many levels. Nutritional balance, which incorporates all these factors, should therefore be considered at the levels of the cell, of organs, of the whole body and of populations. For all the processes that define nutritional balance, there are ranges within which balance can be achieved at a level compatible with health. Outside these ranges malfunction and disease will result from deficiencies, toxicity or competition between nutrients. Within 'normal ranges' of nutrition, research is beginning to define levels which may determine optimal long-term health, and which may differ between individuals according to genetic background. In the presence of disease, nutrient requirements, supplies and metabolism are altered. Nutritional status may be a cause, or part of a cause, of disease, and may in turn be altered by disease. This chapter is concerned with the ways in which nutrient supplies - especially energy - are utilized, and how dietary factors and deficiencies may produce specific clinical syndromes or contribute to the development of disease.

DIETARY REQUIREMENTS AND REFERENCE VALUES

106

Defining the nutritional requirements of individuals, or of groups or populations, is one of the most important aims of nutritional science. Information of this kind is essential for designing food supplies, and therefore it has major economic and political implications.

Certain nutrients are essential, i.e. they are necessary components or precursors of human biological pathways, that cannot be synthesized by humans. Essential nutrients include energy (provided by carbohydrates, fat and protein, together with oxygen and water), essential amino acids and fatty acids, and vitamins and essential trace elements. Foods also contain large numbers of other biologically active compounds which can modify metabolism and function. These are not 'essential', but may have beneficial actions for current or long-term health. Finally, there are compounds in foods which are not nutrients. These include environmental contaminants and inert or non-absorbable compounds. They may still affect health through toxic or allergic mechanisms, so the quantity present in foods may be important. Living cells require a continuous balanced supply of nutrients. However, food and nutrient supply is intermittent, and the body must also accommodate periods of nutrient deprivation, whether through famine or illness. The relationship between storage and tissue requirements determines survival time during nutrient deprivation. For oxygen (with only limited local capacity for anaerobic respiration and no storage), survival is measured in minutes. For carbohydrates and water, a clinical problem may occur in hours or a few days. For most water-soluble vitamins and some minerals (e.g. zinc), nutrient status also declines in days of deprivation. In contrast, in the case of the fatsoluble vitamin B12 and iron, full stores can permit many weeks of dietary deprivation. Traditionally, dietary nutrient requirements have been established for the essential nutrients on the basis of the amount per day which is sufficient to prevent a clinical deficiency syndrome developing, with a margin between 30 and 100% to produce a recommended daily amount, or RDA. This safety margin will depend on the nature of the nutrient. Fat-soluble vitamins, which tend to accumulate in the body and can be toxic, have a lower safety margin than water-soluble vitamins, which are rapidly excreted if consumed in excess. For many nutrients, deficiency syndromes develop slowly and isolated deficiencies are rare, so it has been hard to define RDAs. The 'safe intake' is often quoted. These are intakes which are customarily consumed by apparently healthy people. However, there are variations within the range of intakes in the general population. At a cellular level, most nutrients required for normal metabolism are synthesized. In some situations - in illness when supply is limited or demand is increased - there may be a conditioned deficiency of 'non-essential' nutrients which take on the role of 'conditioned essentials'. It appears that, for some nutrients (notably the antioxidant vitamins), there are levels of intake, well above those needed to prevent deficiencies, which are associated with optimal long-term health. Higher intakes than the RDA of vitamin E, for example, appear to have some protective effect against heart disease and some cancers. The RDA for folate, sufficient to prevent folate deficiency, is insufficient to prevent neural tube defects in susceptible women. Furthermore, within the recommended or 'normal' range

a

FIG. 5.1 Dietary reference values - definitions 95% of people would be adequately nourished if they consumed the RNI, and 95% would be deficient at LRNI.

FIG. 5.2 Dietary intakes and risk of deficiency with and without adaptation The risk of clinical deficiency is reduced when there is adaptation to low intake as it falls below the LRNI.

TABLE 5,2 Criteria for defining dietary reference values of folate intakes there is an inverse correlation with plasma homocysteine, a risk factor for heart disease. For most vitamins and minerals the average intakes in western countries are well above the RDAs, but some - e.g. iron, thiamine, folate, vitamin D, vitamin E - are marginal in certain population groups In 1991, the UK Committee on Medical Aspects of Food Policy, published a new approach to dietary reference values for a total of 40 nutrients. The reference nutrient intake (RNI) is defined as 2SD (standard deviations) above the estimated average requirement (EAR), and the lower reference nutrient intake (LRNI) as 2 SD below it (Fig. 5.1). Thus 95% of people would be adequately provided for by consuming the RNI and, assuming no adaptation to low intakes, 95% would be deficient if they consumed the LRNI (Fig. 5.2). The exception to this general rule is the energy requirement. Assuming the need to maintain a constant weight, only an EAR is quoted. The evidence used included not only the classic clinical deficiency evidence, but also data on blood or tissue saturation and on nutrient-dependent enzyme function (Table 5.2). For a number of nutrients adequate scientific data are still not available, and so 'safe intakes' are quoted. The questions surrounding optimal intakes for future health are still unresolved, apart from fatty acids, which follow the recommendations for cardiovascular health. Dietary reference values (DRVs) are determined for healthy people, and assume normal status for other nutrients. The factors affecting requirements are summarized in Table 5.3.

• • • • •

Amount taken without deficiency developing Amount needed to cure deficiency Amount needed to maintain blood or tissue concentration Amount needed to maintain enzyme saturation Amounts associated with appropriate biological marker of adequacy

TABLE 5.3 Factors which affect requirements of nutrients 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Size Age Growth Sex Physical activity Pregnancy Lactation Availability from specific foods Illness Deficiencies of other nutrients

ENERGY Energy, utilized over time to produce work or build body tissues, is expressed as kcal or kJ (1 kcal = 4.18kJ). It is supplied by fat (9kcal/g), ethanol (7kcal/g), protein (4kcal/g),

107

carbohydrates (3.75kcal/g) and dietary fibre (2-3kcal/g). Because energy cannot be created or destroyed, energy intake always equals energy output, plus or minus changes in energy stores. The body stores some energy as carbohydrates, in the form of glycogen (mainly in the liver), which provides a source of glucose. Total glycogen stores contain about 250 g, or 1000 kcal. The remaining energy stores are as triglycerides in adipose tissue. An average person with 20kg of adipose tissue has energy stores of 20 x 7000 = 140 000 kcal, enough to provide for 70 days' starvation at 2000 kcal/day. Carbohydrate cannot be synthesized from fatty acids. There are no specific stores of protein, but during starvation protein is metabolized from the destruction of muscle and connective tissue. This, and the glycerol from triglycerides, provides some glucose for organs dependent on glucose. Energy is necessary for the maintenance of all metabolic processes, physical activity and growth. Most dietary energy is needed to maintain oxygenation, blood supply and ionic gradients across membranes in all tissues. This is termed the basal metabolic rate (BMR). Energy expenditure can be measured by indirect calorimetry, usually in the early morning in a resting state, the subject having last eaten the previous evening. This resting energy expenditure (REE) is close to the BMR. BMR can be estimated with reasonable accuracy from the sex, age and body weight (Table 5.4). More recent 'doubly labelled water' methods measure energy expenditure in free-living adults using the different excretion rates of the stable isotopes deuterium (3H) as water, and 18O as water and CO2, to obtain an estimate of CO2 production and hence total energy expenditure (TEE). The difference between TEE and BMR measures the energy expenditure in association with feeding and from physical activity (Table 5.5). The energy requirement of an individual subject, which is equal to TEE if weight is to remain constant, depends primarily on body size and composition. Age, physical activity, environmental temperature and physiological state (whether pregnant, lactating or suffering from sys-

SUMMARY 1 Factors affecting basal metabolic rate and total energy expenditure Factors affecting basal metabolic rate • body weight • age/gender

relative proportion of lean body mass to adipose tissue recent weight change or energy imbalance

• thyroid function

108

Factors affecting total energy expenditure • basal metabolic rate • physical activity and body weight (work output) • diet composition (carbohydrate, protein, fat proportions) • inflammatory conditions, altitude

TABLE 5.4 Equations for estimating basal metabolic rate (BMR) BMR (MJ/d)

Males

10-17 yr 18-29 yr 30-59 yr >60yr

BMR = 0.074 W + 2.754 BMR = 0.063 W + 2.896 BMR = 0.048 W + 3.653 BMR = 0.049 W + 2.459

SEE = 0.44 SEE = 0.64 SEE = 0.70

Females

10-17 yr 18-29 yr 30-59 yr >60yr

BMR = 0.056 W + 2.898 BMR = 0.062 W + 2.036 BMR = 0.034 W + 3.538 BMR = 0.038 W + 2.755

SEE = 0.47 SEE = 0.50 SEE = 0.47

W = body weight (kg), SEE = standard error of estimate

TABLE 5.5 Calculated physical activity level (PAL) of adults at three levels each of occupational and non-occupational activity Occupational activity Light

Non-occupational activity Non-active Moderately active Very active

Moderate

Moderate/Heavy

Male

Female

Male Female Male

1.4 1.5 1.6

1.4

1.6

1.5

1.5

1.7 1.8

1.6 1.7

1.6

1.7 1.8 1.9

Female

1.5 1.6 1.7

temic infection) are also important. BMR is determined by the size and metabolic activity of organs in a resting state, and so it has a major component from organs such as the heart, lungs, kidneys and brain. Skeletal muscle is relatively less important in determining BMR than it does in TEE. TEE has a very close relationship to body weight because the energy cost of physical activity depends closely on weight. Variations in BMR for a given weight can reflect metabolic factors such as thyroid function. Energy expenditure may be considerably increased in situations of metabolic stress, such as pyrexia and burns. The factors that control metabolism physiologically and modify it in disease include catecholamines, thyroid hormones, insulin, cortisol and cytokines (e.g. tumour necrosis factor-a). Energy requirements are highest - on a body weight basis or per unit 'lean body mass' - in infants and young children. Energy expenditure starts to fall in the 20s and declines steadily thereafter into old age, principally because of declining physical activity, with consequent reduction in muscle mass. Because the factors affecting energy requirement vary considerably between individu-

TABLE 5.6 Average daily energy expenditure In various lifestyles

a

24-hour energy expenditure in kcal (MJ) Occupation (age in years)

Men

Retired elderly (65+) Housewife (25-45) Office worker Laboratory technician (20-50) Factory worker (20-45) (heavy industry) Labourer (20-45) (heavy physical work)

2300 2500 2800 3200 3600

Women (9.5) 1900 2200 (10.4) 2100 (11.7) 2200 (13.2) 2400 (14.8)

(7.8) (9.0) (8.6) (9.0) (9.9)

TABLE 5.7 Average energy expenditure of an adult in various degrees of physical activity Degree of activity

Expenditure (kcal/min)

Minimal (e.g. sitting, slow walking) Moderate (e.g. walking, gardening) Intermediate (e.g. swimming, brisk walking) Heavy (e.g. digging, squash, running)

1 5 7 10

als, it is impossible to give an exact ideal nutrient intake, but there are clearly differences in daily energy expenditure between adults with different lifestyles (Table 5.6) and at different ages (Fig. 5.3). The energy expenditure for physical activity depends on body size, so it is convenient to express the energy costs of different activities as functions of BMR. Table 5.7 shows the energy required for various physical activities. As the total time in each 24 hours spent in strenuous activity is often rather short, significant increases in energy expenditure are difficult to sustain. Weight reduction by increasing physical activity is therefore hard to achieve. In adults the main indicator of energy balance (and often of overall nutritional status) is change in body weight. Weight varies by up to 1-2 kg from day to day in healthy adults, and is more affected by fluid balance over the short term. Weight change due to altered energy balance is usually accounted for by fat stores, at 9000kcal/kg (75% of weight change), and lean body mass, at l000kcal/kg (25%). Thus each 1kg change in adipose tissue reflects about 7000 kcal shift in energy balance. Although metabolic rate is governed principally by body weight, the alterations in metabolic rate from weight changes are relatively small: a 1kg weight change alters metabolic rate (and thus energy requirement for weight stability) by only about 15kcal/day. In children energy is required for growth, and measurements of weight gain or height gain are useful indicators

FIG. 5.3 Estimated average requirements for energy intake [A] at different ages, and B during pregnancy and lactation for women.

of dietary insufficiency. During the first few weeks of catchup growth from severe malnutrition, the energy cost of weight gain is about 5kcal/g. Thereafter, it becomes more energy expensive to lay down tissue, as fat rather than lean tissue is laid down. Energy balance can be maintained from a wide range of nutrient origins. In the UK, fat and carbohydrate each provide between 40 and 50% of dietary energy, whereas in developing countries the proportion from fat may be less than 15% and from carbohydrates more than 75%. Protein provides 10-15% of energy in most diets, in both western and developing countries.

109

Dietary energy deficiency is that amount which results in negative energy balance. This results in immediate mobilization of energy stores from glycogen and fat. As underfeeding continues a greater proportion comes from protein, principally through muscle atrophy. Excess energy intake will be stored as glycogen if the limited capacity of the liver permits, or as fat in adipose tissue. Gross excess will lend to storage in tissues such as liver, impairing liver function. Minor imbalances of energy balance occur between meals and from day to day. Energy imbalances insufficient to be detected in body weight will have metabolic effects, e.g. on insulin sensitivity and serum lipoprotein.1

PROTEIN Most major human foods include protein, in an amount equivalent to 5-50% of food energy (Table 5.8). Mixed diets usually supply 10-20% of energy as protein. The requirement for dietary protein is determined by tissue growth and repair. The healthy body is efficient in the utilization of amino acids. Total body protein in an adult is about 16kg, which has accrued over 15-20 years of growth. Healthy protein is constantly being broken down and resynthesized in the process of protein turnover, but the amino acids are mostly reutilized. Losses from skin, hair and gut mucosa contribute to an absolute requirement of about 0.6g/kg/day for adults, or about 8% of dietary energy. Although humans cannot incorporate inorganic nitrogen, colonic bacteria can. It is therefore possible for some nitrogenous waste in the bowel to be rescued by bacterial amino acid synthesis and subsequent absorption. There are no body stores of protein or amino acids, but periods of low intake can be tolerated by metabolism of tissues such as muscle, gut mucosa and skin. However, atrophy of these tissues is more likely to reflect total energy deficiency, not specifically protein. The maximal protein requirement for optimal growth in infancy is about 1.5g/kg/day. This is amply provided by human milk, whose protein content is 130g/L, 7.5% energy.

SUMMARY 2 Protein requirements in health • • • •

Protein deficiency impossible with normal food provision Most dietary protein is metabolized like carbohydrate Minimum safe consumption (about 8% dietary energy) Requirements for essential amino acids virtually always met by 1 g/kg/d (= 10% dietary energy)

1

110

MCQ 5.1

TABLE 5.8 The protein content of some foods which are often major protein sources Protein source Cassava Cooked bananas (plantains) Sweet potatoes Human milk (4.1 % fat) Rice (brown or white) Millet flour Potatoes (old, main crop) Maize (sweetcorn kernels) Wheat flour (wholemeal) Groundnuts (peanuts) Cows' milk, full fat (3.9% fat) Beans and peas Lentils

Eggs Cows' milk (skimmed) Fish (fatty) e.g. mackerel Soya (tofu) Meat (lean) (beef) Fish (white)

Proportion of energy as protein (%) 2 3 5 7 7 7 11 15 16 18 19 31 34 34 40 42 45 50 94

It is assumed that human milk contains the amino acids required for growth in optimal proportions, and this is likely to be similar for other milks, and eggs. There are nine amino acids which are not synthesized in the body, and are thus 'essential' nutrients. Others are needed in the diet to optimize utilization, although they are not technically essential. The ability to synthesize specific amino acids may be compromised during illness, such that 'conditioned essentiality' may develop, e.g. for glycine in infants. The metabolic requirement for amino acids is a separate consideration from dietary requirement. Essential amino acids thus form a small part of metabolic requirement, whereas non-essential amino acids (normally readily synthesized in the body) may be required in large amounts. To some extent alterations in colonic bacterial amino acid synthesis may modify essentiality. Because certain foods are deficient in essential amino acids (cereals, for example, are deficient in tryptophan, methionine and cysteine) an optimal amino acid balance can only be achieved by combination with other foods (such as legumes, which are rich sources of tryptophan and methionine). Thus a diet based on maize and beans is consumed by millions of healthy people all over the world. Most mixed diets contain much more protein than is required for health, and the excess is metabolized, mainly as carbohydrates, allowing the correct balance of amino acids to be extracted for protein synthesis. Protein is absorbed in the small intestine as amino acids and small

peptides, after the action of luminal peptidases such as trypsin. Isolated dietary protein deficiency is extremely rare outside famine situations. Low-protein foods such as cassava or plantains are not otherwise eaten in isolation. Even when intakes of ethanol or fat and sugar are extreme, protein requirements in western societies are usually met. There is considerable debate as to how much protein we require. The usual calculation, which has been applied to all nutritional states, is expressed as nitrogen balance - the difference between the dietary intake and the losses of nitrogen in the urine, faeces and sweat. A positive balance is achieved when intake exceeds losses, such as occurs in normal infant and child growth, or during recovery from malnutrition in adulthood. Negative balance occurs when losses exceed intake, as in systemic infection or diarrhoea, or when dietary intake is reduced. The greatest requirements for protein (relative to body weight) occur in young infants. Those less than 3 months of age require about 2.5 g protein/kg body weight/day, and the older infant (9-12 months of age) requires 1.5g protein/kg body weight/day. Children need less than infants but more than adults. Rates of protein synthesis vary with age, being highest in the newborn and declining with time (Table 5.9). Unfortunately, nitrogen balance measurements are difficult to conduct, particularly in the ill or deprived. The main problem is that they depend on having a complete 24-hour urine collection to estimate output. If urine collections are incomplete, as is very common, the results are biased towards an apparently positive nitrogen balance. Thus, improbable claims of positive nitrogen balance have even been made for subjects on very low-calorie diets and evidently losing weight. Nitrogen balance in adults to some extent changes with the customary diet. There is adaptation to low intakes, so that nitrogen balance has apparently been achieved with 0.2g/kg/day after several weeks. However, it is not certain that this is indefinitely sustainable in health. During illness, both protein synthesis and protein degradation are increased, so that total protein turnover is elevated non-specifically and irrespective of whether total body protein content changes. In general, there is net protein loss during disease, with wasting of muscle, and the losses are restored during recovery. The 'ebb and flow'

TABLE 5.9 Variation in rate of protein synthesis with age Age group

Body weight (kg)

Newborn (premature) 2.0 Infant 9.0 Young adult 70.0 Elderly 55.0

Total body protein synthesis (g/kg body weight/day) 15.0 7.0 3.0 2.0

phases of energy and protein balance after metabolic stress relate to the release of metabolically active compounds such as cytokines, and cannot be corrected by increasing the supply of amino acids or dietary protein. There is no simple biochemical test to estimate amino acid or protein status. Plasma amino acids and proteins all reflect stress non-specifically. Serum albumin falls with any stress or inflammatory condition - e.g. myocardial infarction, malignancy or infection - and is not a marker of nutritional status. Serum albumin may, however, be of some value in distinguishing disease-related cachexia from starvation. Low albumin in the presence of starvation is a worrying prognostic sign, usually indicating the presence of infection. This is the case in oedematous malnutrition (kwashiorkor).

a

CARBOHYDRATES There are three major groups of carbohydrates in the diet: sugars, starches, and the plant wall materials that comprise 'dietary fibre'. All carbohydrates are polymers of monosaccharides, with the common formula (CH2O)n. They are almost completely absorbed from the gut, either as monosaccharides in the small intestine after hydrolysis by amylases and mucosal disaccharidases, or (5-10%) as short-chain fatty acids following fermentation by large bowel bacteria. The main absorbed sugars, glucose and fructose, are ultimately metabolized in the Krebs cycle to release energy through oxidation. Each carbon atom combines with two oxygen atoms to produce H2O and CO2 and release 3.75 kcal/g. Of relevance to the management of diabetes and hyperlipidaemia is the 'glycaemic index' of foods. This is the incremental rise (i.e. subtracting this baseline value) in blood glucose over 3 hours after eating 50 g carbohydrate from a food, expressed as % of the rise after 50 g carbohydrate from bread, which is used as a standard. It is possible to predict the glycaemic index of a whole meal from those of its component parts. Low glycaemic index foods, reflecting the content of fat and dietary fibre, consistency, cooking or processing and other undefined factors, are associated with lower blood glucose and insulin (i.e. increased insulin sensitivity) and with improved serum lipoprotein profiles. • Sugars. Sucrose and its components glucose and fructose are the sugars most used as sweeteners. Lactose (glucose-galactose) is the main sugar in milk. Fructose is the sweetest. The monosaccharides glucose, fructose and galactose are readily absorbed at the jejunal surface. The disaccharides sucrose, maltose and lactose are equally well absorbed as the monosaccharide component parts, provided that there has been effective hydrolysis by the intestinal disaccharidases. Sucrose intolerance and lactose intolerance, leading to watery diarrhoea, may occur during starvation or any diarrhoeal disease, as a result of mucosal atrophy and secondary

111

disaccharidase deficiency. This is a strong argument for maintaining gut function by enteral feeding in postoperative and intensive care patients. • Starches. These comprise a group of polysaccharides containing large numbers of glucose units linked together to form branched chains. They are variably hydrolysed as a result of cooking, and are then further hydrolysed within the intestine. The commonest are the dextrins amylose and amylopectin. About 90% of starch is rapidly hydrolysed and absorbed in the small intestine. The remainder (some altered by processing or cooking to become hydrolysis resistant) passes with non-starch polysaccharides to the colon. • 'Dietary fibre' - non-starch polysaccharides. Plant walls contain a range of lignins, celluloses and hemicelluloses that are not digested in the intestine, and which are variably degraded by intestinal bacteria in the colon. Dietary fibre can be measured in a variety of ways in foods, but its function in maintaining healthy colonic flora - an essential function - can be performed differently by the different subclasses, and by other compounds, such as lactose in breastfed infants. Different dietary fibres have important physiological functions in influencing the absorption of sugars and the regulation of serum lipoproteins (soluble and gum-forming fibre from vegetables, fruits, legumes and oats) and in contributing to intestinal bulk and large bowel health (cereal fibre such as bran). Dietary reference values for carbohydrate are shown in Table 5.10. 'Extrinsic' sugars are those outside intact cells in foods. Several varieties of non-sugar sweetener exist. Sorbitol is a sugar alcohol; however, its energy value is similar to that of other carbohydrates. Saccharine has no chemical or nutritional relationship to sugars and provides no energy; it is about 500 times as sweet as sucrose. Aspartame is an amino acid-based non-nutritive sweetener. EU regulations control the amounts of sweeteners that are permitted in foods. None have been shown to be hazardous to humans, but the recommended limits can be exceeded by individuals with unusually repetitive food habits, and by diabetic children who consume large amounts of highly sweetened foods. Very high consumers should vary the types of artificial sweeteners used.

FAT

112

Fatty acids provide the chief store of energy in the body. They are important in the development of all cell membranes and some tissues, and are essential components of some hormones. Dietary fatty acids are mainly supplied as triglycerides. These are esters of glycerol and three fatty acids. Fatty acids are characterized by the chain length, expressed as the number of carbon atoms (short chain C2-C4, medium chain C6-C12, long chain C20-C24), and by the number and position of double bonds (one double

TABLE 5.10 Dietary reference values as percentage of total energy* Population average Saturated fatty acids cis-polyunsaturated fatty acids of which essential fatty acids

10(11)

cis-monounsaturated fatty acids frans-fatty acids Total fatty acids Total fat Non-milk extrinsic sugars Intrinsic and milk sugars and starch Total carbohydrate

12(13)

33 (35)

Non-starch polysaccharide (g/d)

18

6 (6.5) 3(3)

2(2)

30 (32.5)

10(11)

37 (39) 47 (50)

*The average percentage contribution to total energy does not total 100% because figures for protein and alcohol are excluded. Protein intakes average 15% of total energy, which is above the RNI. It is recognized that many individuals will derive some energy from alcohol, and this has been assumed to average 5%, approximating to current intakes. However the Committee on Medical Aspects of Food (COMA) allowed that some groups might not drink alcohol and that for some purposes nutrient intakes as a proportion of food energy (without alcohol) might be useful. Therefore average figurers are given as percentages both of total energy and, in parentheses, of food energy.

bond monounsaturated, two or more polyunsaturated). Most unsaturated fatty acids are cis isomers, but some, mainly synthetically hydrogenated, are trans isomers. Vegetable oils have a greater proportion of polyunsaturated fats and are more fluid than the more saturated fats from animal sources, which are often hard at room temperature (Fig. 5.4) although liquid at body temperatures. Fish oils need to remain liquid at temperatures down to freezing point, and so contain large amounts of long-chain n-3 fatty acids. Dietary fats are normally almost 100% absorbed as they pass through the small intestine, after emulsification through the action of bile salts, and de-esterification of triglyceride by pancreatic lipase. Most dietary triglyceride is absorbed as monoglyceride and free fatty acids. The triglyceride is thus reassembled by re-esterification in the mucosal cell, and secreted by the mucosal cell with the addition of a lipoprotein as chylomicron. Most reaches the circulation via the chyle and the thoracic duct. Some, particularly short-chain fatty acids, are preferentially released discretely into the portal vein. Dietary reference values are included in Table 5.9. Phospholipids - esters of glycerol with two fatty acids are important for the stability of cell membranes, and deficiency may affect satisfactory psychomotor development. Cholesterol is the most important sterol. It is the precursor of the bile acids (which are essential for mixed micelle formation in the intestinal absorption of fat) as well as of adrenal and sex hormones and vitamin D.

FIG. 5.4 Fatty acid composition of some major sources of saturated, mono, and polyunsaturated fatty acids

of carotenoids (the precursors of vitamin A found in vegetables) and other fat-soluble vitamins. The position is more serious when there is intestinal malabsorption of fat (Chapter 15), as fat-soluble micronutrients are excreted in the faeces, dissolved in the malabsorbed fat. In this situation the dietary fat requirement to maintain dietary needs is greatly increased and may need massive supplemental dosages of essential nutrients. Fats provide over twice as much energy per unit weight as protein or carbohydrate, as each carbon atom of the (CH2) chain combines with three oxygen atoms to produce CO2 and H2O, and 9cal/g. Fats are thus useful sources of energy in diets designed to increase the rate of weight gain, e.g. in malnourished children. Conversely, a reduction in the fat content of the diet is a relatively efficient way to lose weight and to maintain weight loss. The intestinal lipase inhibitor drug orlistat utilizes this principle by introducing malabsorption of 15% of dietary triglyceride. With a low-fat diet (<30% energy) this degree of malabsorption is usually asymptomatic. Long-term use permits maintenance of 12-15% weight loss in the 40% of patients who respond to this drug. These findings remind us that patients may lose weight as a result of asymptomatic fat malabsorption from conditions such as coeliac disease or chronic pancreatitis.

a

ALCOHOL Lipoproteins are the chemical complexes, formed by the triglycerides, phospholipids and cholesterol with apoproteins, that circulate in the blood and which transport cholesterol and other fat-soluble compounds such as tocopherol. Lipoproteins can be separated by electrophoresis or according to density on ultracentrifugation, to demonstrate the classes (chylomicrons, VLDL, LDL, HDL) and their subtractions, which are familiar through their roles in atheroma formation. Triglycerides enter the bloodstream via the lymph, as chylomicra, and are converted in the liver muscle and adipose tissue to VLDL and LDL (see Ch. 17). The body is efficient at synthesizing triglycerides, sterols and phospholipids. Many populations whose diet provides only 15% of total energy as fat are able to lay down good stores of subcutaneous fat, reflecting the high capacity for synthesis of fatty acids from carbohydrate (lipogenesis) in liver, muscle and adipose tissue. Lipogenesis occurs whenever a meal provides more carbohydrate than can be stored immediately, but net lipogenesis is relatively unusual with the relatively low carbohydrate intakes in western diets. Most fat storage is from fat. Requirements for essential fatty acids are rather low (under 3% total energy) and deficiency is therefore uncommon. When diets are low in fat the high carbohydrate and protein content satisfies appetite more effectively. This may limit dietary intake and contribute to weight loss and the thinness of populations with lowfat diets. Low-fat diets may also limit the availability

There is no physiological requirement for dietary alcohol (ethanol). However, most adults have sufficient enzymatic capacity to absorb and metabolize much larger amounts than would have been met during evolutionary adaptation. Alcohol contributes 7kcal/g and tends not to satisfy appetite or to stimulate its own oxidation, and so energy from alcohol tends to be additive to the energy balance of the other proximate nutrients (fat, carbohydrate, protein). Alcohol has obvious immediate effects on the central nervous syndrome: it is a diuretic; in excess, alcohol alters lipoprotein metabolism to produce hypertriglyceridaemia; it inhibits folate absorption; and, in excess, supplants other energy sources, leading to multiple micronutrient deficiencies. The most important is thiamine (vitamin B1) deficiency. This constitutes the only true nutritional emergency. Without thiamine, disordered carbohydrate metabolism leads to Wernicke's encephalopathy and Korsakoffs psychosis (Chapter 24 p 1439). Alcoholics are also prone to scurvy because of neglect of their diet.

MICRONUTRIENTS There are large numbers of bioactive compounds in foods. A small number of these have been found to be essential for humans, who lack the ability to synthesize them. This is the basis for the definition of vitamins and essential trace elements. It is not known why nutrition has evolved in this

113

CASE STUDY 5.1 ALCOHOLIC DIETARY DEFICIENCY A 42-year-old professional violinist is admitted following a witnessed grand mal seizure in a hotel bar following a performance. He gives a history of some weight loss, with anorexia and diarrhoea for 10 days. He is a smoker and admits to intermittent heavy drinking, including over the last months. There is no previous history of seizures, but some altered sensation in the feet for several months. A CT scan is reported as normal. A relative expresses concern over excessive and prolonged drinking since the death of his father 18 months previously, related to impaired performance and loss of employment.

Questions 1. What biochemical abnormality could explain the seizure? 2, What emergency nutritional therapy should be initiated on admission?

3. What is the cause of altered sensation in the feet? 4. What long-term management is appropriate?

Discussion The seizure threshold is lowered in alcoholics for several reasons, including head injuries, B vitamin deficiencies and cerebral atrophy, presumed to be a toxic effect of ethanol. In this case 10 days' diarrhoea is likely to have depleted magnesium status as a cause of seizures. There are no body stores of magnesium, which is plentiful in many foods, particularly cereals. Serum magnesium gives a good indication of magnesium status. Low levels block the secretion of parathyroid hormone and thus cause hypocalcaemia. Thiamine replacement should be

way. Humans have the genetic potential to synthesize ascorbic acid (in common with most other species), but the genes are not expressed. Certain micronutrients, e.g. metals such as iron, or fat-soluble vitamins, can be toxic in excess, and their absorption and transport are highly regulated. The absorption of metallic trace elements is generally low (15-30%), but can increase through adaptation to low intakes, or when requirements increase in pregnancy. Dietary reference values for mineral micronutrients are shown in Figures 5.5a and b and 5.6a and b, and 'safe intakes' for those whose true requirements are not known in Table 5.11. Deficiencies of individual vitamins result in specific biochemical and clinical responses, but vitamins are also involved in metabolism in more general ways. Folate, for example, is necessary for cell division and the transfer of

1

114

MCQ5.2

started without waiting for biochemical confirmation of deficiency. Thiamine is particularly rich in cereals and grains, but not in alcoholic drinks. Peripheral neuropathy is a feature of several B vitamin deficiencies, but also of ethanol toxicity with inadequate antioxidant status, and of secondary diabetes when alcoholinduced pancreatic damage is severe. It seldom responds well to B vitamin replacement. Alcoholics are prone to multiple micronutrient deficiencies. Weight loss, often reflecting loss of lean body mass, indicates reduced total food consumption and may point to other chronic complications, e.g. tuberculosis. A multivitamin prescription is appropriate, but most preparations contain insufficient thiamine, and virtually none contain adequate folic acid. Alcoholics may benefit from greater intakes of antioxidant vitamins (C and E) than healthy people, but evidence is lacking.

methyl groups in protein synthesis; thus, a deficiency may impair tissue repair or growth, even if dietary energy and protein are adequate. Thiamine is important in carbohydrate metabolism, and lactic acidosis may occur in deficiency states and result in the brain damage of Korsakoff s psychosis. Ascorbic acid deficiency leads to scurvy, but it is also one of the hierarchy of antioxidant defences, so low vitamin C status reduces defences against free-radical damage (e.g. from smoking) without any signs of scurvy. A number of trace elements, such as zinc, are important in protein metabolism; deficiency of zinc is associated with an increased energy requirement for deposition of lean tissue. Children receiving nutrient supplements for proteinenergy malnutrition grow more slowly relative to the number of calories given, if they are not also supplemented with zinc. Severe individual micronutrient deficiencies may have recognizable clinical features, but even marginal deficiencies may affect metabolism of protein and energy, and the immune system. A variety of non-essential bioactive nutrients are relevant to human health. For example, flavonoids in fruits and vegetables are potent antioxidants, and may contribute to protection against coronary heart disease and cancers.

a

FIG. 5.5 Reference nutrient intakes for consumption of minerals These amounts will be sufficient for the needs of 95% of people.

Some related flavones in legumes are oestrogenic, and may modify the risks of heart disease, osteoporosis and breast cancer. Dietary carnitine may become important in facilitating fatty acid oxidation if carnitine synthesis is impaired during illness. Caffeine and theobromine in coffee, tea and chocolate have several properties, including central nervous system stimulation and diuretic actions.1 NON-NUTRIENT DIETARY COMPONENTS Within foods there are some compounds which are either not absorbed (e.g. the fat substitute olestra) or which are absorbed but not metabolized (e.g. the sugar inulin). Occasionally the presence of such compounds leads to an important difference between the chemical analysis and the nutritional value of a food. It is important to consider the nutrient availability of foods, as well as their chemical composition. Certain plant sterols, particularly stanol esters, inhibit the absorption of dietary cholesterol, although they themselves are not absorbed. These compounds are marked for cholesterol lowering, on average by about 10%. ASSESSMENT OF NUTRITIONAL STATUS BODY COMPOSITION In weight, the composition of the body is about 16% protein, 10-25% fat in men (15-30% in women), 5% min-

TABLE 5.11 Safe intakes of micronutrients where dietary reference values cannot be defined. Nutrient Vitamins Pantothenic acid adults infants Biotin Vitamin E men women infants Vitamin K adults infants Minerals Manganese adults infants and children Molybdenum adults infants, children and adolescents Chromium adults children and adolescents Fluoride (for infants only)

Safe intake

3-7 mg/day 1.7 mg/day 10-200 jig/day above 4 mg/day above 3 mg/day 0.4mg/g polyunsaturated fatty acids 1 ug/kg/day 10 ug/day

1.4mg (26(umol)/day 16ug(0.3umol)/day 5-400 ug/day 0.5-1.5ng/kg/day 25ng(0.5nmol)/day 0.1-1 .0ug (2.20 nmol)/kg/day 0.05mg(3umol)/kg/day

115

Functional measures, such as grip strength, are valuable to complement measures of muscle bulk. Given the wide ranges of normality, changes in repeated measures are more valuable (for example by showing falls, or a reduction in growth velocity in children). Estimation of body composition is a critical factor in human nutrition, both in clinical situations and when the nutritional status of populations is considered. A large number of indirect methods are available, varying in cost and complexity. The situation is complicated by the lack of any true 'gold standard', even for major components such as fat or muscle. For research purposes, estimates can be made from MRI, CT scanning or dual-energy X-ray absorptiometry (DEXA). These approaches can also characterize the physical distribution of tissues. Total body water (found only in non-fat tissues) can be accurately measured from dilution of deuterium-labelled water, and this is probably the most reliable guide to fat mass. Estimation of fat mass or lean body mass in clinical practice and epidemiology has to rely on well-conducted anthropometric measurements (skinfold measurements and waist circumference), either for static measurement or for detecting changes over time. These methods have been validated against underwater weighing (employing the principle of Archimedes) as an independent reference method. A variety of commercially promoted methods (e.g. bio-impedance) generally perform less well than simple anthropometry in estimating body composition.

Anthropometry

FIG. 5.6 Reference nutrient intakes for dietary vitamin consumption These amounts will satisfy the requirements of 95% of people.

116

erals, 1% carbohydrate and the rest water (about 60%). This total body water is either in the cells (intracellular water, 35%) or in the vascular system and tissues (extracellular water, 25%). The most important static measures to assess nutrition are the height, weight, body fat content (usually expressed as percentage of body weight) and fat distribution (specifically the intra-abdominal fat mass). Lean body mass (bones, muscle, brain and other organs) is estimated by difference between weight and body fat.

The key measurements for anthropometry are height and weight, measured under standard conditions (e.g. fasting, without shoes, in light clothes and with an empty bladder). Height and weight are combined in the body mass index (BMI: weight (kg)/height2(m)) as an estimate of body build or fatness which does not depend on height. The 'normal range' for adult BMI is 18.5-25 kg/m2, with BMI <18.5 indicating chronic energy undernutrition and BMI >25 usually indicating increased medical hazards from overweight. The conventional cut-off of BMI >30kg/m2 is used for epidemiological classification of obesity. The normal range of 18.5-25 kg/m2 applies at all adult ages, although the medical hazards change. These criteria do not apply to children under 16, whose BMI is usually much lower than that of adults. Mid upper arm circumference reflects both muscle mass and subcutaneous fat and has been widely used to indicate overall nutritional status. Mid upper arm circumference below 23cm in men or 22cm in women is a guide to the possibility of chronic energy undernutrition. A variety of anthropometric approaches have been used to predict body fat (and hence, by difference, lean body mass). The simplest approach, which gives a high prediction and low error through bias, is measurement of the waist circumference at the level between the lowest ribs and the iliac crest when the patient is standing.

5

TABLE 5.12 Influence of inflammation on biochemical nutritional analyses Analytes

Reference range

Minor inflammation (CRP<15mg/L)

Iron Zinc Selenium Copper Vitamin A Vitamin E Vitamin C Vitamin Eh Vitamin B2 Vitamin B6 Vitamin B12 Folate Lycopene B-carotene a-carotene Lutein

8-30umol/L 12-18n.mol/L 0.8-2.0 umol/L 11.5-23.5 umol/L 1.0-2.8 umol/L 15-40umol/L 10-115nmol/L 280-540 ng TDP/gHb <70% activation* 18-135nmol/L (PP) 188-1059 pg/mL 3.5-16.0 ng/mL 100-300 ug/L 92-312 ug/L 14-60ug/L 82-202 ng/L

Decreased by 40% Decreased by 10% Decreased by 10% Increased by 10-15% Decreased by 30-40% Negligible No data No data No data No data No data No data Decreased by -50% Decreased by -50% Decreased by -50% Minimal change

Major inflammation (CRP~ 100-200 mg/L)

Comments

Decreased by 60-90% Decreased by 40-60% Decreased by 40-60% Increased by 30% Decreased by 30% Decreased by 20-30% Minimal to >75% dec. No change No change Decreased by 40-50% No change No change Decreased by >90% Decreased by >90% Decreased by >90% Decreased by >70%

Liver uptake Liver uptake ? Redistribution ? Liver extraction Probable redistribution More data needed More data needed More data needed More data needed More data needed More data needed More data needed More data needed More data needed More data needed More data needed

An approximate guide to the extent of change seen in analytes with different degrees of inflammation. The figures quoted are collated from disparate sources and the effects of diverse factors in different pathophysiologies may be important. TDP = thiamine diphosphate: * = of red cell glutathione reductase; PP = pyridoxal phosphate.

Body fat percentage (men) = (0.567 x waist in cm) + (0.101 x age in years)-31.8 Body fat percentage (women) = (0.439 x waist in cm) + (0.221 x age in years) - 9.5 Because the waist circumference incorporates a measure of the internal fat mass as well as of total body fat, it has recently found a place in health promotion to alert people to the need for weight management. Health risks start to develop when the waist circumference exceeds 80cm (32 in) in women or 94cm (37 in) in men. Cardiovascular risks are approximately tripled, and most subjects have symptoms from overweight when the waist exceeds 88cm (35 in) in women or 102cm (40in) in men. Waist circumference does not depend on height.

Biochemical nutritional assessment Biochemical measures are of relatively limited value in nutritional assessment. Blood or plasma is seldom the appropriate compartment to detect functional nutritional abnormalities, and extracellular fluid nutrient contents give only a crude (and sometimes misleading) indication of nutrient status. For some nutrients, e.g. zinc or amino acids, the plasma concentration may actually increase in starvation, reflecting tissue destruction. Abnormal plasma vitamins or trace elements are most frequently found when there is clear clinical evidence of malnutrition. An exception to this generalization is the measurement of haemo-

globin and the nutritional haematinics - transferrin iron, folate and vitamin B12 - where subclinical deficiencies are relatively common and demand treatment. Clinical biochemistry services now offer assays for vitamins (commonly vitamin A and carotenoids, B1, B2, B6, B12, folate, C, D, E) and trace elements (e.g. iron, copper, zinc, selenium). These are used mainly to confirm clinical suspicion of deficiency, and to monitor artificial feeding. Results always have to be interpreted in relation to clinical condition, and to indicators of stress or acute-phase reaction, such as C-reactive protein. Many vitamin measurements, e.g. vitamin C, are strongly influenced by inflammatory states (Table 5.12). Several proteins produced by the liver may be used in association with nutritional assessment, namely albumin, prealbumin, transferrin, retinol-binding globulin and thyroid-binding globulin. Plasma levels are of limited value on their own because they are affected more by short-term energy intake and by stress or inflammation.

DIETARY INTAKE ASSESSMENT Estimation of dietary intake, both of individuals and of groups or populations, is critically important to ensure health. It should not be considered an assessment of nutritional status but it complements nutritional status assessment. Dietary intake can be assessed in terms of eating

117

pattern, foods, food groups and nutrients. All of these elements are important to different aspects of health by reference to standard recommendations or to dietary prescriptions for individuals. The dietary assessment of individuals is a specialized activity which needs to be conducted by trained dietitians.

Dietary intake of individuals The habitual eating pattern of an individual undergoes great variation from day to day. It is therefore usually inadequate to base any judgement on the eating habits of a single day, or even a small number of days. Certain foods are often consumed on only one day per week or less (e.g. fish on Fridays in some communities, carrots as the major contributor of B-carotene on Sundays, and liver as the major source of vitamin A much less frequently). The dayto-day variability of major nutrients is often of the order of 30-50%. Two dietetic approaches are used:

118

• Retrospective dietary assessment This approach relies on the skill of the dietitian in establishing eating habits, depending on the memory of the subject. Retrospective approaches are sometimes used for a single 24-hour period or for a week, or through questionnaires to estimate habitual intake over a much longer period, perhaps 1 year. As a general rule, the more questions asked and the longer the interview, the more foods are remembered and a greater value is obtained for food intake. Cross-check questions should be employed in the dietary history in order to remind subjects of foods that may have been omitted (for example, asking how often milk is purchased may show a disparity from the answers to a question about the amount of milk consumed). The latter approach, which gives some indication of the nutritional quality of the diet, is usually employed to express the fat, carbohydrate and protein intake as percentage calories. Retrospective dietary assessment has a place in clinical practice when there is an interest in the consumption of individual foods, and learning about an individual's meal patterns is often more valuable than attempting to establish nutritional adequacy, which can often give misleading results. Patients currently referred for dietary intake assessment are mostly overweight or obviously undernourished. In both contexts misreporting is common. Overweight people, and those with restrained eating, systematically under-record or underreport by up to 50%. Brief questionnaires attempting to estimate the intake of individual nutrients, e.g. calcium, have been developed, but give rather poor results compared to fully quantitative methods. • Prospective dietary assessment. This approach does not rely on subjects' memories but does rely on a high degree of cooperation over several days, during which they are asked to complete a full inventory of all foods and drinks consumed, together with details of cooking methods and portion sizes, which may be weighed using food scales or again related to household measures.

The 'gold standard' for dietary intake assessment is the 7-day weighed intake. This approach is often impossible to apply to children, to busy working people, to the confused or demented, and to those who are socially deprived or have literacy problems. For this reason, dietary assessment is often problematic in sectors of the population where nutritional problems are most prevalent. Records over 3-4 days, including weekends and weekdays and appropriately adjusted to estimate 7-day intake, are often used. A dietitian compiles a list of foods consumed by the individual and then makes some assessment of the portion sizes and quantities eaten, usually by reference to simple household measures. Standard food tables (for example McCance & Widdowson's Composition of Foods) are now available on computer databases, so the foods can be identified by code number, and the nutrient composition per 100 g is then available for 40-60 nutrients. Nutrient intake data can be extracted in absolute amounts (in g or mmol per 24 hours), or in nutritional terms, expressed as % energy mg or mmol per l000kcal of food energy. It is helpful to be able to evaluate nutrient intake by meal type, in order to identify meals (e.g. lunches) which are the cause of suboptimal total diet composition.

Dietary assessment of populations and groups Several levels, of differing complexity, cost and precision, are available. The best plan is usually to use more than one approach and to check that the results corroborate each other. • Food disappearance rates. The food and agriculture organization of WHO compiles food balance sheets for each country annually. If the total food supply (commercial production, home production, importation) and the population base are known, then an estimate can be made of food consumption per head of population and therefore of nutrient consumption. Data of this kind are often incomplete, particularly with reference to home food production. A number of assumptions have to be made about the ways in which foods are used, wastage in preparation and cooking, etc. This approach gives no guide at all to the intake of individuals or of population subgroups which may be of concern. On the other hand, broad generalizations about dietary adequacy can be made, and comparisons between countries with different disease patterns. • Household food purchasing. The National Food Survey of the United Kingdom, established in 1945, set up a model which has been employed in many countries: a random selection of households is taken every year and their total food purchases surveyed for 1 week. Analysis requires a knowledge of the numbers of individuals within the household. Knowledge of the social setting may also be of value in establishing the eating patterns

of population subgroups. Assumptions still have to be made about the uses to which food purchases are put. The biggest problem in the past has been that household food surveys have not included any measure of foods purchased and eaten outside the home, or of home food production. Therefore, for a number of nutrients diet intakes are spuriously low. This applies particularly to nutrients such as fat and saturated fat, which are present in high concentrations in foods consumed as meals and snacks outside the home. • Food frequency questionnaires. A number of questionnaires have been validated to estimate the food and therefore the nutrient consumption of groups by enquiring about the frequency of consumption of a range of foods. In the computer analysis, standard portion weights are applied to enable estimation of per capita nutrient consumption. Food frequency questionnaires are not sufficiently precise to be used in the assessment of food intakes of individuals, but they are relatively cheap to use. Inevitably, they address only a limited range of major foods and will therefore tend to underestimate food intake, particularly in individuals whose range of food choices is wider. • Prospective dietary intake. Epidemiological approaches have employed the same prospective weighed intake methods used in the assessment of individuals. Although arguably the most reliable approach, this is an expensive and professionally intensive method, and problematic to more deprived sectors of the population. In clinical practice, all methods of dietary intake assessment tend to underestimate the habitual food intakes by 20-50%, particularly those who are overweight. The reason for this is uncertain and there may be selective misreporting (e.g. for high-fat foods). During the assessment period of prospective methods, overweight people tend to cut down on their usual food intake. There is also an element of self-deception and an unwillingness to reveal true eating habits. It is interesting that in many cultures high social status is associated with a denial or minimization of a number of vegetative and physical bodily functions, including food intake.

NUTRITION AND FEEDING PRACTICES FOR GROWTH Standards for child growth - compiled from subjects living in a clean environment with an abundance of food - show a considerable range in rates of gain in weight and stature. There are also differences in dietary intake in such children, which is also true in healthy adults who are not excessively over- or underweight. A man in his 30s who continues to eat the same amount of food that he ate in his late teens is likely to become overweight. In each case, the difference in nutritional status is the result of the balance between energy intake and energy expenditure.

Early infancy In the young infant nutritional requirements are particularly high in relation to body weight. Breast milk certainly provides enough energy and appropriate nutrients for at least the first 4 months, and has psychosocial and immunological advantages over artificial formula feeds. Recent evidence also points to better brain development in breastfed infants, related to essential fatty acid intake. Artificial formulae are frequently dangerous when prepared under unhygienic conditions, as in the homes of many poor families in the developing world. Cows' milk contains a dangerously high sodium content and should not be given before 1 year. Weaning should be gradual from 4-6 months: overfeeding in infancy may programme an excessive appetite.

a

Later infancy and early childhood Nutritional requirements are also high in infants over 4 months of age, when children can begin to receive solid foods in addition to breast milk. In developing countries the growth rate frequently falters at this stage, either because the energy given is not sufficient (the energy density of weaning foods in traditional societies is often low) or because of gut infection, so that food is not so well absorbed. Specific nutrient deficiencies (e.g. zinc, amino acid, folate) may also play a role. Cows' milk consumption in young children has been associated with bleeding into the bowel and anaemia, leading to the development of 'follow-on' formula feeds for use in the first year of life. There is very weak evidence that exposure to common dietary allergens in infancy may provoke allergies so foods with low allergenic potential (e.g. rice rather than wheat) are preferred.

Later childhood and adolescence Nutritional requirements for protein and energy are usually met in older children, but micronutrient deficiencies (e.g. of iron, folate and zinc), which may lead to anaemia, impaired intellectual performance and short stature, are relatively common even in western populations, and easily go unnoticed.

Pregnancy The nutritional stress of pregnancy is created by the need to produce both an infant and a placenta, plus some maternal fat stores. A 12.5kg weight gain in pregnancy would require a total of about 7000kcal (Table 5.13). There are also increased demands on maternal stores, especially of iron, zinc and folate. On the other hand, carefully conducted studies have not shown a greatly increased food intake in pregnancy. For some nutrients, e.g. calcium and iron, the efficiency of intestinal absorption is increased during pregnancy. Energy intake often falls in the first trimester, and then needs to increase to about 200kcal/day

119

above non-pregnant intakes. The additional energy requirements of pregnancy are met from reduced physical activity and greater resting in the later stages. Fetal growth certainly depends on maternal nutritional status, both from starting weight and from weight gain, and may be specifically reduced by physical exertion in late pregnancy. Intriguingly, it has recently been shown that food deprivation in late pregnancy leads not only to smaller babies, but also to these offspring having smaller babies themselves. Good fetal growth is important for later, adult, health (to protect against diabetes, heart disease, stroke), and so maternal nutrition is doubly important. The Institute of Medicine recommends that, at term, women with BMI <19.8 should have gained 12.5-18 kg, those with BMI 19.8-26, 11.5-16kg, and those with BMI 26-29, 7-11.5kg. A BMI 28kg/m2 is a reasonable target at term. Table 5.13 summarizes the extra nutrient requirements of pregnancy, and suitable sources.

Lactation Babies should be exclusively breastfed for 4 months, and according to WHO recommendations breastfeeding should be continued for 12-24 months. The need to produce breast milk (up to 800kcal/day) for many months has obvious nutritional implications. Some of this can be provided from maternal stores laid down in pregnancy, but an additional 500kcal intake and extra protein are often

TABLE S.13 Summary of special nutrient requirements in pregnancy 1. Energy - an increase of 200kcal/day is required in the 3rd trimester. Aim to reach BMI 28 kg/m2 at term (2310kcal for women aged 15-18) (2140kcal for women aged 19-50) 2. Protein - an increase of 6g per day throughout pregnancy (51 g per day) Two small/moderate portions of protein-rich foods, plus milk, bread, pulses, etc. 3. Iron - 14.8mg per day required throughout adulthood. No extra routinely needed during pregnancy: absorption increases 4. Calcium - 100mg per day (19-50) 800 mg per day (15-18) No extra routinely needed during pregnancy: absorption increases Throughout life - half a pint of milk daily as a good insurance 5. Vitamin D-10ug per day Supplement. Encourage women to go outdoors Use fortified foods, e.g. margarine, and vitamin D rich foods, e.g. oily fish 6. Folic acid - supplement from before conception up to first 12 weeks (0.4 mg routinely, 4mg if previous neural tube defect) 7. Vitamin C - increase intake by 10mg to 50mg per day More fruit and fruit juice 8. Vitamin A - increase vegetables, increase intake by 100ug to 800jag

120

9. Dietary fibre - to avoid constipation of pregnancy, increase by 10g/day More vegetables, wholemeal bread

recommended. Many women are able to produce adequate volumes of breast milk despite poor dietary intakes. However, when maternal nutritional status is severely decreased, there is a corresponding decrease in the volume and nutritional content of the milk. It is thus essential to improve the diet and lifestyle of women in developing countries if breastfeeding is to be promoted. When there has been substantial fat accumulation during pregnancy, as is usual in western countries and where obesity is more of a problem than undernutrition, prolonged breastfeeding (4-12 months) helps with weight loss.

MALNUTRITION MALNUTRITION IN CHILDREN The term protein-energy malnutrition (PEM) was used historically to describe a variety of clinical syndromes believed to result from a diet inadequate in both energy and protein. The term is misleading, since dietary protein deficiency is virtually never seen, and low serum proteins in malnutrition are not caused by dietary protein insufficiency. Undernutrition of energy is often accompanied by micronutrient deficiencies, electrolyte imbalance and metabolic disorders. When it occurs in the UK it is usually precipitated by a severe illness (e.g. gastrointestinal disease, surgery). In developing countries it is the result of a poor diet, often exacerbated by the added stress of infection (especially diarrhoea and measles). The term malnutrition/infection syndrome may be used to describe this situation. Classification of the severe clinical syndromes (Table 5.14) depends on the child's weight calculated as a percentage of the median international standard for a child of the same age. Milder grades of malnutrition also exist according to weight/age. Undernourished children may be further classified according to whether, or to what degree, they are short for their age (stunting); alternatively, they may be underweight and thin, i.e. a low weight/height (wasting) (see Table 5.14 and Fig. 5.7). The inference is that children who are stunted are likely to be chronically

TABLE 5.14 Classification of stunting and wasting Degree

Malnutrition (weight/age)

Stunting (height/age)

Wasting (weight/height)

Normal

Over 80%

Over 95%

Over 90%

Mild Moderate Severe

60-80%* Less than 60%*

87.5-95% 80-87.5% Less than 80%

80-90% 70-80% Less than 70%

* Defined as kwashiorkor if oedema present

FIG. 5.7 Growth chart of a severely malnourished child A Start of nutrition rehabilitation. B After discharge from nutrition rehabilitation.

infected or have been deprived of an adequate diet, usually for many months (sufficient to cause linear growth retardation). The child who is wasted but not stunted has usually lost body tissues because of a recent infection or poor diet, for example due to harvest failure.

Marasmus In marasmus (weight for age <60% without oedema), which is usually seen in older, fully weaned children, there is loss of both lean tissue and subcutaneous fat, with characteristic wrinkling of the skin. This is sometimes difficult to differentiate from dehydration. The child is usually very cachectic, and has long eyelashes and lanugo hair on the forearms despite loss of hair on the head. However, the child is generally alert and keen to eat. The plasma albumin is usually maintained, but plasma sodium and potassium levels are often low. Marasmus describes the result of dietary energy shortage, but specific nutrient deficiencies commonly coexist, as in anorexia nervosa.

Vitamin A becomes sequestered in the liver because of insufficient synthesis of the transport protein, retinolbinding protein, and patients are consequently more prone to infection. Plasma albumin is low, as are plasma sodium and potassium. The low plasma potassium indicates a reduction in the total body potassium, owing to the loss of muscle mass. The low plasma sodium, however, is misleading, as it is due to an excess of body water: the total body sodium is in fact elevated. Several factors contribute to the development of oedema. A decrease in plasma albumin decreases colloid osmotic pressure, whereas an increase in the generation of free radicals by infection (from systemic or intestinal pathogens, or toxins, such as aflatoxin) can damage the capillaries sufficiently to cause them to leak fluid into the tissues. The deleterious effects are greatest in children with impaired scavenging mechanisms for free radicals, owing to a relative excess of iron and copper but a deficiency of vitamins C, E and carotenoids.

5

Associated deficiencies Zinc deficiency is common in severe malnutrition, and may account for the skin lesions and impaired cellular immunity. Vitamin A deficiency is frequently precipitated by malnutrition, probably because of lack of synthesis of retinol-binding globulin by the liver, and anaemia often occurs as a consequence of folate and/or iron deficiency and systemic infection (especially malaria). Cellular immunity is frequently depressed, permitting infections such as tuberculosis to become rampant. Antibody production is also impaired, but to a lesser extent, so that even severely malnourished children generally produce satisfactory levels of antibodies following immunization. However, the general reduction in immunity is a key factor in the development of bacterial overgrowth syndromes in the intestine, pneumonia and septicaemia. The decreased heat production in malnutrition is often severe enough to cause hypothermia and prevents a pyrexial response to infection.

Specific organ changes Kwashiorkor The main feature of kwashiorkor, which usually affects infants at the time of weaning, is oedema, which conceals the loss of lean tissue. There is little evidence that dietary differences explain why some children develop kwashiorkor and others become marasmic; kwashiorkor is often precipitated by measles and intestinal and other infections. These children are ill and anorexic, and have very high mortality. Weight for age <60% indicates marasmic kwashiorkor. The oedema usually starts in the legs but may extend to the trunk and head. There is often an associated scaly cracking of the skin, and some children have discoloured or orange hair (caused by associated zinc deficiency). The liver is sometimes enlarged, as it becomes infiltrated with fat which cannot be removed because of inadequate production of lipoproteins (see Ch. 19).

There is atrophy of all organs, including alimentary tract mucosal atrophy, which causes hypochlorhydria and flattening of the intestinal villi with secondary malabsorption. Nearly all children are heavily colonized by intestinal parasites and bacteria. Renal function is often impaired, leading to difficulties in excreting sodium and water. Experimental studies show effects of childhood malnutrition on brain size and psychomotor development, but longitudinal studies of severely malnourished children show that the social environment is more important than nutrient deficiency in determining subsequent intellectual achievement.

Management The following principles should be followed:

121

TABLE 5.15 Predicted rate of weight gain* in malnourished child weighing 7kg Total intake (kcal/day)

Excess (kcal/day)

Weight gain (g/day)

Days to gain 300 g

0

0

-

760

60

10

300

820

120 240

20 40

150

480 960

80 160

700

940 1180 1600

75 38 19

* Assuming maintenance requirements of 100kcal/kg/body weight

• The child requires resuscitation for dehydration and electrolyte restoration (in most cases using oral rehydration therapy). • Antibiotics (this is still an indication for chloramphenicol) should be given for systemic infections, together with treatment for malaria where appropriate. It is wise to assume that all children with severe malnutrition have bacterial infections, tuberculosis and intestinal infestation, even when classic signs of infection are absent. • Nutritional therapy, which depends on the introduction of special feeds providing maintenance requirements (100-150kcal/kg body weight, including 10% energy as protein, and avoiding lactose), should be accompanied by mineral mixes (especially potassium and zinc). In children with kwashiorkor, maintenance requirements are given until the oedema is cleared and then energy intakes are increased. In marasmic children the intake can be increased earlier until, ideally, about 200 kcal/kg body weight/day are given (Table 5.15).

CACHEXIA AND MALNUTRITION IN ADULTS Cachexia of dietary origin, or malnutrition in adults, is surprisingly common in medical and surgical practice. A recent survey in the UK found that 13% of admissions to general medical and surgical wards had malnutrition according to BMI criteria. Higher figures have been found using other criteria. Malnutrition was only recognized in a small proportion, and most patients lost more weight during their admission without any nutritional action being taken. Inadequate intake is the most important cause,

owing to psychological factors (such as depression, isolation or anorexia nervosa), immobility and social deprivation, pain, oral or intestinal cancer, arthritis, nausea/ vomiting, liver metastases, chemotherapy, gastrointestinal disease, fever or postoperative complications (which often cause anorexia). Nutrient intakes are often reduced drastically in those who can only take fluids. 1 Requirements may be considerably increased by sepsis. Energy expenditure increases by at least 10% for every 1°C rise in body temperature, and muscle protein is mobilized to provide amino acid precursors for the inflammatory response. Infection also increases the production of acute-phase proteins such as C-reactive protein, but levels of plasma albumin may decrease as a result of decreased synthesis by the liver and increased catabolism of circulating albumin. Any cause of malabsorption may cause a serious decrease in absorption of protein and energy. Disease of the lower small intestine and colon may cause serious leakage of endogenous protein and protein loss from blood. To produce cachexia, these factors must be chronic, or relapsing without adequate periods of remission, and often more than one factor is operating at any time. The classic causes of malnutrition in adults are cancers, tuberculosis and gastrointestinal problems, including short bowel syndrome and fistulae. Depression is another important cause, but organic disease should always be excluded. More recently AIDS and chronic obstructive airways disease (COAD) have become familiar as causes of apparently intractable cachexia. The most consistent finding is the dominance of anorexia during acute infective episodes, leading to stepwise weight loss. In HIV, as in inflammatory bowel disease, there is also a background of steady weight loss through malabsorption diarrhoea. Elevation of metabolic rate in these conditions plays only a small part in producing weight loss. The most important single manoeuvre in preventing and managing malnutrition is to have good measurements of the weight of every patient at every contact with health services (inpatient and outpatient). The diagnosis must be recorded, as it has profound medical implications, and failure to diagnose malnutrition might incur medicolegal problems. Dietetic intervention, social support and artificial feeding where necessary are much more effective if used early in the process of weight loss. Even with AIDS or COAD, effective nutritional interventions which lead to weight gain will improve muscle function, improve quality of life and reduce susceptibility to infection. Correction of malnutrition takes time, so management started in hospital must be continued after discharge. ©

VITAMIN DEFICIENCIES AND EXCESSES 1 Fig. 5.1 122

0 Case 5.1

A checklist for the methods of assessment of vitamin deficiencies is given in Table 5.16. The important food sources

SUMMARY 3 Vitamin deficiency syndromes Vitamin Syndrome A Thiamine Niacin Riboflavin Pyridoxine Folate B12

C D K

Night blindness, xeropnthalmia Wernicke's encephalopathy, Korsakoff's psychosis Pellagra Mucosal lesions Glossitis, neuropathy Megaloblastosis, villous atrophy Megaloblastosis, neuropathy Scurvy Rickets, osteomalacia Hypoprothrombinaemia

TABLE 5.16 Assessment of nutritional deficiencies: vitamins* Vitamin

Method

Comment

A

Clinical

Infections common Night blindness is a good indicator Corneal lesions are very late-stage findings

• Plasma carotenoids • Plasma retinol Thiamine

Clinical • HPLC • Erythrocyte transketolase

Sensitive indicator of liver stores but reduced in infection Neuropathy and optic atrophy are late-stage findings Specific Sensitive but rapidly altered by amount of dietary thiamine

of the various vitamins are shown in Table 5.17, and their roles in the body in Table 5.18. A table of recommended daily intake of vitamins is given in Table 5.19.

Riboflavin

Clinical • Erythrocyte gluthathione reductase

Mouth lesions are non-specific Assay related poorly to clinical state

VITAMIN A DEFICIENCY

Niacin

Clinical • Urinary n-methylnicotinamine/creatine excretion

Glossitis, stomatitis, dermatitis Sensitive. Difficult assay

Folate

• Serum folate • Erythrocyte folate

Marked changes after dietary intake Sensitive assay of body stores

B12

• Plasma vitamin level

Sensitive assay of body stores

C

Clinical • Plasma or leucocyte ascorbate (HPLC)

Specific but occur at late stage only Sensitive. Rapid response to diet, Falls non-specifically in illness

D

Clinical • Plasma alkaline phosphatase • Plasma 25hydroxycholecalciferol • Bone marrow histology

Rickets, osteomalacia Non-specific. Sensitivity may be improved by isoenzyme analysis Very specific

E

• Plasma or erythrocyte tocopherol (HPLC)

Sensitive and specific. Must be corrected for plasma LDL

K

• Prothrombin time

Non-specific, as it is influenced by hepatocellular dysfunction

Vitamin A (retinol) can be synthesized within the liver from several carotenoids (cc-carotene, B-carotene, lutein) which occur in a variety of vegetables and certain fruits. Vitamin A deficiency leads to night blindness and xerophthalmia (damage to the conjunctiva, the major cause of blindness in children). It also affects the immune system - increasing the severity of respiratory infections and diarrhoea - and the formation of epithelium in the sweat glands, lung and gut. It contributes to the high rates of child mortality in certain populations, particularly among poor communities, where dietary intakes of carotene are low and where the availability of the nutrient may be further reduced by intestinal parasites and diarrhoea, and where undernutrition suppresses the synthesis of retinol-binding protein, preventing utilization of hepatic retinol stores. Vitamin A deficiency sometimes complicates inflammatory bowel disease, coeliac disease and cirrhosis. Measles is an important contributory factor in xerophthalmia because of its direct effect on the conjunctiva, where it increases the local requirement for vitamin A to repair the damaged conjunctiva, and also by its effect on synthesis of retinol-binding protein, which normally carries vitamin A from the liver stores to the eye. Supplementation with vitamin A reduces the mortality from measles in malnourished children, even when their liver contains vitamin A (because its transport to other tissues is inhibited). Night blindness alone is easily reversible, but xerosis (the dry appearance of the conjunctiva) and Bitot's spots (heaped-up desquamated cells causing flecks on the conjunctiva) are signs that, unless treatment is started soon, the lesion may progress to irreversible corneal ulceration. In areas where dietary vitamin A intake cannot be increased by horticulture and nutrition education pro-

5

Specific

* Pyridoxine is omitted as this is not a routine assay

grammes, the administration of vitamin A capsules (100 000 IU every 4 months) will largely prevent xerophthalmia. In established xerophthalmia, retinol palmitate (100 000 IU) is given intramuscularly. Vitamin A should not, however, be given to pregnant women because of its teratogenic effect. Breastfeeding tends to prevent vitamin A deficiency.

123

TABLE 5.17 Vitamins and their important food sources

TABLE 5.18 Major roles of vitamins in the body

Vitamin

Food sources

Vitamin

Function

A

Carotenoids (esp. p-carotene), in carrots, dark green leafy vegetables, pumpkin, mango Preformed vitamin A in fish oils, liver, eggs, margarine

A

Maintenance of mucosal surfaces; production of mucus; immune system and retinal function; fat soluble

Thiamine (B1)

Thiamine (B1)

Fortified wheat flour (much of the thiamine in flour is removed by milling, but is then replaced in most flours), fortified breakfast cereals, milk, eggs, yeast extract, fruit

Metabolism of carbohydrate to provide energy; maintenance of cardiac muscle and peripheral nerves

Riboflavin (B2)

Control of intracellular oxidation; maintenance of skin, especially around the mouth

Riboflavin (B2)

Milk, cheese, eggs, fortified breakfast cereals, liver, kidney, whole grain cereals

Nicotinic acid

Utilization of energy from food; maintenance of skin, especially in light-exposed areas

Niacin

Liver, kidney, milk, cheese, eggs, beef, chicken, pork, yeast extract, instant coffee, peas, beans

Pyridoxine (B6)

Metabolism of amino acids, especially tryptophan to nicotinic acid

Pyridoxine (B6)

Liver, meat, fish, whole grain cereals, milk, peanuts

Folate

Cell multiplication, in the blood and mucosa

Folate

Liver, kidney, green leafy vegetables, peas, beans, oranges, bananas

Cobalamin (B12)

Cell multiplication, especially blood cells; control of fatty acid synthesis in myelin sheaths of nerves

Cobalamine (B12)

Liver, kidney, fish, beef, pork, lamb, eggs, cheese

Ascorbic acid (C)

Ascorbic acid (C)

Oranges, lemons, potatoes, green vegetables, fortified fruit drinks

Maintenance of connective tissue; stimulates iron absorption

D

D

Fortified milk (e.g. evaporated), fatty fish, fortified margarine, eggs, liver

Maintenance of calcium and phosphate levels in the blood; control of absorption of dietary calcium and movement of calcium between blood and bone; fat soluble

E

Vegetable oils, wholegrain cereals, eggs, margarine

E

K

Vegetables, peas, beans, liver

Contributes to antioxidant capacity of body; important in maintenance of erythrocyte membranes in neonates; fat soluble

K

Control of blood clotting; fat soluble

Thiamine deficiency

124

Thiamine is an essential factor in many glycolytic pathways and is especially important for normal function of the peripheral nervous system and the heart. Deficiency can cause a peripheral neuropathy, especially in the legs, ataxia and optic atrophy. Most cases in the UK occur in malnourished alcoholics. Such patients may develop two additional syndromes. In Wernicke's encephalopathy nuclear or supranuclear ophthalmoplegia occurs with nystagmus, pupillary abnormalities, aphasia and mental confusion. In Korsakoffs psychosis a permanent confused state, characterized by short-term memory loss and confabulation, develops (see Chapter 24, p. 1439). In beriberi, a highoutput cardiac failure develops (wet beriberi) and may accompany the peripheral neuropathy (dry beriberi). Deficiency is more common in populations which change to eating highly milled rice, as this process removes the thiamine. Curiously, Korsakoffs psychosis does not seem to occur commonly in populations prone to beriberi, suggesting that in alcoholics an additional factor may be acting. Diagnosis is usually made on clinical history and examination, but subclinical thiamine deficiency should always be suspected in alcoholics, and supplements given prior to feeding carbohydrates. If carbohydrate (e.g. dextrose infusion) is given without thiamine, then lactic acidosis can

rapidly lead to brain damage. Elevated levels of erythrocyte transketolase activity and blood pyruvate occur. Thiamine (100 mg) should be given intravenously (a multiple B vitamin parenteral infusion is appropriate), followed by repeated doses of 20 mg three times daily. Cardiac beriberi and ophthalmoplegias improve rapidly, but Korsakoffs psychosis tends to recover only partially. Many routine vitamin supplements do not contain enough thiamine to correct a deficiency. Long-term oral thiamine supplementation is a reasonable preventative strategy for alcoholics.

Niacin (nicotinic acid and nicotinamide) deficiency Nicotinic acid occurs in many plant and animal foods and can also be synthesized from tryptophan. Nicotinic acid deficiency tends to occur in those who eat a monotonous maize diet (which is low in tryptophan), such as is often supplied to refugees. The resulting deficiency syndrome is pellagra. Pellagra often starts with either a sore mouth, mild diarrhoea or confusion. The full clinical syndrome of 'dermatitis, diarrhoea and dementia' is not common, except in well-established cases. The skin lesions are initially erythe-

TABLE 5.19a Reference nutrient intakes for vitamins Males over 11 years Thiamin (mg/d) Riboflavin (mg/d) Niacin (nicotinic acid equivalent) (mg/d) Vitamin B6 (mg/d) Vitamin 812 (ug/d) Folate (ug/d) Vitamin C (mg/d) Vitamin A (ug/d)

1.0 1.2

17

Females over 11 years

Lactation (0-4 months)

0.8 1.1 14

1.4

1.2

1.5

1.5

+0.2 +0.5

+2

+0.5

200

+60

40

40

+30

700

600

+350

* no increment

TABLE 5.19b Reference nutrient intakes for minerals (SI units)

Calcium (mmol/d) Phosphorus (mmol/d) Magnesium (mmol/d) Sodium (mmol/d) Potassium (mmol/d) Chloride (mmol/d) Iron (umol/d) Zinc (umol/d) Copper (umol/d) Selenium (umol/d) Iodine (umol/d)

Females over 11 years

Lactation (0-4 months)

25

20

+14.3

25

20

+14.3

12.3

12.0 70

70 90 70 180 140

16 0.9 1.0

+2.1

90

* *

70 260**

* *

140 16 0.9 1.0

a

*

200

Males over 11 years

thus accompany general malnutrition or alcoholism and is especially liable to occur during pregnancy. The main lesions are mucosal dryness and cracking around the mouth, ears and eyelids, with the genital region also affected. Red blood cells are affected, causing a normochromic anaemia. Diagnosis is usually on clinical grounds but can be confirmed by reduced glutathione reductase activity in red blood cells. Lesions respond rapidly to riboflavin (6mg daily).

+90

+5 +0.2

*

*no increment ** 160 for females over 50 years

matous, but dry, cracked appearances then occur, especially in areas exposed to the sun, producing 'Casal's necklace' around the neck. There may be villous atrophy in the small intestine and malabsorption. If untreated, later-stage deficiency may cause severe encephalopathy and spasticity. Pellagra is rare in the UK, but may occur in malabsorption syndromes such as coeliac disease, in patients treated with isoniazid, in Hartnup's disease, in the carcinoid syndrome, in patients with very low-protein diets, or in conditions giving rise to considerable loss of protein such as the nephrotic syndrome. Deficiency responds rapidly to daily doses of nicotinic acid (500 mg).

Riboflavin (vitamin B2) deficiency Riboflavin deficiency occurs when a diet is continually deficient in legumes, pulses and animal products. It may

Pyridoxine (vitamin B6) deficiency Pyridoxine deficiency occurs in malabsorption syndromes, alcoholism and secondary to drug toxicity (especially isoniazid), but rarely as a primary deficiency disorder. Clinically, there is mental confusion, glossitis, dry skin lesions and peripheral neuropathy. Pyridoxine deficiency has rarely been associated with infantile convulsions and sideroblastic anaemia. Pyridoxine deficiency is popularly believed to be a cause of premenstrual syndrome but there is little evidence for this. Treatment of deficiency in adults is by 5-20 mg daily. Excess intake can lead to neurotoxicity.

Folate deficiency Folate deficiency occurs in those whose diet lacks fresh vegetables, or in which the vitamin is destroyed by cooking or lost through canning. It may also occur secondary to upper small bowel malabsorption, in situations of increased requirements (i.e. pregnancy or haemolytic conditions), or in patients taking anti-folate drugs (e.g. methotrexate, trimethoprim). Alcohol inhibits the absorption of folate, then aggravating the effect of poor diet in alcoholics. Clinical features, diagnosis and treatment are described in Chapter 23. As well as the megaloblastic anaemia and pancytopenia, depression has been consistently associated with folate deficiency. Folic acid supplementation 400ng/day is recommended for all women who may become pregnant, to prevent neural tube defects (spina bifida, anencephaly). This is not to treat a deficiency, but a pharmacological effect to overcome defects in folate metabolism present in a minority of women. Availability of folate from foods is lower than that of folic acid from supplements.

Vitamin B ]2 deficiency Vitamin B12 deficiency is described in Chapter 23. Dietary deficiency is rare, but can occur in strict vegans.

Vitamin C deficiency The clinical features of scurvy have been identified as part of general malnutrition for centuries, but the specific features of scurvy and its cause only became apparent after it decimated crews on long-distance sea voyages after the discovery of America. In the 18th century, Lind carried out a controlled trial, treating two cases of scurvy with oranges

125

and lemons and using mouthwashes of dilute vinegar or sea water for four other patients. Complete recovery occurred in the treated patients but none in the other patients. Today, deficiencies occur in the elderly - especially those on 'tea and toast' or 'tinned spaghetti' diets - in alcoholics and in food faddists. It may also occur in young children when requirements are high, and in smokers, as ascorbate is consumed in removing the free radicals from smoke. Low ascorbate levels are found in diabetic patients. The clinical features of tiredness, hyperkeratosis of the hair follicles, petechial haemorrhages and bleeding gums are only present in severe deficiency; bleeding gums are more usually due to poor oral hygiene than to vitamin deficiency. The classic features of infantile scurvy - irritability, pain on moving the legs, and tenderness and enlargement of costochondral epiphyses, together with megaloblastic anaemia - are now rarely seen. Diagnosis is by measurement of plasma ascorbate or leukocyte ascorbate content. Ascorbic acid is the antioxidant vitamin whose plasma concentration correlates best with diet. Low concentrations are very common in hospitalized or ill patients, and this is in part an acute-phase response. Ascorbic acid 100 mg three times per day is rapidly effective in clinical deficiency. There is limited evidence that large doses of vitamin C may reduce symptom severity in common cold but this does not point to an underlying nutritional deficiency.

Vitamin D deficiency Vitamin D deficiency occurs in those who fail to produce enough in their skin (from the precursor 7dehydrocholesterol in response to UV light), or who have inadequate dietary intake. In the UK, the high-risk group is immigrant vegetarian Asians, who tend not to expose themselves to the limited sunlight available. Although they do eat fatty fish (a good source of vitamin D), their consumption of vitamin D-fortified foods, e.g. margarine, is low, and the high phytate content (in chapatti flour) of the diet decreases its bioavailability. The result of vitamin D deficiency is rickets (in children) or osteomalacia (in adults) (see Ch. 18, p. 962). 1

Vitamin E deficiency Vitamin E deficiency can occur in low-birthweight infants and in children with malabsorption syndromes (especially cystic fibrosis). It is very rare in adults but has been reported on inadequate total parenteral nutrition, following gastroplication surgery for obesity and in abetalipoproteinaemia, as tocopheral absorption and transport depend on low-density lipoprotein (LDL). Deficiency

1

126

Fig. 5.2

2

Fig. 5.3

permits free radical damage to a number of membranes, such as red blood cells (causing haemolytic anaemia) and nerve cells (causing ataxia and peripheral neuropathy, which is usually permanent). These conditions are exacerbated if extra nutrients, such as linoleic acid and iron, are given without vitamin E. There is no consistent scientific evidence of any effect of vitamin E on athletic success, sexual potency or atheroma. The only indication for therapy is if the level of plasma tocopherol is shown to be low in an appropriate clinical setting. Plasma concentration needs to be adjusted for LDL concentration for interpretation. Deficiency can be treated by 200 mg daily doses in adults, but much larger doses (parenterally) are needed to overcome the effects of abetalipoproteinaemia.

Vitamin K deficiency This may occur in neonates because placental transfer is limited and breast milk is often deficient. It also occurs in obstructive jaundice and certain malabsorption syndromes. The result is hypothrombinaemia and bleeding, which is rapidly reversed by injections of vitamin K (phytomenadione). The prolonged prothrombin time associated with liver disease seldom responds to extra vitamin K.

Fluoride deficiency Dental enamel has a specific requirement for fluoride, whose deficiency in drinking water (under Ippm) is the main cause of dental caries in children. Where fluoride consumption is high - e.g. from tea in Ethiopia - dental caries is rare even with high sugar consumption. Fluoridecontaining toothpaste virtually abolishes the problem in areas of fluoride deficiency. To reach children who do not brush their teeth, successful strategies include fluoride normalization of water supplies to 1 ppm, or fluoridization of salt (Switzerland). Most of the UK is fluoride deficient.

VITAMIN EXCESS SYNDROMES Toxicological syndromes associated with vitamin excess (Table 5.20) are mainly found with fat-soluble vitamins (A, E, D, K) which accumulate in the body.

Vitamin A Toxicity occasionally occurs acutely in children given large doses (e.g. 200000IU) of vitamin A intramuscularly. It takes the form of irritability, and even convulsions, as a result of increased intracranial pressure. Recovery occurs within a few days. Excessive self-medication is the usual cause. Levels of vitamin A in the livers of farmed animals have risen in recent years because of feeding practices. It is quite possible to consume, in a single meal containing liver, much more than the maximum permitted of 10000IU per day for

TABLE 5.20 Main consequences of vitamin toxicity Vitamin

Symptoms of toxicity

A

Headache, convulsions, muscular stiffness, hepatomegaly, teratogenic effects Not described Not described Vasodilatation Peripheral neuropathy Not described Not described Increased urinary oxalate Hypercalcaemia, renal failure Nausea Hyperbilirubinaemia

Thiamine (B1) Riboflavin (B2) Nicotinic acid Pyridoxine (B6) Folate Cobalamine (B12) Ascorbic acid (C) D E K

a woman during pregnancy. Large doses of vitamin A are associated with an increased prevalence of craniofacial abnormalities in the infant. For this reason it is recommended that pregnant women do not eat liver, which used to be encouraged for pregnant women to provide iron and other nutrients.

Pyridoxine (vitamin B6) Pyridoxine toxicity has been occurring more frequently in recent years owing to its use - successful or otherwise - in attempts to alleviate premenstrual syndromes (depression and fluid retention). A peripheral neuropathy is likely if the daily dose exceeds 500 mg, and has been reported at lOOmg/day. Some cases have neurological abnormalities which persist even after stopping the drug.

Vitamin C Toxicity occurs among those who take enormous doses of vitamin C in the unfounded belief that it will prevent the common cold or cancer. Doses above 2g/day may cause abdominal discomfort and osmotic diarrhoea. There is increased urinary oxalate excretion - which may dispose to oxalate stones - and some evidence of damage to the intestinal epithelium.

Vitamin D Vitamin D toxicity occurred in the UK after World War II in some children given too much fortified milk. (The quantities added are now more strictly controlled.) Mild forms of vitamin D toxicity are characterized by weakness, weight loss and growth failure. Severe forms cause nephrocalcinosis, radiological changes in the epiphyses and mental retardation.

TRACE ELEMENTS

a

The clinical relevance of trace elements in health is limited, but zinc deficiency appears important. Zinc, and probably also copper, is important in protecting cell membranes against free radical damage. Zinc is a cofactor in over 100 enzymes, and there are no body stores. Absorption from the diet is about 30%, but adapts to needs in health. Deficiency is seen most clearly in the congenital disorder aerodermatitis enteropathica, which is characterized in infants after weaning by a scaly skin rash, secondary bacterial and fungal infection, diarrhoea and severe growth faltering. All are rapidly reversed by zinc therapy. Deficiencies are also seen in children with severe malnutrition, and were formerly seen in patients receiving parenteral nutrition without sufficient trace element mixes. © Selenium deficiency can cause a cardiomyopathy, although its main role is as a cofactor in the antioxidant enzyme glutathione peroxidase. The selenium content of plasma falls during stress or illness as part of the acutephase response.

Iodine deficiency Most iodine deficiency is due to low levels in the soil and water. There are many regions in the world, e.g. Nepal, Peru and Bangladesh, where iodine intakes are so low (<50-75 mg/day) that thyroid function is impaired. There is a high prevalence of cretinism in neonates in these regions, which may present as deaf-mutism, mental retardation and ataxia, with the characteristic morphological features of thickened facial appearance, stunted stature and enlarged tongue. The brain damage is irreversible if not treated early. It is now also realized that the poor intellectual performance and deafness (without any unusual morphological features) found in many children in iodine-deficient areas is due to marginal iodine deficiency. This may be partially reversible. In some areas an already limited iodine availability is made worse by the consumption of goitrogens, e.g. in certain brassicas, cassava or soya beans. The extent of the problem can be assessed by measuring the prevalence of goitre (enlargement of the thyroid gland), which is particularly common among adolescents and pregnant women. Measurements of urinary iodide and iodine/ thiocyanate levels can determine whether the problem is primary dietary iodine deficiency (low iodine) or the result of a dietary goitrogen. Richer people in goitrous areas generally obtain enough iodine by eating food (e.g. fish) brought in from other areas, but the main strategy is to iodize salt. Where this is impractical, injections of iodized oil (lipiodol) may be given, especially to women of childbearing age. A single dose (5ml) will usually last for up to 5 years. Oral doses (400 mg) of iodine preparations may provide enough to last for 2 years.

127

OVERWEIGHT AND OBESITY Obesity is the accumulation of excess fat with multiple organ-specific pathology. The disease and its clinical consequences are obviously progressive, but for epidemiological purposes cut-offs of BMI 25 and 30kg/m2 are used, corresponding with the clinical hazards increasing and becoming frequent. Alternatively, waist circumference can be used as a simple indicator of total body fat, which is largely independent of weight.

Causes Obesity is due to an interaction between genetic constitution and the environment. The genetic component is clear, as is shown in studies of families and twins. The genes involved are not identified but are probably multiple, and they provide only the predisposition to more or less rapid weight gain. They may affect appetite regulation or metabolic efficiency. Together, genetic factors account for about 30% of predisposition to weight gain. The remainder comes from environmental influence - food attractiveness and availability, need for exercise etc. The fat accumulation itself can of course only come from excess energy consumption above needs over a prolonged period. 1 Obesity may be precipitated by an obvious decrease in energy expenditure due to a change in lifestyle (e.g. to a more sedentary job), or an injury or arthritis which limits mobility. However, in most obese individuals it seems that intake and expenditure problems occur together. Many people admit to eating excessively. A sympathetic request for a confidential 2-week diary of everything the patient eats may be very revealing, but most obese people underestimate their intake. Depression and psychosocial problems are common, so a brief history in the outpatient clinic will usually fail to give an accurate assessment. The metabolism of food energy has been studied in detail over the last 20 years, and the evidence gives some clues as to how some individuals are more prone to weight gain than others with a similar energy intake. • People with a high fasting respiratory quotient (RQ) (CO2 production divided by O2 utilization) are most prone to weight gain. • When volunteers are given extra calories, some put on weight faster than others. • Dieting leads to an adaptive decrease in energy expenditure, and this may vary between individuals. • Obese people have a decreased thermogenic response to stimulation by noradrenaline (norepinephrine) or food, and this persists even when they have lost weight.

1 128

Figs 5.4, 5.5

2

Fig. 5.6

There are well-described genetic obesities in experimental animals, but the mechanisms appear to be different in humans. To cause obesity there needs to be an increase in appetite without a compensatory rise in energy expenditure. The ob gene codes for a peptide hormone, leptin, which is released from adipose tissue as a long-term signal of fat storage. Leptin signals satiety via hypothalamic receptors modifying neuropeptide Y activity, and as a signal to other functions, such as the onset of puberty, and tends to downregulate energy expenditure. Overweight humans have high circulating leptin concentrations and may therefore be resistant to it, although a complete absence of leptin, usually from a very rare defect in the ob gene, results in colossal childhood obesity. Several other genetic obesities are known, such as the Prader-Willi syndrome, with a defect on chromosome 15. No common gene defect has yet been discovered in the general population of obese patients. Studies of families, twins and adoptees all point to a strong genetic influence, accounting for about 30% of variability in weight gain. Multiple genetic factors are involved, with no single common pathway controlling energy balance yet identified.

Consequences It is often overlooked that overweight people start to develop symptoms (Table 5.21) as BMI increases above 25kg/m2. Most people with BMI >30kg/m2 are chronically troubled by symptoms, including tiredness, breathlessness, sweatiness, low back pain, and hip and knee pain. Depression is a common complaint but is unlike typical endogenous depression, and is probably due in part to fatigue. Obesity is associated both with functional difficulties in simple domestic activities and with discrimination in every walk of life. This poor quality of life is only exceeded by the worse quality of life for an obese patient on a protracted diet without visible results. The metabolic effects of overweight, leading to increased risks of heart disease (hypertension and hyperlipidaemia) and cancers (related to increased adipose tissue aromatase and increased free sex hormone) are well documented. These metabolic hazards of being overweight are exaggerated in those who also have a central (appleshaped) fat distribution, with a high intra-abdominal fat

TABLE 5.21 Symptoms of overweight • • • • • • •

Tiredness Sleeping poorly Sweating Breathlessness Snoring Low back pain Hip and knee pain

• Hirsutism • Menorrhagia • Amenorrhoea • Stress incontinence • Angina • Depression • Infertility

TABLE 5.22 Medical consequences of overweight and obesity Physical symptoms

Metabolic problems

Social problems

Anaesthetic/surgical problems

Endocrine problems

Psychological problems

Tiredness Breath lessness Varicose veins Back pain Arthritis Oedema/cellulitis Sweating/intertrigo Stress incontinence

Hypertension NIDDM Hepatic steatosis Hyperlipidaemia Hypercoagulation IHD and stroke

Isolation Agoraphobia Unemployment Family/marital stress Discrimination

Sleep apnoea Chest intections Wound dehiscence Hernia Venous thrombosis

Hirsutism Oligomenorrhoea/ infertility

Low self-esteem Self-deception Cognitive disturbance Distorted body image Depression

Metromenorrhoea Oestrogen dependent cancers: breast, uterus, prostate

a

IHD, ischaemic heart disease; NIDDM, non-insulin dependent diabetes mellitus

mass shown by high waist circumference. Table 5.22 shows the range of metabolic and mechanical consequences of being overweight. They have an impact in virtually every branch of medical practice. © The extra medical costs of the overweight are estimated to account for an astonishing 2-80% of total health service budgets. For this reason the UK Government Health of the Nation White Paper made the prevention and correction of obesity one of the highest priorities for the health service in the 1990s. Most Local Health Authorities have programmes for weight management. The World Health Organization has included obesity as a worldwide epidemic needing coordinated preventive measures, as well as a disease in its own right which requires treatment to relieve distress.

Criteria for treatment The management of weight problems should address the process of weight gain. Many people say they want to lose weight: most are able to do so using a variety of approaches, but many are not prepared to adopt the permanent lifelong change in diet and lifestyle needed to prevent regain. The groups of patients who need professional (medical) weight management are those who are unable to control a weight problem using self-help methods, who are prepared to make substantial lifestyle and diet changes for relatively modest cosmetic gains, and those who should be advised on medical grounds to lose weight. People with BMI below 25kg/m 2 cannot be considered overweight, and so weight loss is usually inappropriate. A request for help from these people may indicate an underlying eating disorder, such as anorexia nervosa. There may be patients with BMI a little under 25 when some weight loss could be valuable, for example those with metabolic syndrome or non-insulin dependent diabetes. Advice on weight management should be offered to all patients whose BMI exceeds 25kg/m2, aiming first at restricting further weight gain and second at a period of weight loss, followed by a period of supervised weight maintenance. Between BMI 25 and 30, the principal aim

should be the further weight gain. Those who wish to lose weight or those who have a high risk (e.g. through family history) of developing diabetes, heart disease, hypertension or hyperlipidaemia, and those who already have these conditions, should be given advice on how to lose weight. Above BMI 30, where disease risks are high, weight loss will bring benefits in relation to metabolic and cardiovascular problems and, more importantly in older patients, from the point of view of symptoms and impairment of quality of life. With a view to health promotion for the general public the waist circumference provides more accessible criteria for treatment. Above waist 80cm (women) or 94cm (men) the risks are increasing and people should avoid further gain. Above 88cm (women) or 102cm (men) health impairment is almost universal, and professional help should be offered (Table 5.23).

WEIGHT MANAGEMENT Recognizing the aetiology of obesity, and its major medical consequences, it is vital that doctors take a leading role in advising patients about the need for and methods of weight management. The skills of several professional groups can be valuably employed, including nurses, dietitians, clinical psychologists, physiotherapists, and even commercial slimming and exercise clubs. Weight management is best conducted in the primary care and community environment, but up-to-date training should be provided for all those involved. All patients with weight problems require assessment for other related and compounding risk factors, such as smoking, blood pressure, lipids and blood glucose. Although many smokers gain some weight when they stop, this weight is usually lost over the subsequent year or so and the gain can be prevented by simple dietetic advice. The medical hazards of smoking are only equalled by weight gain equivalent to a doubling of body weight, but the combination of smoking and overweight has multiply

129

TABLE 5.23a Waist watchers - Men Action level 2 Action level 1 <94cm >102cm 94-102 Increasing health High health risk Green zone risk -aim for - seek advice - no need to green zone to lose weight lose weight Prevalence (%) of Type 2 diabetes Cardiovascular risk factors Shortness of breath Poor physical functioning Chronic low back pain

0.6

1.8

26.0 10.4

47.4

4.5 64.2

17.1 10.2

31.5 17.5

19.5

21.1

4.8 14.1

TABLE 5.23b Waist watchers - women Action level 1 Action level 2 <80cm 80-88 >88cm Green zone Increasing health High health risk - no need to risk - aim for - seek advice lose weight green zone to lose weight Prevalence (%) of Type 2 diabetes Cardiovascular risk factors Shortness of breath Poor physical functioning Chronic low back pain

0.4 13.6 17.4 7.1 16.5

active effects on risks. Serum y-glutamyl transferase (y-GT) should be checked, as alcohol consumption is often an important component of weight gain in men, and it is often difficult to identify overweight drinkers. It should be remembered that y-GT will be elevated in untreated non-insulin dependent diabetes mellitus or metabolic syndrome with fatty liver. Serum TSH is useful, as subclinical hypothyroidism is common and greatly impairs ability to lose weight. Treating hypothyroidism does not automatically lead to weight loss.

Diet modification

130

A first step towards weight management is healthful balanced dietary advice. Many of the features of current health promotion dietary advice are highly appropriate to weight loss or weight maintenance. Even during slimming, patients should eat at least five portions of fruit and vegetables daily. Extra helpings, and even snacks of vegetables, can be encouraged. This will help to reduce fat intake. Fat is the nutrient with the highest energy content (kcal/g), so its major dietary origins need to be restricted, particularly full-fat milk and dairy products, processed meat products such as sausages and pies, and biscuits and baked

0.5 26.0 25.5 10.0 18.0

2.7

42.5 39.4 19.7 24.1

food. Bulky and relatively nutritious foods such as bread, potatoes and pasta can still be taken in reasonable amounts to satisfy appetite; it is the spreads or sauces that go with them that tend to contain unnecessary calories. The style of eating recommended for weight management should be continued for life. It should be explained to patients that they have a metabolic profile which will present a problem if they revert to eating a diet high in fat. Alcohol contains 7 kcal/g, so alcohol restriction is directly valuable, but alcohol consumption rapidly leads to non-compliance with diets. It is best avoided during periods of weight loss, particularly by drivers.

Eating behaviour Many patients with weight problems have a rather undisciplined eating pattern, often skipping meals such as breakfast, and even lunch, and then eating large amounts later in the day and at night. A useful management strategy is to concentrate on shopping, to advise patients not to buy what is not needed. It is helpful to plan eating patterns in advance, based on three meals a day and complete avoidance of all snacks. A breakfast which is filling and nutritious but low in calories, such as porridge, is effective.

SUMMARY 4 Practical advice for weight management • • • •

Become a non-smoker (top priority) Eat at least five portions fruit and vegetables daily Eat fish at least twice a week 20-30 minutes moderate activity daily (to breath hard or break sweat) • Avoid future weight gain (before and after loss) • Achieve weight loss 5-10 kg in 3 month stages • Feel proud to have done each of these

Porridge and 'breakfast cereals' are useful as standbys for energy-saving evening meals, and are preferable to lowcalorie commercial meal replacement products, which are often expensive and unpalatable. Individual strategies can be discussed, negotiated and used with patients. 'Rules', such as only eating off a plate, never eating with the fingers, or never eating with the television on, may be helpful.

Physical activity It is extremely difficult to lose weight simply by increasing exercise and physical activity. Each kilogram of adipose tissue (fat) contains about 7000 calories, so 7kg (a stone) contains 50000 calories. Sustainable increases in physical activity would need to be maintained for well over a year to lose this amount of weight. On the other hand, physical activity can be enormously valuable in improving quality of life and helping people to feel better, and in reducing blood lipids, blood pressure and the risks of diabetes. Increased physical activity is vital for weight maintenance to avoid regain, and it is essential to protect muscle from wasting during weight loss. Regular physical activity may help regulate appetite. The metabolic rate of a patient who has lost 7kg (one stone) will be reduced by about 120kcal/day. This can be matched by 20-30 minutes of brisk walking daily if the patient were to continue to eat the previous intake. It is useful to advise the patient to trade food at snack time for activity, even of a very limited extent. Negotiated plans should be agreed, for example to walk all journeys which are less than one mile. Organized physical activity is helpful for many patients but seldom conducted on a daily basis. For those unable to take some exercise daily, it is acceptable to undertake longer blocks of exercise less frequently (e.g. 1 hour three times a week). In addition to taking specific exercise, the avoidance of complete inactivity is important. Television watching (average 29 hours per week in the UK) relates closely to weight gain and to difficulty in slimming, even in people who take some exercise.

Drugs The main function of drug therapy for obesity is to restrict long-term weight gain below the untreated rate. Weight loss is a brief (3-6-month) phase, but no available drug can produce major weight loss alone, and a drug that did would

be dangerous. Initial weight loss is probably most efficient using a very low-calorie diet, under supervision, with drugs for the long-term maintenance phase. No drug therapy can achieve weight loss or maintenance at a lower level without dietary and physical activity changes, and so drugs should not be prescribed unless patients are achieving success with dietary and lifestyle modification as an indication of motivation. One of the main aims of using drugs for weight management is to reduce the need for drug therapy for complications of obesity, such as antihypertensive drugs or hypoglycaemic agents, lipid-lowering drugs and analgesics. These preparations are usually taken indefinitely and often in combination. There is no place in weight management per se for the use of diuretics: oedema is usually a local problem, not the result of sodium retention, and responds poorly to diuretics. Thyroxine does not produce weight loss except in large doses with dangerous cardiac side-effects, and should not be used for obesity. With recent recognition of the extent of pathology and healthcare costs attributable to obesity, there has been intense research interest in developing new drugs for weight management. Drugs may affect appetite or satiety and energy intake, or energy expenditure. Because of the physiological complexity of energy balance, no single drug is likely to have profound effects, and responsiveness varies widely between subjects. Ideally, to reduce body weight, energy intake must fall and energy expenditure must rise. Most effective agents have some effect on both sides of the energy balance equation, but pharmaceutical management will always rely on dietary restraint. The old appetite suppressant drugs, based structurally on amphetamines and sharing noradrenergic releasing activity (e.g. phentermine), have fallen out of favour, largely because of the historic association with amphetamine, despite their lacking addictive or mood-altering actions. They were never subjected to long-term clinical trials, but undoubtedly led to substantial weight loss in patients otherwise unable to achieve it. These drugs are still available in the USA and elsewhere. Sibutramine, available in most of America and Europe, has both noradrenaline and serotonin reuptake inhibition activity. Patients who respond (2 kg weight loss in 4 weeks) can expect to lose about 10 kg on average over 3-6 months, and to maintain the differences from placebo for at least 2 years. Greater weight loss, achieved by an initial period on a very low calorie diet can also be maintained. The main side effects are increase in pulse rate and in blood pressure. In patients who fail to lose weight blood pressure tends to rise, but blood pressure falls with weight loss on sibutramine, particularly in hypertensive patients. About 10% of patients do not lose weight with sibutramine. They were included in the clinical trial results, but should be identified early in a therapeutic trial in routine practice. Orlistat aids weight loss and maintenance by inhibition of intestinal lipase, leading to 30% fat malabsorption. Patients who lose 5% body weight after 3 months treat-

131

ment, together with a 30% fat diet, can expect 10-15% weight loss, maintained for 2 years. However, only 30-40% of patients are able to pass this hurdle. The main sideeffects are symptoms of fat malabsorption if patients eat too much. The drug is not absorbed. A wide variety of other weight loss agents are prescribed in other countries, or sold over the counter in pharmacies and supermarkets. For none of these is there evidence of safety or efficacy, with the possible exception of an ephedrine and caffeine combination, widely used in Europe, but not in UK.

Surgery In patients with BMI above 35kg/m2 and where metabolic or other complications are crippling or life-threatening, surgery is sometimes justified. The most widely used, operations at present are vertical banded gastroplasty and laparoscopic gastric bonding which reduce the size of the stomach to only about 50ml. Gastroplication requires the backup of a high dependency unit, but has a relatively low complication rate in experienced hands. Following operation, patients need to follow a liquid-only diet for 3 months, otherwise the staple line is likely to rupture and the surgery will be ineffective. It is a good plan to insist that all patients spend 3 months on the liquid-based diet before operation; during this time they should be expected to lose approximately 15kg in weight. Supplementation with iron and vitamins is necessary, and long-term medical nutritional and dietetic supervision is required. This form of operation usually results in a deranged eating pattern for life. For this reason psychiatric or psychological assessment and subsequent supervision should be available. Weight loss of 30-100 kg can be achieved but because regain after 2-3 years is so common gastric bypass with Roux-en-Y is becoming preferred. Long-term weight loss is more assured, but at the cost of chronic loose stools.

1 kg of fat in a week through strict dieting, but to gain the same amount temporarily from water, and this can be demoralizing.

Weight maintenance Very few patients can continue to lose weight beyond 3 months, and it is unreasonable to ask them to attempt more. Any result other than weight gain should be regarded as successful at 3 months, and then a weight maintenance programme introduced to prevent regain. This follow-on programme should include active support and lifestyle and psychological counselling for a further 3 months. Commercial and lay slimming groups may provide valuable support. Where the medical risks of weight gain are specially high (e.g. active, or family history of diabetes, hypertension, coronary heart disease), long-term medication may be considered. 1 2

NUTRITION SUPPORT Malnutrition delays wound healing and the recovery of muscle tone after an injury or operation, and may increase susceptibility to infection. Adults with a BMI below 16 have three times the postoperative mortality of betternourished patients. The striking changes in body composition from partial starvation in illness are shown in Figure 5.8. It is better to avoid malnutrition than to treat it, and it is essential to make a nutritional assessment of every patient. If the patient is already underweight, or if there is weight loss and the clinical condition gives concern, some form of nutrition support - such as dietetic supplements, enteral feeding or parenteral feeding - can be used to prevent further deterioration (Table 5.24). Every hospital should have a nutritional assessment and support policy, with a nutrition team comprising a physician and/or sur-

Expected rate of weight loss Many patients, and indeed health professionals, have unrealistic expectations during weight loss. Following a strict dietary restriction of 500 calories a day below normal maintenance, patients will lose about 0.5kg (1 Ib) per week or 7kg (1 stone) in about 3 months. More restricted diets and faster weight loss can only be achieved by a minority and for a fairly short period. More rapid weight loss may be dangerous, as it implies excessive loss of muscle or other body tissues. Weight loss of 0.5-1 kg per week should be expected and encouraged. Patients should not weigh themselves more often than every 2-4 weeks, as bathroom scales give a poor guide to body weight. It is also possible to lose

1

132

Case 5.2

2

MCQ 5.3

FIG. 5.8 Body composition of a man in health and after partial starvation

TABLE 5.24 Examples of conditions in which nutrition support may be necessary Illness associated with nausea and abdominal pain Gastrointestinal disease Malignancy Chemotherapy

Inability to eat Coma

Weakness Bulbar/pseudobulbar palsy

facing malnutrition, a formal prescription process for nutritional supplements may be as important as that for drug administration.

a

Intestinal disease

Enteral feeding Maiabsorption Inflammatory bowel disease Fistulas Thoracic duct damage Ischaemic colitis Short bowel syndrome Increased requirements Burns Sepsis Trauma

geon with special experience in clinical nutrition, a nutrition nurse/technician and a dietitian, together with strong support from the pharmacy, biochemistry and pathology laboratories.

Food and sip feed supplements Many patients find hospital food unappetizing even when they feel relatively well. In such cases, it may be possible to increase energy and nutrient intake by snacks provided by the diet kitchen between main meals. Attractive food and company can contribute greatly to maintaining nutrition; impersonal meal delivery and having to eat in bed have negative influences. If solid food is not well accepted, it is often possible to give liquid feeds for sipping but energy intake can be compromised. Semi liquid feeds can be prepared for patients with swallowing difficulties. A considerable range of commercially prepared feeds are available, presented as 200ml cartons. Some are complete feeds, containing a balance of all essential nutrients in proportion to requirements; others are incomplete, such as flavoured sugar solutions. Palatability is very variable. Sip feed supplements and enteral feeds are available in a vast range of compositions. For virtually all patients, the macronutrient contents should be similar to that of a healthy balanced diet, i.e. about 10-15% energy from protein, 50-60% from carbohydrate and 30-40% from fat. There is no clinical indication for high protein feeds: preparations with up to 40% energy as protein are available but should not be used as they suppress appetite and may lead to impaired renal function. Evidence is very limited to justify the use of 'elemental' feeds containing pre-digested peptides or oligosaccharides, or for feeds enriched with particular amino acids such as glutamate. Sip feed supplements should be served chilled and their volumes recorded accurately. They must be taken as supplements to an optimized food intake, not used as meal replacements. Patients often need a nurse to help and encourage supplement consumption. For patients

If oral intake is not satisfactory usually because of swallowing problems, a fine-bore silastic nasogastric tube may be passed. Because the enteral route should always be used when the gastrointestinal tract is intact, surgeons should also be encouraged to place a feeding nasogastric or nasojejunal tube at the time of any upper gastrointestinal surgery, or to leave a feeding jejunostomy. Even very limited enteral feeding after surgery or in the ITU nourishes the intestinal mucosa, reducing bacterial overgrowth and translocation, and permits more rapid weaning on to full enteral feeding. Enteral feeding solutions are passed down the tube at controlled rates. Small battery-operated pumps can control the exact flow rate, but they are not essential. The feed should be started in small volumes, and no more than 1000 mL given in the first 24 hours, otherwise diarrhoea may develop. The volume is then increased to the desired amount (probably a maximum of 2500mL/day). When patients are beginning to take food, the feed should be stopped for 1 hour before meals. Overnight enteral feeding may usefully supplement a limited oral intake, but care must be taken to avoid aspiration of gastric contents. A semirecumbent position is advisable. A large range of commercially available enteral feeds are available, made up from a variety of protein sources, including casein, soya and mixtures of amino acids, with glucose polymers and triglycerides added to increase the energy content. Most feeds provide about Ikcal/mL and have an osmolality of 280-420 mmol/kg. Low-lactose and cows' milk-free feeds are available. The commercial preparations have micronutrients in amounts that match dietary reference values, and so additional vitamins and minerals may be needed for patients whose micronutrient status is poor, or who have increased losses, utilization or malabsorption. Biochemical monitoring is required. Energy provision should match the metabolic rate, measured by indirect calorimetry or predicted from standard equations (Table 5.4) (BMR x 1.3 for a bedbound patient, plus 10-50% according to degree of illness or need for weight regain). It is impossible to overcome the increased metabolic rate of an ill, infected patient and to promote muscle regrowth at the same time. Protein should account for 10-15% of energy, the remainder coming in approximately equal amounts from carbohydrates and fat. The use of oligosaccharides or polydextrose may reduce the osmotic effects of glucose delivered enterally, but the evidence for using expensive peptide feeds is scanty. Enteral feeding can be performed at home; the reservoir bottles and lines should be changed every 24 hours to avoid the risk of contamination of feeds, but the silastic tubes can remain in situ for several weeks. Even so, complications of enteral feeding occasionally occur and should be anti-

133

cipated. Some patients prefer to use enteral feeding intermittently and become expert at introducing nasogastric tubes. Oesophagitis caused by the tubes is rare with modern silastic tubing, but the tube may be misplaced and a clinical check (aspiration of gastric fluid, or detection of bubbling with a stethoscope) should be carried out before feeding is commenced. When enteral feeding needs to be continued beyond a few weeks, gastrostomy feeding is usually easier and more acceptable to the patient. Percutaneous endoscopic gastrostomy is now a simple and safe procedure, increasingly used in a range of medical conditions, including pseudobulbar palsy following stroke, advanced multiple sclerosis or other neurological disorders. The most common complications of enteral feeds are diarrhoea, abdominal distension and pain, all of which relate to excessive infusion; the volume infused should then be decreased. More protracted diarrhoea can be controlled by codeine phosphate (30-60 mg t.d.s.) or imodium (2-4mg t.d.s.). Aspiration pneumonia is relatively frequent in debilitated patients (especially those with swallowing disorders) receiving enteral feeding when lying flat. Such patients should be fed sitting, if possible.

Parenteral feeding If enteral feeds cannot be given, parenteral feeding is an option; however, the dangers of infection, thrombosis and metabolic problems with this method are considerable. Enteral feeding should be commenced as soon as possible (even if very limited) to prevent mucosal atrophy. Parenteral feeds are ideal culture media for bacteria. They must be made up under sterile laminar flow conditions by specialist pharmacists (the vitamin and mineral preparations have limited stability, so feeds must be made up shortly before use). A dedicated intravenous line must be inserted under sterile conditions and only handled by trained staff. -Long-term parenteral nutrition requires a central line, as the high osmolality (often l000mmol/L) causes thrombosis in slow-flowing peripheral veins. For shorter-term feeding (up to 10-14 days), perhaps before it is ascertained that long-term parenteral nutrition will be required, peripheral veins can be used. Sterile techniques should still be used, and blood flow can be increased by a distal GTN patch. A lower glucose content allows reduced osmolality and so reduced thrombosis: a limited energy supply is often acceptable for 10-14 days for peripheral line feeding. Some commercial peripheral feeds are incomplete in terms of micronutrients, and so monitoring is important. The energy requirement for a malnourished patient receiving nutritional support is often difficult to assess. For long-term feeding, energy requirements can be estimated from standard equations published by the WHO (Table

1

134

MCQ 5.4

5.4). For ill patients and the malnourished, measurement of resting metabolic rate by indirect calorimetry is valuable, largely to avoid overfeeding, which causes a fatty liver. The amino acid content should be 10-15% of energy, and the remainder given as glucose or lipid emulsion (which permits a lower volume and lower osmolality than with glucose alone). A standard single 24-hour bag (2-3 L) containing 1500-2500 kcal is available. Parenteral feeds can be given either continuously over 24 hours, or over 12-18 hours to allow a daytime period off feeding, when the line is sealed and the patient can be mobile. Micronutrient requirements depend on clinical condition. All the essential nutrients (amino acids, fatty acids, vitamins, trace elements) are required. The matching of provision to need is more critical than with enteral feeding, because the 'buffering' influence of absorption in the gut does not apply. If the patient is malnourished and tissue growth is required, a greater than normal provision of phosphate, potassium and zinc will be needed to avoid 'refeeding syndrome' deficiencies. Because it is difficult to exclude subclinical vitamin B1 deficiency, extra thiamine should be given to patients when starting parenteral nutrition. For some, the intravenous delivery results in more rapid renal excretion, and so greater than normal dietary amounts are needed. Many patients receiving parenteral nutrition have tissue damage, and renal failure is relatively common, demanding special consideration. A number of metabolic complications may occur in patients receiving parenteral nutrition: • Hyperglycaemia is common in patients with severe surgical stress (e.g. burns and intravenous trauma) and sepsis, but plasma glucose can be controlled with carefully monitored insulin, adjusted to maintain plasma glucose 4-8mmol/L. Rebound hypoglycaemia may occur if intravenous feeding is abruptly halted. • Metabolic acidosis is rare if fructose and excessive quantities of amino acids are avoided. • Parenteral feeding, especially with lipid solutions, may cause hepatic abnormalities. The commonest reason is overfeeding, leading to fatty liver, which responds to a reduction of energy in feeds. However, most patients receiving parenteral feeding have many other potential causes for hepatic damage, and there is rarely any indication for parenteral nutrition to be stopped on account of mild biochemical changes. • Fat is metabolized more favourably than glucose by patients with sepsis. • Patients with respiratory failure should not receive large glucose loads because of their impaired ability to excrete the extra CO2 produced. • Mineral and electrolyte abnormalities occur variously according to the underlying disease processes. Fortunately, deficiencies can be made up relatively easily during preparation of the bag that contains the feeding solutions. • The stability of feed constituents is precarious, and problems have arisen with 'home-made' feeds and with

CASE STUDY 5.2 POSTOPERATIVE MALNUTRITION A 54-year-old man, height 1.75m, weight 65kg (BMI 21kg/m2) underwent a revision of a mitral valve replacement. His postoperative course was complicated by mediastinal wound infection and he had an episode of pneumonia, hypotension and acute renal failure, necessitating prolonged ventilation. During recovery the question of his nutritional status, complicated by 7 days of infection is raised at a caseconference, 10 days postoperatively, at which stage feeding had not been commenced. His weight has fallen to 60kg. Questions 1. What nutrient deficiencies are likely to be present? 2. How could management be optimized? Discussion Failure to eat for 10 days postoperatively, when metabolic rate is elevated about 50%, will have resulted in negative energy balance of about 10 (days) x 1.5 (increased metabolic rate) x 1600 (minimum daily requirement) = 24000k/cal. This would be equivalent to almost 4kg adipose tissue, but excess lean tissue is lost in infection, at about l000kcal/kg. The weight

measurement is likely to be seriously confounded by oedema, and changes in fluid balance (which may not be dependent). The patient's 'true' BMI may therefore be as low as 16kg/m2, making death from malnutrition likely. Specific deficiencies of many micronutrients are likely. Blood biochemistry is likely to be difficult to interpret owing to acute changes in protein and electrolyte balance. Deficiencies of nutrients which do not have body stores will be particularly depleted (e.g. vitamin C, B complex, magnesium, zinc, selenium). Preoperative status of thiamine, selenium may have been marginal thereby potentially impairing cardiac function. Immune function may have been impaired by a variety of nutritional factors, including a deficiency of essential fatty acids, vitamin A and zinc. Preoperative nutritional assessment and optimization should be routine. Common hazards with late, and inexperienced, nutritional support include inadequate thiamine to overcome a deficiency state, and the overprovision of energy. During illness the capacity to utilize nutrients and to synthesize body tissues is reduced, partly from cytokine action, partly from secondary nutrient deficiencies and altered nutrient essentiality. Excess energy supply

storage. Micronutrients should be added shortly before infusion. Home parenteral nutrition is occasionally necessary for patients with postsurgical short bowel syndrome and other conditions. It represents an enormous undertaking, but with family, social and medical support can be continued satisfactorily for many years. Commercial organizations exist to deliver supplies and to provide nursing and pharmacy support. In many cases total parenteral nutrition is not necessary - more limited feeding, e.g. 3 or 4 nights per week, may be sufficient. 1

Monitoring nutrition support The measurements available for monitoring support are shown in Table 5.25.

nutrition

leads to acute hepatic steatosis and impaired liver function. Measurement of metabolic rate is possible using a ventilated hood calorimeter, but it is more difficult when the patient is on a ventilator. Using estimated BMR is usually adequate. At this stage this patient requires assessment and management by a multidisciplinary Hospital Nutrition Support Team (including representation from specialist trained nursing, pharmacy, dietetic, biochemical and clinical staff). Biochemical nutritional data should be assessed at least weekly, with daily weight and assessment of state of hydration and for oedema. If the GI tract is accessible and intact, then this route is preferable to intravenous feeding. Because of mucosal atrophy after 10 days without feeding, enteral feeding will need to commence at low rates, increasing to approach energy requirement after 2-5 days. A nutritionally complete feed is required. Evidence does not support enrichment with individual amino acids. An approximate breakdown of energy 15% protein, 55% carbohydrate, 30% fat is appropriate. If increased CO2 production is clinically problematic, then higher fat, lower carbohydrate feeds might be considered.

a

TABLE 5.25 Measurements for monitoring nutrition support • Anthropometric: weight, mid-arm muscle circumference (MAC) • Minerals and electrolytes: uric acid, phosphate, sodium, potassium, zinc, magnesium, iron • Resting metabolic rate • Plasma vitamin concentrations • Nitrogen balance: corrected 24 hour urinary urea excretion • Glucose

Regular measurements of weight are necessary but are difficult to perform in the very sick without special bed balances. Changes in body weight may indicate alterations in fluid balance rather than an increase in lean body mass.

135

Skinfold thicknesses and mid upper-arm circumference are valuable but affected by oedema. Plasma proteins (CRP, albumin) are useful indicators of progress in treating the underlying disease processes. Glucose should be measured regularly. Plasma phosphate and potassium levels may decrease rapidly and even precipitate cardiac arrhythmia if nutrition support is given rapidly in large quantities to a severely malnourished subject who has previously received very little dietary intake. Most modern amino acid solutions, however, have sufficient phosphate to avoid this. A low plasma sodium is often present in severely malnourished subjects. Total body sodium, however, is often high, and extra sodium should only be given with great caution. An approximate assessment of urinary nitrogen loss can be obtained by measurement of the 24-hour urinary urea excretion, corrected according to the formula: Nitrogen (g/24h) = urinary urea (mmol/L) x urine volume (L) x 0.028 x 6/5 This formula allows for the conversion of mmol of urea to g of nitrogen, and corrects for the fact that urea accounts for approximately five-sixths of total urinary nitrogen loss. Parenteral feeding solutions are ideal culture media. The previously high prevalence of infection has been reduced by the single-bag system, but a high standard of nursing and medical care of the lines is essential if sepsis is to be prevented. Different types of intravenous catheters are available - the best are those that are tunnelled subcutaneously. The feeding line must not be used for any other purpose. If other infusions are required a multilumen Hickman line may be used, keeping one lumen sterile for parenteral feeding. Hickman line insertion requires skill and experience. Complications include pneumothorax, central vein thrombosis, nerve injury, haematoma and arrhythmias. Similarly, only staff trained in the care of central venous feeding lines should be allowed to touch them, and the need for aseptic changing of sets and meticulous care of the catheter entry site should be emphasized. Feeding line infection should be suspected if fever develops, and is confirmed by cessation of fever when feeding is withheld and by a drawback blood culture (conducted under sterile conditions). Infected feeding lines can often be rescued by stopping feeding and by infusing vancomycin for 4-5 days. Central vein thrombosis can be treated by streptokinase infusion, but complete sterility must be maintained if the line is to be reused for feeding.

Refeeding syndrome

136

A lesson learned from work in the developing world, and now applicable in western hospitals, is that refeeding of severely malnourished and ill patients can often precipitate rapid decline and death. The 'refeeding syndrome' relates to the dynamic changes induced by introducing nutrition support. Metabolism is reactionary to rapid shifts from

extra- to intracellular spate of potassium, magnesium, zinc and phosphate. Low plasma levels can then lead to cardiac and central nervous system problems. These shifts need to be anticipated, and clinical problems can be avoided by careful monitoring and extra provision of these nutrients during early refeeding.

NUTRITION AS A COFACTOR IN DISEASE AND IN THERAPY

Hypertension There are direct relationships between BMI, sodium intake and blood pressure. Humans evolved on very low-sodium diets. The main effect seems to be a programming effect from exposure to high sodium intake in infancy. Most patients with 'essential hypertension' are in fact obese, and the increased sodium intake is secondary to increased food intake. There is a threshold effect of additional salt in adult volunteers, with those on very high doses showing an increased blood pressure. However, there are also important cofactors, including family history, stress, obesity and potassium, calcium and magnesium intakes, which may obscure the relationship in an individual. In the UK the average daily salt intake in adults is 5-10 g, i.e. sodium intake (150-250mmol). This is more than twice the amount associated with lower rates of hypertension experimentally, and 10 times more than is required daily. Salt restriction and weight loss should always be the first steps in hypertension management. Halving sodium intake to 80-100 mmol/day has an effect similar to that of routine antihypertensive drugs (5-10mmHg) in hypertensive patients, but without the side-effects. This reduction would lower blood pressure by 25mmHg immediately for normotensive people, and would prevent a l0mmHg rise 30 years later. A weight loss of 5-10 kg will also reduce blood pressure by 3-5mmHg. Greater effects are seen during acute energy restriction, but blood pressure tends to return to pre-weight loss figures over a period of years.

Atheroma International comparisons of mortality rates from coronary heart disease (CHD) show that these are highest in populations with the highest saturated fat intakes. This is true even when the effects of obesity, hypertension, cigarette smoking, diabetes, stress and family history are allowed for. The dietary factors leading to coronary heart disease are shown in Figure 5.9. The major lipid risk factors for the development of atheroma are high levels of plasma cholesterol and lowdensity lipoproteins (LDL), and low HDL. The pattern of high triglycerides, small dense LDL and low HDL is particularly atherogenic, and is associated with insulin resistance and non-insulin dependent diabetes in the 'metabolic syndrome', characterized by a high waist circumference

Environmental factor

Metabolite or process affected

Pathophysiological change

Pathology

Disease

FIG. 5.9 Factors leading to coronary heart disease

and intra-abdominal fat accumulation. Consumption of saturated fatty acids is particularly likely to increase both LDL and cholesterol. Diets low in saturated fatty acids lower these levels. Saturated fats are prominent in butter, lard, sausages, meat pies, pastries and chocolates, whereas unsaturated fats predominate in oils or margarines prepared from olives, rape seed, maize, sunflower or soya. Not all margarines are unsaturated - those that contain coconut oil (used in many hospital canteens in the UK), for example, are highly saturated. Synthetically hydrogenated fats

contain high amounts of trans isomer fatty acids, which behave like saturated fats. Plasma cholesterol mainly indicates the production of cholesterol by the liver. Cholesterol-containing foods (such as eggs) do not necessarily increase the synthesis of cholesterol in the liver. Cholesterol absorption can be reduced 10% by plant sterols (sitosterol and sterol esters) used in non-functional foods. Dietary antioxidants (e.g. vitamins A, C, E, flavonoids in fruits and vegetables) reduce atheroma by protecting LDL from oxidation. Fibrinogen is elevated in the overweight, as are coagu-

137

lation factor VII and plasminogen activator inhibitor (PAI-1). These are all predictors of thrombosis, and other dietary factors also have an influence (e.g. dietary fat on factor VII). Certain n-3 polyunsaturated fats - such as eicosapentanoic acid, present in cod liver oil, mackerel and pilchards - tend to inhibit thrombosis and should be particularly encouraged. People who eat a lot of fish have less heart disease. Major dietary changes in the population of Finland between 1970 and 1990, increasing fruit and vegetables and fish, and reducing sodium and saturated fat, together with reduced smoking, have been accompanied by 60-70% reductions in CHD. Intentional weight reduction in overweight or obese subjects, particularly in diabetic patients, is followed by a reduced risk of mortality from CHD and an increased life expectancy.

Diabetes Insulin-dependent diabetes is not principally linked to diet in its aetiology, although prolonged breastfeeding seems to be protective. The major need in management is to provide regular meals and snacks, every 2 hours or so, to match the profile of insulin action and to avoid hypoglycaemia. Alcohol impairs glucose release from the liver and thus potentiates any risk of hypoglycaemia. Non-insulin dependent diabetes (type II) is closely linked to diet. In nearly every society, those who are overweight and whose diet is high in fat and low in soluble dietary fibre have a significantly increased risk of diabetes. The greatest risk is in those who were born small for dates and show poor growth in the first year of life. They seem specially prone to weight gain and diabetes. The major hazard in non-insulin dependent diabetes is premature heart disease and stroke. Weight loss improves the prognosis for the overweight, and physical activity probably plays a major part. For all patients with diabetes, the need to follow a low saturated fat diet, with a high fruit and vegetable intake, is even more pressing than for people without diabetes. Sugar avoidance is no longer considered necessary, but general recommendations for a healthy diet apply (Table 5.26). There is sufficient circumstantial evidence to urge high intakes of antioxidant-rich fruits and vegetables to protect against microvascular disease. There is no benefit from the use of so-called 'diabetic' foods, which exploit a vulnerable group.

The metabolic syndrome

138

The 'metabolic syndrome' describes a clinical context which predisposes to premature CHD and stroke by a variety of coexisting abnormalities, including hypertension, dyslipidaemia (low HDL, high triglycerides, high small dense LDL), type 2 diabetes and hypercoagulability. The clinical syndrome is characterized by a predominantly central fat accumulation (intra-abdominally) even in those

TABLE 5.26 Summary of nutritional recommendations for people with diabetes compared with current average UK diet Approximate content of usual UK diet

Recommendations for diabetes

Energy

maintains BMI 25 kg/m2

to approach BMI 22 kg/m2

Carbohydrate (% energy) Added sucrose or fructose (g/day) Dietary fibre (g/day) Total fat (% energy) Saturated fatty acids Monounsaturated fatty acids Polyunsaturated fatty acids Protein (% energy) Salt (g/day) normotensive hypertensive So-called 'diabetic' foods

45 50-100

50-55* <25

19-25 40

>30g 30-35* <10 10-15 <10 10-15

17 11

6 12-15

10 -

None

<6 <3 None (avoid)

* Lower amounts of energy from carbohydrates and up to 35% from fat may be acceptable provided that most of the fat derives from monounsaturated fatty acids such as olive oil in 'Mediterranean-style' diets.

who are thin - with a high waist circumference. The unifying biochemical feature is probably early insulin resistance, at least in tissues other than the intra-abdominal fat mass, and there seem to be a reduced activity of PPAR (peroxisome proliferator-activated receptor-y) enzymes. The metabolic syndrome is at least partly genetic, but is precipitated by weight gain and physical inactivity, by smoking and by high alcohol consumption. Another key element is related to poor fetal and infant growth (low birthweight and size at 1 year of age). This may relate to under- or imbalanced nutrition at critical stages of fetal organ development the 'Fetal programming' hypothesis.

Cancers It is estimated that malnutrition and cachexia play a part in 70-80% of all cancer deaths, and diet contributes to 35% of all cancer development. Obesity is associated with the development of colonic cancer and also the sex hormonerelated cancers (breast, endometrium and prostate). The WHO produced a summary of dietary influences on different cancers, revealing a rather consistent pattern among major cancers of associations with dietary fat and alcohol, and with protection from fruits and vegetables. There is some suggestion that high intakes of meat perhaps burned meat containing cyclic hydrocarbons may be hazardous. It is not certain which factors in fruits

and vegetables are protective, but antioxidants probably play a part. It is possible that vitamin E supplementation may help reduce cancers. On the other hand, studies with P-carotene supplementation have actually shown increases in lung cancer in smokers, especially if they are also heavy drinkers. Vitamin tablets should therefore not be taken, or prescribed, in doses far exceeding healthy dietary intakes. It is better to promote fruits and vegetables, which also contain other bioactive compounds, such as carotenoids and flavonoids, that probably also help to protect against cancer when consumed throughout life. Advice for cancer prevention is thus essentially the same as for reduction in heart disease.

Gallstones The deposition of cholesterol to form gallstones depends on the relative concentrations of total bile salts, lecithin and cholesterol. The fine balance between these is affected by obesity and low dietary fibre intake, but the precise mechanisms of the imbalance or of the formation of gallstones are not clear. Rapid weight loss may precipitate biliary colic.

Liver disease The most important association between dietary factors and liver disease is that between alcohol and cirrhosis. However, mycotoxins may also affect the liver: aflatoxins, which are widely prevalent in cereals kept under poor storage conditions, may be a factor in causing primary liver cell cancer and may also contribute to the development of the fatty liver in kwashiorkor. Malnutrition in children does not account for the high prevalence of cirrhosis in developing countries: hepatitis B infection seems to be the major cause.

Pancreatic disease Chronic pancreatitis is frequently found in populations where malnutrition is rife. Periods of undernutrition may cause stasis of the exocrine juices within the pancreas, leading to acinar cell destruction and loss of digestive function. Pancreatic calcification often follows, with secondary diabetes. However, in western societies alcohol excess seems more important than undernutrition as a cause of chronic calcific pancreatitis.

Respiratory disease Being overweight, and particularly having a central (intraabdominal) predominance of fat, causes breathlessness, and also aggravates symptoms in patients with COPD. Sleep apnoea and hypoxia is a potentially serious consequence, with snoring as an early symptom. Weight loss is very effective. A proportion of patients with COPD and chest bellows disease become severely cachectic. The chest problem is

aggravated by wasting of the respiratory muscles, and by malnutrition-related immunosuppression. The main cause is suppression of appetite by recurrent infection exacerbations. Hypermetabolism is a lesser factor. Weight gain is extremely hard to achieve by dietetic means, but can improve respiratory function: appetite is poor, and the effort of feeding worsens hypoxia. Starvation ultimately leads to death by pneumonia in most cases, and it has been suggested that malnutrition may contribute to the development of emphysema.

5

Renal disease Patients with renal failure accumulate potassium, phosphate and nitrogenous products from protein catabolism. Traditional conservative management included a diet restricted in protein potassium (limited meat, dairy products, fruit and vegetables). Former advice to avoid fruit and vegetables is now tempered by anxieties about reduced antioxidant status, as the main cause of death is cardiac disease. A daily limit of three items of fruit and two portions of vegetables (or vice versa) is acceptable for most patients. Most patients with advanced renal diet have poor appetites and tend to undereat. The dietetic emphasis should probably be on increasing starchy carbohydrates, rather than on restricting foods. During dialysis, small molecules including vitamin C are removed, and supplements should be given afterwards. Nephrotic syndrome and proteinuric diabetic neuropathy are aggravated by high protein intakes. Restriction to about 0.8-1 g/kg/day (10% energy) is appropriate and will still allow nitrogen balance. This amount is much less than usual in UK diets.

Dental caries Fluoride is an essential nutrient, required to form healthy dental enamel. Much of the UK, particularly Scotland, has a fluoride-deficient water supply (below Ippm). Dental caries occurs when Streptococcus viridens adheres to teeth with fluoride-deficient enamel and metabolizes sugar into lactic acid. This lowers the pH acutely and erodes the enamel. Low-pH foods can be similarly damaging. All sugars can act as substrate; thus 'natural' sugars such as fructose are just as cariogenic as sucrose. The most important factor is the length of time that sugar-containing foods remain in contact with the teeth; adhesive foods and sugary drinks given in bottles to infants are particularly damaging. The low pH of many fruit drinks and colas leads to rapid enamel loss. There has been a marked decrease in dental caries among children in certain social groups in the UK over the last 20 years, but increased social division within the population reveals very high rates in deprived areas and groups. The fall in caries has been associated with an increase in the use of fluoride (in toothpaste, in drinking water or as tablets), and possibly also an increased prevalence of teeth cleaning. Perhaps paradoxically, sugar con-

139

sumption has increased over this period, although very recently low-sugar products have become more prevalent.

TABLE 5.27 Foods associated with food Intolerance* Food

% of cases

Urinary calculi

Milk

Calcium or oxalate stones are sometimes precipitated in the urinary tract as a result of increased dietary intake. Oxalate stones are those most frequently associated with dietary factors, occurring most commonly in populations with high phosphate intakes. In Thailand, for example, oxalate calculi occur most commonly in the rural populations, where the high phosphate intakes, associated with a predominantly vegetarian diet, contrast with the lower phosphate intakes of the better-off urban families. Oxalate stones may also be precipitated by dehydration in extremely hot climates, and by excessive vitamin C consumption. 1

Egg

25 22

Nuts/peanuts Fish/shellfish Wheat/flour Chocolate Artificial colours Pork/bacon Chicken Tomato Soft fruit Cheese (but not milk)

12 12

6 5 3 3 3 3 3 3

*From. Truswell AS 1986 ABC of nutrition

FOOD SENSITIVITY AND INTOLERANCE

The clinical importance of food allergy is difficult to assess as little is known of the scientific basis for hypersensitivity to food, there have been few controlled clinical trials of exclusion diets, and the mechanisms by which they work are not understood. A variety of foods have been implicated in food intolerance (Table 5.27). There has been considerable public interest in peanut allergy. Reports of food allergy can lead to needless anxiety. Parental concern can lead to poorly designed exclusion diets and nutrient deficiencies. The great majority of self-diagnosed food allergies or food intolerances turn out to be erroneous when professionally assessed. However, dramatic, even life-threatening allergies do occur. The major clinical symptoms associated with food sensitivity are shown in Table 5.28. A number of food allergies are associated with immunological reactions - usually an IgE response. A radioallergosorbent test (RAST) may detect antibodies in plasma against a specific food protein, but there is little correlation between the clinical features and the results of a variety of RAST tests. Skin tests of purified antigens may detect an urticarial response within a few minutes of intradermal inoculation, but positive skin tests do not correlate well with sensitivity reactions in other organs, such as the lungs and the gastrointestinal tract. In short, there is as yet no defined mechanism in most cases of food sensitivity, and very few laboratory tests are of any use. Management therefore depends on careful history-taking and a critical approach to elimination diets.

1

140

MCQ 5.5

TABLE 5.28 Symptoms in 100 patients with food intolerance* Symptom

Number of patients affected

Asthma Gastrointestinal Eczema Urticaria Rhinorrhoea Angioedema

58 41 37 35 31 8

* From: Truswell AS 1986 ABC of nutrition

Eczema Exacerbations of eczema are frequently associated with reactions against eggs and milk. Infantile eczema is sometimes improved by withdrawing eggs and cows' milk products; families with a strong history of eczema are advised to breastfeed as long as possible and to delay the introduction of eggs and cows' milk. Dietary manipulation in older children and adults is also worth trying but is often disappointing, and must be supervised to ensure nutritional adequacy.

Gastrointestinal symptoms The best-defined gastrointestinal hypersensitivity syndrome is to gluten in coeliac disease (see Ch. 15). Onset is sometimes precipitated by a gut infection, especially in children. Transient gluten intolerance following mucosal damage from giardiasis or rotavirus infections may occur in other patients. Secondary lactose intolerance may occur

in untreated coeliac disease through mucosal atrophy and loss of lactase. A more frequent problem is cows' milk protein intolerance. Here, mucosal damage permits access to mucosal lymphocytes, which become sensitized to the dietary protein. Subsequent exposure then produces a severe response in which lymphokines are released, damaging enterocytes. Villous atrophy occurs and malabsorption develops. There is also some evidence of exacerbation of the symptoms of ulcerative colitis by cows' milk; it is sometimes suggested (though reliable evidence is lacking) that infantile colic is caused by dietary factors transmitted in breast milk or contained in cows' milk. Early exposure to cows' milk in infancy has been related to gastrointestinal blood loss and anaemia, but the mechanism is obscure. The principal carbohydrate in milk is lactose. The expression of lactase in intestinal mucosa varies widely in adults, and in some races (particularly eastern) it is virtally absent, leading to lactose intolerance. Fermented milk products are preferred and are less liable to be malabsorbed. The very rare genetic fructose intolerance syndrome is usually identified by patients who develop symptoms after sweet foods.

Management of suspected food sensitivity If the patient suffers a reaction each time exposure to the allergen occurs, the latter should be easily identifiable. However, in the case of the general systemic symptoms described above, identification of the allergen is more difficult. There are two approaches:

5

• Exclude a suspected allergen for at least 4 weeks and then reintroduce it (this is mostly satisfactory where the condition is mild); or • Work out an elimination diet (one based on lamb, rice and pears is a classic starting point) as advised by a dietitian which ensures full nutritional adequacy. Then reintroduce a potential allergen every 2 weeks and observe the response. It is important to monitor body weight regularly and be alert to the possibility of an underlying eating disorder. Multiple food 'allergies' are less likely to have organic origins, and multiple exclusion diets can be dangerous. Restricting food choices for growing children is particularly hazardous, and needs careful supervision and counselling.

Migraine Migraines sometimes appear to be precipitated by foods such as chocolate, cheese, red wine, nitrates, cows' milk and wheat. Some of these contain amines or stimulate histamine release, which may produce an effect on cranial arteries.

NUTRITION IN HEALTH PROMOTION: CURRENT DIET AND PROPOSALS FOR CHANGE Health promotion

Respiratory tract responses Respiratory tract responses - including excessive production of tears, nasal secretions, sneezing and asthma are usually related to environmental factors, e.g. chemical sprays, cosmetics and food or plant pollens, but eggs, cows' milk and chocolate may also precipitate symptoms. Recent studies suggest that a variety of food preservatives, including tartrazine and sulphur dioxide, may also be responsible.

Urticaria and angioedema Most cases of urticaria or angio-oedema do not have dietary orgins. Acute or chronic urticaria may occur in response to eggs, nuts, fish (particularly shellfish) and cows' milk, as well as certain fruits, wine and cheese. Angioedema is a feared and potentially lethal complication.

Neurological and behavioural symptoms A range of syndromes of 'hyperactivity' in children and tiredness in adults are commonly attributed to intolerance, especially to tartrazine. Although some cases may occur, social factors are probably more important.

The promotion of good health (physical, psychological and social) incorporates an increasing emphasis on disease prevention through dietary change. Many of the major disabling illnesses and causes of premature mortality in western societies - heart disease, stroke, cancer - are progressive, not amenable to cure by drug therapy, and to a substantial extent preventable through dietary change. Two approaches are possible: • High-risk screening strategy. This approach requires a screening procedure to identify individuals at high risk, who then receive specific dietary advice and professional follow-up. The screening procedure must have very high sensitivity and must be able to reach high-risk groups, such as those in deprived areas and communities. It is dependent on high levels of professional input and therefore expensive. Such a programme must be sustained indefinitely, as there will always be new individuals entering the high-risk categories. Although this type of approach has been heavily promoted for cholesterol lowering, the principles outlined above cannot in fact be met. In particular, the sheer numbers who would require to be screened and followed up would make the system impractical (almost half the UK population has undesirably high blood lipids), and

141

the low sensitivity and specificity of cholesterol screening means that among those with low serum cholesterol the commonest cause of death is still ischaemic heart disease. The high-risk strategy approach is therefore more suited to less frequent conditions that require specific advice, such as phenylketonuria identified by a Guthrie test at birth, or the identification of women who have had a child with neural tube defect, who need high-risk folate supplementation (4mg daily) in the periconceptional period to prevent a recurrence. • Population-directed strategy. This approach does not involve screening but requires, by a combination of health education and modification of food supply, a change in the nutrient consumption of the whole population or of targeted subgroups. Examples include the abolition of marketing of high-sugar bedtime drinks for infants, as a way of reducing dental caries, and the healthy eating food and health policies designed to reduce major diseases such as ischaemic heart disease, diabetes, cancers and osteoporosis, which have been urged by many governments and umbrella organizations such as WHO. Population-directed programmes do not demand major health service inputs, although high-risk strategies may operate alongside for special groups. Because the diet of the whole population will be modified, it is essential that the changes promoted should be advantageous or safe for everybody. Given the very limited resources of health promotion agencies, effective population-directed nutritional policies must include cooperation from the food production and supply industries.

142

The average dietary intakes in western countries, including the UK, have changed substantially over the last 50 years, with greatly increased intakes of foods of animal origin, and declining intakes of fruits and vegetables and starch-rich staple foods (cereals, potatoes). There is now little scientific argument in opposition to health-directed measures aimed at reducing intakes of total fat (particularly saturated fat), reducing sodium and some increase in potassium and calcium consumption, while promoting increased consumption of fruits and vegetables and, to some extent, unsaturated fats, particularly those from oilrich fish. Although the benefits to any individual from dietary change are relatively small, the burden of disease to whole populations would be reduced substantially, particularly through reductions in heart disease, stroke and diabetes, and in the prevention of obesity and osteoporosis. A comprehensive package of quantitative proposals for dietary change have been made for Scotland, a country which has a very high burden of preventable disease (Table 5.29). To some extent .this has been influenced by the observational evidence from countries such as Norway and Finland, where long-term programmes of dietary change have resulted in major reductions in heart disease, stroke and cancer. It is important to recognize that nutrient tar-

TABLE 5.29 Intermediate nutrient and dietary targets for Scotland for the year 2005 Dietary component Vegetables and fruit excluding potatoes (g/d) Carbohydrates: total % E Starch (% E) Non-starch polysaccharides (g/d) Extrinsic non-milk sugars (% E) Fatty acids (% E): Cis-monounsaturated Cis-polyunsaturated (Total) n-6 n-3 Saturated Trans Total fat (% E) Sodium consumption (mmol/d) Potassium consumption (mmol/d) Alcohol consumption* (g/d) Men Women Dietary cholesterol (mg/d)

1989-91 intake

Targets

181

>400

25.3

>40

10.5 16.3

>16 <10

12.6 6.3 5.2

<10

0.8 16.6

13 <8.8 >1.2

2.1

<11 <2

40.7

<35

163

70

62

80

34.5 11.1*

<24*

-

-

<16*

f

Current intakes relate to estimates derived from a variety of sources. % E signifies the proportion of the nutrient as a percentage of total food energy, i.e. excluding alcohol. Changes in vitamin intakes are those expected from the dietary changes; women of child-bearing age have been recommended to supplement their diet with 400 ug folic acid, * Alcohol is specified for consumers only, intakes from Adult Survey, maximum intakes proposed by Royal College of Physicians (1987).

gets can only be met if there are changes in food choices, and that these must be matched by changes in food supply. Successful intervention programmes thus require multidepartment government action over a long period. For some of the nutritional consequences of dietary change, the benefits would be seen principally in those who are currently consuming very poor diets; however, it is likely that there would be some improvement even for groups with better diets at present. Extensive studies have shown no appreciable health disadvantage in the changes proposed. In particular, the effects of cholesterol lowering, which were disputed for many years, have now been shown clearly not to be harmful. The benefits from sodium restriction had been difficult to quantify because of the variety of research addressing different groups with different levels of blood pressure and different levels of sodium intake. Large meta-analyses have clarified that a major reduction in sodium intake should be made. This would partially benefit those who already have relatively high blood pressure and would be likely to reduce the development of hypertension in younger people.

RECENT ADVANCES IN HUMAN NUTRITION Recent years have seen an important expansion of research in human nutrition, stimulated by recognition that the major sources of illness and premature mortality in the world, and specifically in highly developed Western countries, will not be amenable to cure by drug therapy. If these problems (cancers, heart disease, stroke, diabetes, osteoporosis) are to be alleviated, the answer must lie in prevention, and diet plays a major role. These are the themes that underlie the application of modern nutritional science to medicine. Most diseases develop as a result of interaction between genetic factors peculiar to the individual, and environmental exposure or conditions. Thus within human nutrition the concept of 'gene-nutrient interactions' has emerged. It is very difficult for people to change their diets, and satisfy targets to maintain good health. The science of food choice, which relates closely to food science in the context of functional foods, designer foods and therapeutic foods or diets, has become of major importance. Dietary and nutritional influences operate over a very long timespan in modifying health and disease. Longterm observational follow-up studies from cohorts of relatively young people, are already demonstrating new dietary influences on health in periodic followup surveys. A number of recent advances in human nutrition can be identified as arising from these new understandings: • Importance of fetal nutrition and development for later health: poor fetal growth predicts diabetes, hypertension, heart disease, stroke. • Folic acid supplementation in the perioconceptional period overcomes metabolic abnormalities that prevent neural tube development. • Dietary folate reduces serum homocysteine, a risk factor for CHD. • Antioxidants - vitamins C, E, carotenoids and phenolics - have the potential to be involved at several stages against development of cancers (free radical DNA damage) and coronary heart disease (protection against LDL oxidation). • Vitamin supplements given unphysiologically as tablets may have different effects from giving foods which contain these vitamins. Foods contain many other related bioactive compounds such as flavonoids, carotenoids, etc. This balance is likely to be important for health. • Specific fatty acids have different effects on plasma lipo-proteins and coronary heart disease risk. For example stearic acid, although saturated, does not elevate cholesterol. Trans-fatty acids, found mainly in synthetically hydrogenated fats such as margarine and in baked foods, elevate LDL cholesterol and are linked to heart disease prevention in the same way

as saturated fatty acids. Individuals with different apolipoprotein E subtypes respond differently to dietary fatty acids. • The human placenta can probably transfer fatty acids to the fetus and there may be specific fatty acid transfer proteins. • Immune function appears to be determined to some extent by the balance of fatty acids in the diet. New non-absorbable drugs (e.g. orlistat) or novel food products (e.g. sitosterols and sterol esters) modify fat absorption and consequent health. • The two-way interaction between nutritional status and disease may lead to 'conditioned essentiality' for specific nutrients during illness, e.g. for amino acids such as glycine or glutamine, which are not essential during health.

Some of the dietary changes required to improve health demand action principally from individuals, to select more appropriate foods and to cook and prepare them in more healthy ways. Thus an increased consumer demand for fruits and vegetables and for lower saturated fat food can be encouraged to help health promotion, and is likely to be met by increased provision from the food industry. To reach other nutrient targets, however, there will need to be greater emphasis on programmes of change within the food manufacturing and catering industries: for example, about 80% of sodium intake derives from foods that have had salt added during manufacture or processing. A programme of stepped small reductions in the salt content of common foods is likely to be acceptable and beneficial to consumers. The composition of many foods could be modified with health benefits and without impairing palatability or commercial success. Examples of effective national nutritional policies are available from countries such as Norway and Finland, where substantial changes in average diet composition over a period of 10-20 years have been accompanied by major reductions of the order of 60-70% in ischaemic heart disease and cancers before the age of 65. There are far greater effects that could be achieved by lipid lowering or other drug interventions. The three dietary components with the greatest impact are reduced saturated fat as a proportion of energy, reduced sodium consumption, and substantially increased fruit and vegetable consumption. The success of these programmes has depended on the activities of committed health professionals prepared to take on public positions and to interact with national and local governments and with the food supply industries. Responsible and well-educated media involvement, which does not seek to promote mavericks and sensationalists, has been influential. High levels of general education and minimization of social deprivation are critical for the effectiveness of health promotion.

a

143

FURTHER READING Department of Health. Nutrition: A core curriculum for nutrition in the education of health professionals. London: HMSO, 1994. Department of Health. Nutritional aspects of the development of cancer. Report of the Working Group on Diet and Cancer of the Committee on Medical Aspects of Food and Nutrition Policy. Report on Health and Social Subjects No. 48. London: HMSO, 1998. Dietary reference values for food energy and nutrients for the United Kingdom. Report on Health and Social Subjects, 41 1991. Department of Health COMA. London: HMSO, 1991. Garrow J S Obesity and related disorders. Edinburgh: Churchill Livingstone, 1988. Holland B, Welch A. McCance and Widdowson's compositions of foods, 5th edn. Cambridge: Royal Society of Chemistry. 1991: See also supplements dated after 1991. James W P T, Garrow J S Human nutrition and dietetics, 10th edn. Edinburgh: Churchill Livingstone, 2000.

144

Mann J, Truswell A S. Essentials of human nutrition. Oxford: Oxford University Press, 1998. Manual of dietetic practice, 2nd edn, Oxford: Blackwell Science, 1994. National Institutes of Health (NIH) obesity guidelines. Available online at http://www.nhlbi.nih.gov/guidelines/obesity/ob_home.htm The Scottish Office Home & Health Department. The Scottish diet: Scotland's health - a challenge to us all. Report of a Working Party to the Chief Medical Officer for Scotland, 1994. Scottish Office Department of Health. Scotland's Health - A challenge to us all. Eating for health - a diet action plan for Scotland, 1996. Truswell A S ABC of nutrition, 2nd edn. London: British Medical Association, 1992. WHO Diet, nutrition and the prevention of chronic diseases. Technical Report Series 797. Geneva: World Health Organization, 1990.

6

Cancer Medicine Robert L Souhami

Incidence and epidemiology 145 Aetiology 146 Tumour growth and development 149 Pathology 151 Clinical features 151 Staging and investigation 155

Supportive care in cancer management 165 New approaches to management 167 Trials of treatment 167 Care of the dying 168 Alternative therapy 168

Principles of cancer management 157

Each year in the UK there are approximately 220000 new cases of cancer. Almost one-third of the population will develop cancer during their lifetime, and of these 70% will die of the disease - 150000 deaths a year. As a cause of mortality, cancer is second only to cardiovascular disease. It is most common in the elderly: the age-specific incidence rates for men and women are shown in Figure 6.1. The frequency of cancer and the increasing complexity of management mean that every clinician must have a firm grasp of the principles of diagnosis, investigation and management. In recent years, the approach has been changing rapidly as a consequence of fundamental advances in cancer biology, the isolation of genes responsible for hereditary disposition to cancer, increasing accuracy in diagnosis and staging, and major developments in the use of radiotherapy and chemotherapy. At the same time, clinicians have developed a better awareness of the problems faced by patients who are incurable, and professional attitudes have changed in the direction of more open discussion with patients. Cancer medicine therefore demands a combination of a high degree of technical competence and a sympathetic insight into the feelings of the patients. It is this double challenge that makes it a most rewarding part of medical practice.

INCIDENCE AND EPIDEMIOLOGY Incidence means the proportion of a defined population developing a disease in a stated period of time; prevalence

is the proportion which has the condition at a single point in time. Crude incidence or prevalence rates refer to a whole population and specific rates to selected groups, for example those defined by age or sex. The age-specific incidence rates of some cancers are shown in Figures 6.2 to 6.4. It can be seen that adenocarcinomas in women have quite different age-specific incidence rates. Cancer of the breast increases rapidly in incidence before the menopause, and at a slightly slower rate thereafter (Fig. 6.2). Uterine and ovarian cancer also show an increasing incidence premenopausally, but the rate does not rise in later life. Gut cancers (Fig. 6.3), on the other hand, show no relation to the menopause but rise rapidly in incidence in old age in both sexes. The biphasic age-specific incidence of Hodgkin's disease (Fig. 6.4) contrasts markedly with the findings in adenocarcinoma. Lung cancer is at present more common in men, but the incidence in women is rising rapidly, approaching that of breast cancer. Cancer is the second main cause of death in childhood, with acute lymphoblastic leukaemia accounting for nearly half the deaths. The commonest sites of cancer in the west are the lung, skin, large bowel, prostate, stomach and rectum in men; and the breast, large bowel, skin and lung in women. The mortality rates are, however, different from incidence rates (Fig. 6.5), with skin cancer having a low mortality rate and lung cancer a very bad prognosis. However, the relative incidence of cancers is changing with time. For example, the incidence of lung cancer has been rising rapidly since the mid-1960s, especially in women, whereas that of stomach cancer has been falling for both men and women (Fig. 6.6). The rise in incidence of lung cancer is undoubtedly due to smoking, but the reason for the decline in stomach cancer is not clear. Cancer incidence varies in different countries (Table 6.1), and this variation may provide insights into aetiology. Carcinoma of the oesophagus has a high incidence in an area extending from the Caspian Sea to Central Asia and the Far East; it is also more common in the Transkei. In these areas it has an incidence 200 times that found in the UK. In Japan, gastric carcinoma is 30 times more common than it is in the UK, but pancreatic cancer is much rarer. The reasons for these variations are as yet unknown, but dietary and nutritional factors are suspected to play a part. The migration of populations has helped to determine how much of the geographical variation is genetic and how much is environmental. When Japanese people settle in the USA, the incidence of cancers in their offspring resembles that of the adopted country: the high incidence of gastric cancer falls, and the lower incidence of cancers of the breast, colon, ovary and prostate rises. These findings suggest that environmental, rather than genetic, causes account for most of the observed differences between countries.

6

145

FIG. 6.3 Age-specific incidence of adenocarcinoma of the colon FIG. 6.1 Age-specific incidence of cancer (England)

FIG. 6.4 Age-specific incidence of Hodgkin's disease

FIG. 6.2 Age-specific incidence of carcinomas of the breast, ovary and uterus Note change in rate of increase at the time of the menopause.

AETIOLOGY Evidence suggesting aetiological factors in cancer has come from epidemiological and laboratory investigation. 1 It is probable that, for many types of neoplasm, an inter-

1

146

MCQ 6.1

FIG. 6.5 Incidence and mortality of common cancers in men and women in England and Wales Note that whereas skin cancer is common, its mortality is low. Cancer of the lung is common and accounts for a large number of total cancer deaths.

6

FIG. 6.6 Cancer of the lung and stomach A Cancer of the lung: age-standardized death rate/10000 in age group 65-74,1905-85 in the UK. B Cancer of the lung: mortality 1941-91 for different age groups. C Cancer of the stomach: change in mortality in England and Wales and in Japan, 1940-1990.

TABLE 6.1 Geographical Incidence of cancer Cancer

High incidence

Low incidence

Ratio (high Mow)

Oesophagus

Iran, Transkei, Kazakhstan Australia Mozambique Far East (China) UK Japan Europe, USA

Holland

200:1

India UK Europe Nigeria UK Japan

200:1 100:1

Skin Liver Nasopharynx Lung Stomach Breast

100:1 35:1 30:1 6:1

action of genetic and environmental factors is responsible for the development of a cancer (Fig. 6.7).

CHEMICAL CARCINOGENS

The frequency of scrotal carcinoma in chimney sweeps was noted by Sir Percivall Pott in 1775, and the high incidence of bladder cancer in dye workers in 1895. The mechanisms

of action of chemical carcinogens have been intensively studied. There are several classes of experimental carcinogens (Table 6.2) which are possible causes in humans: • Polycyclic hydrocarbons are formed from combustion of organic compounds. They are present in cigarette smoke, soot, car exhaust fumes and some foods. Dibenzanthracene was one of the first to be shown to be carcinogenic. The carcinogenic effect results from reaction with bases in DNA. • Aromatic amines and azo dyes, such as B-naphthylamine and benzidine, are metabolized to an active form. (3Naphthylamine is glucuronated in the liver. Glucuronidases in the urine liberate the carcinogen in the bladder. • Nitrosamines are formed when nitrites in food are converted to nitrates, which react with amines in food. The substances can cause stomach cancer experimentally. • Alkylating agents, such as those used in cancer management, bind to DNA. Use of these drugs in cancer treatment has been associated with the development of secondary malignancies, the most important of which is secondary leukaemia, which may also be caused by treatment with etoposide, a widely used anticancer agent.

147

FIG. 6.7 The sequence of genetic events contributing to invasive cancer

TABLE 6.2 Chemical carcinogens Class

Example

Source

Polycyclic hydrocarbons Aromatic amines Azo dyes Nitrosamines Toxins Alkylating agents Chromates Asbestos

Dibenzanthracene

Smoke, cigarettes

p-Naphthylamine Dimethylaminobenzene Dimethylnitrosamine Aflatoxin Melphalan

Aniline dyes Butter yellow ?Food Aspergillus flaws Drug treatment of cancer Industrial exposure Insulation and heating installations

• Organic toxins include the best-studied example, aflatoxin, produced by the mould Aspergillus flavus growing on grain. This substance has been suggested as an aetiological agent in liver cancer in Africa. • The most important occupational carcinogen is asbestos, which causes cancer of the pleura, lung and peritoneum in workers exposed to the dust. When carcinogens act on a cell they may cause a permanent change which predisposes to cancer, a process known as initiation. Other agents can then produce transient changes which with continued exposure lead to cancer; this is known as tumour promotion. This multistage process has been shown in some experimental cancers, where mutagens act as initiators and hormones or dietary constituents act as promoters. The more prolonged the exposure to a carcinogen, the more likely the tumour is to develop. Several genetic events in sequence are necessary for the tumour to develop.

IRRADIATION

148

The carcinogenic effect of ionizing irradiation was indicated by the skin cancers that developed on the hands

of early radiologists, and by the increased frequency of acute and chronic leukaemia, breast and thyroid cancer in Hiroshima during the 30 years after the atom bomb was dropped. The mechanism of carcinogenesis is probably related to the chromosomal damage caused by X-rays. It seems likely that 'background' ionizing irradiation does not contribute greatly to common cancers in humans. Therapeutic irradiation may give rise to bone and soft tissue sarcoma (after a latent period of 10 years), and possibly to lung cancer. Ultraviolet light probably plays a role in the development of skin cancers. This is seen most clearly in the autosomal recessive disorder xeroderma pigmentosum, where a defective DNA repair mechanism is associated with the early development of skin cancers on skin exposed to sunlight. Malignant melanoma and other skin cancers are uncommon in black people but are particularly frequent in fair-skinned Celts. In Australia, skin cancers are common among the non-indigenous population and increase in incidence in the north nearer the Equator.

VIRUSES Ribonucleic acid (RNA) viruses can cause a variety of cancers in animals. These viruses contain a gene which codes for the enzyme reverse transcriptase, which is produced in the infected cell and results in the cellular synthesis of DNA complementary to the viral RNA and thus to the production of more viruses. For this reason they are called retroviruses. Tumours caused by these viruses include avian erythroblastosis, rat sarcoma, feline leukaemia and murine leukaemia. The insertion of the viral genome into DNA leads to malignant transformation. A growth-factor receptor is encoded by the erythroblastosisB virus (v-erb B), and a phosphokinase by Rous' sarcoma virus (sis). RNA viruses are involved in the development of some forms of T-cell leukaemia and lymphoma (human T-cell leukaemia virus, HTLV types 1 and 2). The human immunodeficiency virus (HIV) is the cause of the acquired

immune deficiency syndrome (AIDS), in which there is a greatly increased frequency of skin sarcoma and B-cell lymphoma. These tumours are probably not caused by the virus but are a consequence of oncogenesis in patients with greatly reduced immunity. Hepatitis C is an RNA virus which is implicated in the high incidence of hepatoma in chronically infected patients. Research into RNA viruses led to the discovery that the viral oncogenes had homologues in normal cellular DNA. These cellular proto-oncogenes appear to code for products very similar to those produced by the equivalent virus. These findings have led to the search for increased expression of cellular oncogenes in human cancers, for the gene products, and for the link between carcinogenesis and oncogene activation. There is strong circumstantial evidence implicating DNA viruses in the pathogenesis of some malignant neoplasms. For example, Epstein-Barr virus (EBV) causes proliferation of normal human B lymphocytes; EBV-coded proteins are found in Burkitt's lymphoma cell lines; and African children with the lymphoma have high titres of antibody to viral antigens. Worldwide, hepatoma is one of the commonest cancers. Infection with hepatitis B virus (HBV) is associated with its development both in Africa and in Europe. Hepatoma is particularly likely to occur in chronic active hepatitis associated with HBV and hepatitis C, and in alcoholic cirrhosis where there is evidence of previous HBV infection. Cancer of the uterine cervix is increasing in frequency in young women. There is strong evidence that human papillomavirus is associated with this disease, possibly as a result of sexually transmitted infection.

GENETIC FACTORS Numerous inherited genetic abnormalities predispose to the development of malignancy (Table 6.3). These syndromes include examples of autosomal dominant and recessive inheritance, and X-linked disorders. Inherited chromosomal abnormalities may also predispose to cancer (Table 6.3). The most common is Down's syndrome (trisomy 21), where there is an increased incidence of acute leukaemia. In some cases the relationship between the genetic abnormality and the induction of cancer has become clearer. In hereditary retinoblastoma the retinoblastoma gene (Rb) acts as a growth suppressor gene. Inherited loss of this gene is followed by mutation in the other allele, which sets the scene for malignant transformation. Germline mutation, followed by loss of heterozygosity of the p53 gene, occurs in the Li-Fraumeni syndrome, in which the families of children with sarcoma show an excess of other cancers. It has become apparent that there may be an important inherited genetic component to many other common cancers. Very recently, two genes responsible for familial breast cancer have been isolated. The first, BRCA1, is prevalent in Ashkenazi

RECENT ADVANCES IN GENETIC DIAGNOSIS OF FAMILIAL CANCER, SCREENING, CANCER PREVENTION

6

The cloning of two breast cancer genes that are responsible for early-onset and aggressive breast cancer, the identification of the genes responsible for hereditary colon cancer, and the likelihood that other genes are responsible for some cases of testicular, lung and prostate cancer, will lead to increasing demand for genetic testing for cancer risk. In satisfying this demand it will be necessary to provide counselling and advice to patients and their families. The identification of individuals at especially high risk may allow screening services to be directed more efficiently towards those who have extra need and in whom considerable expenditure will be fully justified. It is now feasible to consider the use of drugs which may prevent cancer developing in those at high risk. An example of such intervention is the use of tamoxifen as a preventative agent in breast cancer.

Jewish women and has an early age of onset. A wide variety of mutations have been detected. BRCA2 predisposes to both breast and ovarian cancer. In hereditary nonpolyposis colorectal cancer the syndrome arises as a result of mutation in one of a family of DNA mismatch repair genes. Acquired genetic abnormalities are frequently found in cancer cells, with mutation, transposition and deletion of genetic material. These abnormalities are sometimes consistent, such as the translocation from chromosome 22 in chronic granulocytic leukaemia, and the reciprocal translocation involving chromosome 8 in Burkitt's lymphoma. In many tumours (lung cancer, leukaemia, hepatoma) there is mutation of the p53 gene, which codes for a growth regulatory transcription factor. In colon cancer a cascade of genetic events is responsible for the transformation of polyps to invasive cancer.

TUMOUR GROWTH AND DEVELOPMENT Some tumours exhibit properties which imply that they have arisen as a result of clonal expansion caused by an oncogenic event in a single cell. Lymphomas, for example, usually show restriction to one type of surface immunoglobulin light chain; and studies in individuals who are heterozygotes for the enzyme glucose-6-phosphate dehydrogenase have shown that tumours contain only one form of the enzyme. Whether all cancer are monoclonal in origin is not known, and the development of some

149

carcinomas in a polyclonal field of premalignant change remains a possibility. Even if most tumours are monoclonal, heterogeneity of structure and function arises during the course of tumour growth. Cells from a tumour are diverse with respect to resistance to damage by irradiation or cytotoxic drugs, surface antigen expression, growth rate, and expression of biochemical markers such as embryonic proteins or peptide hormones. This diversity is due to genetic instability in the cancer cells. Heterogeneity within tumours presents a formidable problem for treatment. In normal tissues, self-renewing stem cells replenish a proliferating population of cells which follow a pathway of differentiation resulting in a mature tissue (Fig. 6.8). Cell death is balanced by the formation of precursors, and the tissue is static in size or else grows and diminishes in response to such external controls as hormonal environment. In cancers the restraint on growth is disrupted, so that sustained expansion of the tumour occurs. For clinically detectable cancers, the kinetics of growth appear to be exponential and the rate of tumour volume doubling varies widely. In Burkitt's lymphoma, the doubling time of visible tumour may be as short as 1-2 days, in contrast to 100-400 days for some squamous carcinomas and adenocarcinomas. It seems unlikely that growth is exponential from the beginning of the cancer, as this would imply very long subclinical growth periods for some tumours (Fig. 6.9). Cancers are clinically apparent after about 32 doublings, and lethal after 36-38. It is possible that tumours grow more rapidly when they are very small and slow progressively as they enlarge (gompertzian growth). If true, this may have implications for therapy, as the more rapidly dividing small tumour might be more susceptible to chemotherapy. During growth, the tumour becomes vascularized and lymphatic drainage is established. Cells more than 150 um from a capillary are relatively hypoxic and resistant to radiotherapy. They are also exposed to a lower concentration of cytotoxic agents. The vascularization may be insufficient to sustain viability, and areas of necrosis and haemorrhage are frequently found in large cancers. The process of vascularization allows tumour cells to penetrate blood and lymphatic channels, and thence to metastasize to regional lymph nodes and distant sites.

Metastasis Metastasis is a remarkable process and one which is still very poorly understood. The risk of metastases increases as tumours become larger. The cells must survive tissue invasion, circulation, passage across the capillary wall, and establishment in tissues. The process of tissue penetration

1

150

Figs 6.1, 6.2

Table 6.3 Inherited genetic abnormalities and malignancy Inherited disorder Skin disorders Neurofibromatosis (AD) Tylosis palmaris (AD) (keratosis of palms and soles) Peutz-Jeghers syndrome (AD) Immune deficiency X-linked agammaglobulinaemia (XLR) Wiskott-Aldrich syndrome (XLR) Ataxia telangiectasia (AR) X-linked lymphoproliferative syndrome (XLR) Gut disorders Polyposis coli (AD) Gardner's syndrome (AD) Hereditary non-polyposis Colotectal cancer (AD) Neurological disorders Tuberous sclerosis (AD) Retinal/cerebellar angiomatosis (von Hippel-Lindau syndrome) (AD) Other syndromes Bloom's syndrome (AR) (short stature, telangiectasia) Hemihypertrophy Fanconi's anaemia (AR) (skeletal abnormality, patchy pigmentation, mental retardation) Multiple enchondromata (S) (Oilier syndrome) Chromosomal abnormalities Down's syndrome (trisomy 21) Klinefelter's syndrome (47.XXY) Aniridia - Wilms' syndrome (11pMosaicism (45,XO/46,XY) 13q- (multiple malformations)

Cancer

Sarcoma, phaeochromocytoma, medullary carcinoma of thyroid Oesophageal cancer Intestinal and ovarian cancer

Lymphoma Lymphoma Lymphoma and stomach cancer Lymphoma

Colonic carcinoma Colonic carcinoma Duodenal carcinoma Colonic carcinoma

Glioma Hypernephroma, ependymoma, phaeochromocytoma

Leukaemia Many other cancers Wilms' tumour, hepatoblastoma Acute leukaemia

Chondrosarcoma

Acute leukaemia Breast cancer Wilms' tumour Gonadoblastoma Retinoblastoma

Hereditable component of other cancers Breast cancer BRCA1 BRCA2 Breast cancer Unknown (recessive?) gene Testicular cancer p53 mutation Sarcoma in children Breast and other cancers AD Autosomal dominant AR Autosomal recessive

XLR X-linked recessive S Sporadic

• sarcomas, which arise from mesodermal cells making up bone and connective tissue; • leukaemias and lymphomas, which arise in cells of the bone marrow and immune system.

FIG. 6.9 Diagrammatic representation of exponential and gompertzian growth

FIG. 6.8 Diagrammatic representation of cellular proliferation in a tumour A Pluripotent self-renewing stem cell. B Proliferative compartment in which cell number increases but gradually loses reproductive potential. C Mature tissue.

appears to be by secretion of enzymes known as metalloproteinases (such as collagenase). Inhibitors of tissue metalloproteinases are now in early trials as anticancer agents. The precise location of a metastasis is probably due in part to chance. Nevertheless, clinical patterns of bloodborne metastasis are apparent: gut cancers spread through the portal venous system to the liver; ovarian cancers seed into the peritoneal space; breast cancer has a predilection for bones of the axial skeleton; sarcomas characteristically spread to the lungs.

PATHOLOGY The major categories of malignant tumour are: • carcinomas, which arise from ectoderm or endodermal cells;

6

The distinction between benign and malignant neoplasms is based histologically on the cellular pleomorphism, frequent mitoses and tissue invasion that are characteristic of malignancy. The clinical manifestations of malignancy are tissue infiltration and lymphatic and bloodborne metastases. The degree of differentiation of a tumour is often an important guide to prognosis, and therefore to therapy. Poorly differentiated carcinomas have a worse prognosis than well-differentiated tumours. This is due in part to an association with a more rapid growth rate and more early and widespread metastases. About 5% of cancers present as metastases where the primary site of the tumour is unknown. The histology is usually a poorly differentiated carcinoma, often with features suggestive of adenocaicinoma, but the pathologist may find the tumour difficult to classify. The development of monoclonal antibodies has led to a greater precision in diagnosis by using immunocytochemical methods. Many anaplastic tumours can now be diagnosed as epithelial tumours, melanomas or lymphomas by using antibodies against epithelial antigens, or antigens associated with melanoma, leukocytes or T and B lymphocytes. The correct identification of lymphoma will lead to important changes in management of a patient (Fig. 6.10). Many antibodies will not bind to antigens on formalin-fixed and paraffin-embedded sections, and require fresh tissue. Other specialized investigations can be carried out on fresh tumour tissue - measurement of hormone receptor content (e.g. oestrogen and progesterone receptors) for example, which may be of importance in prognosis and in deciding treatment. Using immunocytochemical techniques, the presence of substances such as oc-fetoprotein can be detected in the tumour. The increasing complexity and versatility of histopathological diagnosis means that it is now essential for there to be close collaboration between clinician and pathologist before specimens are obtained. Ultrasound and CT scanning now allow accurate needle biopsy and aspiration cytology of hitherto inaccessible intra-abdominal and intrathoracic tumours. 1 Cytological diagnosis does have some limitations: in the diagnosis of lymphoma, for example, it may be difficult to distinguish tumour cells from normal reactive lymphocytes. For this reason, histological diagnosis is preferred, especially in situations in which there is clinical doubt about the diagnosis.

CLINICAL FEATURES Cancers cause symptoms due to the local spread of the primary tumour, and as a result of metastasis and of metabolic and remote effects of the malignancy.

151

Metastatic disease in bone causes pain which is often intense and distressing. One characteristic feature of malignant bone pain is that it is often worse at night. Bone pain may occur in the absence of a pathological fracture or vertebral collapse, but becomes worse when this occurs. In the liver, metastases cause pain by stretching the capsule. Cerebral metastases give rise to raised intracranial pressure, fits, focal signs and symptoms, and mental confusion. Pulmonary spread may cause breathlessness due to large intrapulmonary tumour masses, diffuse lymphatic permeation (lymphangitis carcinomatosa) and massive pleural effusion.

SYSTEMIC SYMPTOMS Primary and metastatic cancer may cause a variety of metabolic and other disturbances which are important causes of symptoms. In general, these are produced in one of three ways:

FIG. 6.10 Section of nasopharyngeal tumour A Staining with anticytokeratin; the epithelium of normal respiratory gland stains strongly but tumour cells are negative. B Staining with antibody to the leukocyte common antigen reveals that the tumour cells are expressing the leukocyte antigen and that the tumour is a lymphoma.

SYMPTOMS DUE TO LOCAL DISEASE

152

The primary tumour is clinically silent either until it becomes a visible painless lump that the patient notices (this is a common presentation of breast cancer and Hodgkin's disease and other lymphomas), or until it impinges on surrounding tissues. Thus, carcinoma of the lung will cause increasing bronchial narrowing, with breathlessness, cough and collapse of the distal lung; colonic carcinoma may present with disturbed bowel habit and obstruction; brain tumours may cause raised intracranial pressure or focal neurological signs; cancer of the head of the pancreas will cause obstruction of the bile duct, leading to jaundice. As the primary tumour gets larger it may ulcerate and bleed. Haemoptysis is a common symptom of bronchial carcinoma, and chronic blood loss leading to iron deficiency anaemia frequently occurs with carcinoma of the colon. The tumour may infiltrate nerves and bone, leading to pain, or pain may arise from stretching of the capsule of an organ - for example pain over the liver with advanced hepatoma. Nerve entrapment frequently occurs with pelvic tumours, such as carcinoma of the cervix. Similarly, carcinoma of the bronchus occurring in the superior sulcus (Pancoast's tumour) may be accompanied by intense pain in the arm, owing to involvement of the brachial plexus.

• As a result of the tumour mass, both primary and metastatic; • As a result of secretion of hormones by the tumour (Table 6.4); • By ill-understood mechanisms which give rise to specific cutaneous, neurological and other syndromes remote from the tumour and not due to metastasis - paraneoplastic syndromes (Table 6.5). The commonest metabolic disturbance associated with advanced cancer is cachexia. This is a state of profound weight loss, muscle wasting and weakness. Many factors contribute to the state, of which anorexia is perhaps the major cause. Loss of appetite occurs early with tumours of the upper gastrointestinal tract, even before metastases have occurred. For most tumours loss of appetite is an accompaniment of advanced disease, particularly hepatic metastases. Anorexia may be made worse by treatment with radiation and cytotoxic drugs, by psychological factors such as depression, and by unrelieved pain. The tumour may promote weight loss by direct metabolic effects. For example, there may be excess lactate production by the tumour which is recycled into glucose (a process wasteful of energy), and an increased glucose consumption at a time when carbohydrate intake has declined. Some tumours may cause malabsorption (carcinoma of pancreas) or intestinal obstruction (colon cancer), which contributes to weight loss. Unlike patients who are starving, many cancer patients have a normal or high metabolic rate but the reason for this is not clear. Fever may accompany cancer and is not always due to infection, although this must always be excluded. Fever is particularly likely to occur with lymphomas, renal carcinoma, sarcomas, and any tumours that have metastasized to the liver. Both fever and anorexia may be due to cytokine production by the tumour.

6

CASE STUDY 6.1 COUGH AND BACK PAIN A 65-year-old retired warehouseman presented with a 2-month history of troublesome cough productive of clear sputum. One week previously he had had a sudden onset of midthoracic back pain while bending to clean the bath. The pain was worse on moving and on coughing. He smoked 20 cigarettes a day and for 2 years had had intermittent pain in his left calf on walking 500 yards. On examination he looked well but had back pain on undressing. BP 145/90, pulse 76 regular. Foot pulses absent except for the right posterior tibial. Both external jugular veins were distended. The heart was normal and there was no peripheral oedema. Chest expansion was symmetrically diminished, with increased anteroposterior diameter of the chest. There was an expiratory wheeze, more marked on the right. He was tender on pressure over T8, with a suggestion of angulation at this site. The rest of the examination was normal.

Questions 1. What is the probable underlying diagnosis? 2. What might be the cause of the venous distension? 3. What is the probable cause of the back pain? 4. What are the most important diagnostic investigations? 5. What should be the immediate treatment?

Discussion The patient is a cigarette smoker who has a history and signs of peripheral vascular disease in his legs. Any new

symptoms in such a patient should raise the question of whether these might be due to the development of a bronchial carcinoma. The local symptoms caused by a bronchial carcinoma depend on the site of the tumour. At the periphery of the lung it may remain silent until it invades the pleura, causing pain or an effusion, or obstructs a peripheral airway, causing breathlessness or infection. When centrally placed, tumours may obstruct an airway, causing severe breathlessness, wheeze, or an exacerbation of cough. The tumour, or associated lymph node masses, may expand in the superior mediastinum and cause superior vena cava obstruction (SVCO). Bronchial carcinoma typically metastasizes to liver, bone and brain. About 50% of cases present with symptoms or signs of metastasis. Metastasis to bone typically causes pain that may not be alleviated by rest - indeed, it is often troublesome at night. Presentation with pathological fracture may occur. One of the most important sites of bone metastasis is the cervical and thoracic vertebral column, because at this site there is a grave risk of spinal cord compression. In this case the worsening cough might equally be due to an exacerbation of the obstructive airways disease from which smokers suffer. The back pain after minor exertion raises the possibility of a vertebral collapse and therefore of metastasis to the vertebral column. Physical examination further increases the suspicion that there might be an underlying malignancy. The wheeze is more prominent on the right, suggesting a unilateral cause, and there is tenderness over

Anaemia may be due to blood loss, malabsorption, or dietary deficiency of iron or folate. The most common mechanism is a failure of release of iron (stored in macrophages) into the plasma - the anaemia of 'chronic

the spine of the eighth thoracic vertebra. The distended external jugular veins, in the absence of signs of heart failure, may indicate the early onset of SVCO. Investigations are therefore directed to establishing the diagnosis and the cause of the back pain. A chest X-ray may show appearances suggestive of a carcinoma and indicate if this is centrally placed. If a mass is present the involvement of the SVC can be further investigated by CT scan. Sputum cytology may reveal malignant cells. If negative, a bronchoscopy will be necessary provided that the clinical condition permits. A bone scan will show increased uptake in the thoracic spine if there is a vertebral metastasis, and a lateral X-ray may show vertebral collapse. The appearances will be further delineated by a thoracic CT scan. This will also give an indication of incipient pressure on the spinal cord which is the greatest threat to this patient. Routine full blood counts and biochemistry are unlikely to add to the diagnosis but will be helpful in indicating if other complications of malignancy are present, such as liver metastases or hypercalcaemia. The essential initial treatment is relief of early SVCO and treatment of the vertebral metastasis, if present. To some extent the method of treatment will depend on the histological type of lung cancer. Chemotherapy is rapidly effective in relieving SVCO if the tumour is small cell carcinoma. Other types of lung cancer will require radiotherapy treatment. The vertebral metastasis should be treated with radiotherapy, whatever the histology, especially if the CT appearances suggest that the cord may be in danger.

disease' (p. 1226). This anaemia is unresponsive to iron therapy. Chemotherapy and radiation both depress bone marrow function and contribute to anaemia. The hormonal syndromes that accompany cancer (Table

153

TABLE 6.4 Hormonal syndromes in non-endocrine cancers

TABLE 6.5 Paraneoplastic syndromes

Hormonal syndrome

Tumour

Mechanism

Syndrome

Hypercalcaemia

Squamous carcinomas Carcinoma of kidney

Parathyroid hormone related peptide

Neurological Neuropathy (usually sensory)

Hyponatraemia

Small cell lung cancer Lymphoma Pancreatic carcinoma

Ectopic ADH produced by tumour

Small cell lung cancer Bronchial and thymic carcinoids

Ectopic ACTH produced by tumour

Gynaecomastia

Teratoma Hepatoma Large cell bronchial cancer

Excess oestrogen activity

Hyperpigmentation

Small cell lung cancer

Ectopic MSH produced by tumour

Cushing's syndrome

Cerebellar degeneration Eaton-Lambert syndrome Musculoskeletal Clubbing Hypertrophic pulmonary osteoarthropathy Dermatomyositis

Polyarthritis Haematological Multiple venous thrombosis

6.4) are important because they are a cause of symptoms which are frequently treatable. In many patients hypercalcaemia is due to bone destruction by metastases, but some cancers produce hypercalcaemia by a hormonal mechanism. In these cases the calcium level falls if the tumour is removed. The tumour elaborates a peptide (parathormone-related peptide) and other factors (such as growth factors) which can mobilize bone calcium. Hypercalcaemia associated with cancer can be very severe, requiring urgent management (see Ch. 18). Hyponatraemia, caused by the production of antidiuretic hormone (ADH) by the tumour, is a frequent accompaniment of small cell lung cancer. Malaise, fatigue leading to fits, and coma may result. The condition can be reversed by demeclocycline and water restriction (Ch. 21). Adrenocorticotrophic hormone (ACTH)-like peptides may also be produced by small cell lung cancer. In this disease the high levels of ectopic ACTH produce profound metabolic disturbances, with proximal muscle weakness, hyperglycaemia and hypokalaemic alkalosis. These symptoms can be reversed by management of the tumour. The paraneoplastic syndromes (Table 6.5) can cause diagnostic difficulty in two ways. First, the syndromes must be distinguished from other diseases causing a similar clinical picture. For example, the peripheral neuropathy accompanying cancer may be clinically indistinguishable from other causes. Second, the underlying carcinoma may not always be clinically apparent. Tumours such as carcinoma of the ovary or pancreas may be associated with multiple migratory venous thromboses at a time when the primary disease is not detectable. The diversity of symptoms caused by primary and

0 Case 6.1

154

O Fig. 6.8

0 Figs 6.3-6.6

Q Fig. 6.7

Erythrocytosis

Disseminated intravascular coagulation

Associated tumours

Small cell lung cancer Cancers of breast and ovary Small cell lung cancer Small cell lung cancer

Carcinoma of the lung (particularly squamous cell) Lung cancer Oesophageal cancer Other cancers Adenocarcinomas of breast, gut and ovary

Pancreatic cancer Other adenocarcinomas Renal carcinoma Hepatoma Cerebellar haemangioblastoma Uterine cancer Mucinous adenocarcinomas

metastatic cancer, together with the hormonal, metabolic and paraneoplastic syndromes that accompany many malignant tumours, mean that a diagnosis of cancer must be considered in any unexplained illness, particularly in a middle-aged or elderly patient.

METASTASIS FROM AN UNKNOWN PRIMARY SITE From 2 to 5% of patients with cancer present with a metastasis, in lymph nodes, bone, liver, brain or lung, the primary site not being clinically apparent. Usually the histological appearances are of adenocarcinoma or poorly differentiated cancer. This common clinical problem is a considerable test of medical skill, because the patient is understandable extremely anxious and it is probable that only limited treatment will be available if the primary is discovered. The site may suggest the source of the metastasis. Lymph node metastasis in the head and neck will suggest a head and neck or lung cancer. Metastases in the liver commonly arise from the gut. An axillary node containing adenocarcinoma 1 may be due to a primary in the breast. Detailed histological analysis may be necessary to rule out important and curable cancers such as germ cell tumours or poorly differentiated lymphomas. Investigation is directed towards discovering a treatable primary source, for example in breast or prostate, where hormone therapy

may be effective. Multiple invasive investigations seldom produce information that will affect management. In 20% of cases the primary site is never discovered. Empirical treatment with chemotherapy may be given, but the risk of toxicity is considerable and the response rates are poor.

STAGING AND INVESTIGATION Once a diagnosis of cancer has been established, further investigation is usually needed to establish the degree of spread of the primary tumour and to determine whether metastasis has occurred. It has been known for many years that the prognosis of a patient depends, among other things, on the size of the tumour and the presence of metastases. Our ability to locate the tumour and distant metastases accurately has improved immeasurably in recent years with the introduction of radionuclide and ultrasound scanning, CT scanning and magnetic resonance (MR) imaging. These techniques have shown that in many patients the cancer is metastatic at the time of presentation. This realization has led to a greater emphasis on systemic treatments such as hormone therapy and chemotherapy and, in some cases (e.g. breast cancer), a tendency towards less radical surgical procedures for the primary tumour. The determination of the degree of spread at presentation is generally known as staging. The international acceptance of staging systems has meant that, in treatment trials, comparable groups of patients can be treated allowing a comparison between different methods.

TNM CLASSIFICATION

A generally accepted system which is applicable to some (but not all) cancers is the TNM (Tumour, Nodes, Metastases) classification. 0 The T component defines certain characteristics of the primary tumour, such as its size, site or depth of invasion; the defining criteria will be different with each type of cancer. N refers to the presence or absence of involved nodes; the degree of node involvement may be designated as NO (no involved nodes) and Nl, N2, N3 etc. as more nodes are involved. If no metastases are present the designation is MO, and Ml if they are detected. Clearly, the TNM stage assigned to a tumour will depend on the thoroughness of the clinical and radiological investigation, and this will change as new techniques are invented and perfected. The TNM stages can be grouped into categories which have prognostic significance. These categories or 'stage groups' have been of great value in the design of treatment trials. The TNM stages and related stage groupings for carcinoma of the kidney are shown in Table 6.6 and the prognosis related to the stage grouping in Figure 6.11.

TABLE 6.6 Renal carcinoma TNM staging and stage

grouping Stage

Definition

T1 T2 T3 T4

T stage Small tumour surrounded by normal kidney Large tumour with kidney deformity Perinephric infiltration or renal vein extension Invasion of local structures (e.g. abdominal wall)

NO N1 N2 N3

N stage No nodes involved Ipsilateral nodes involved Contralateral nodes involved Fixed nodes

M1

M stage No distant metastases Distant metastases

Stage I Stage II Stage III Stage IV

T1 T2 T3 T4

M0

6

Stage grouping NO NO NO-2 NO-3

MO MO MO M1

The stage grouping refers to the worst prognostic feature of the case: e.g. a T1 N1 tumour would be stage III

The TNM system is not applicable to some tumours and is of limited value in others; it is clearly inappropriate, for example, in acute leukaemias, and prognostic factors in these diseases are based on other criteria (see Ch. 23). Nevertheless, staging classification is possible and useful in diseases such as chronic lymphatic leukaemia and myeloma. The criteria have nothing to do with a TNM system, but are related to tumour mass. In lung cancer the TNM system is useful and widely used in squamous carcinoma, but is of considerably less value in small cell carcinoma. In this disease, a rough categorization is made into limited disease (confined to one hemithorax) and extensive disease (spread beyond one hemithorax, including metastases). There is a clear difference in prognosis in these two categories (Fig. 6.12).

STAGING INVESTIGATION Much information can be obtained from a chest X-ray. © As techniques are developed, the sensitivity (ability to detect an abnormality if it is present) and specificity (ability to distinguish a metastasis from other pathologies) of the investigations changes. At present, radionuclide bone scanning is the most sensitive method of detection of bone metastases (Fig. 6.13A). 4 CT scanners are better than,

155

FIG. 6.12 Prognosis of small cell carcinoma of the lung related to the disease extent at presentation

Limited disease = disease confined to one hemithorax. Extensive disease = more widespread local extent or metastasis.

CT has proved of value in detecting lymph node enlargement in the thorax and the abdomen. Unfortunately, nodes can be replaced by secondary tumour and yet not be enlarged. Lymphography - in which contrast material is injected into lymphatics in the dorsum of the feet can show filling defects in normal size nodes, as well as node enlargement. For tumours that spread transperitoneally, such as carcinoma of the ovary, laparoscopy (peritoneoscopy) is the most sensitive method of detecting spread of tumour if a laparotomy is not carried out. Although invasive, this procedure can also be used to obtain liver biopsy material under direct vision.

TUMOUR MARKERS 4

FIG. 6.11

A Carcinoma of the left kidney. T = degree of local spread; N = involvement of adjacent lymph nodes; M = metastasis to bone or other sites. B Prognosis according to stage grouping.

or at least equal to, ultrasound scanning in the detection of hepatic metastases (Fig. 6.13B), 1 but ultrasound is much cheaper and can be repeated regularly to assess response. CT is the best method of detecting pulmonary metastases and is able to demonstrate metastases of 2-3 mm in size (Fig. 6.13C). MRI is superior to CT in detecting deepseated brain tumours, especially in the brainstem and spinal cord, and is the preferred method for showing brain metastases (Fig. 6.13D). 2 MRI is of great value in demonstrating cord compression (Fig. 6.13E) 0 and the soft tissue extent of tumours such as sarcomas (Fig. 6.13F).

156

1

Fig. 6.9

4

MCQ 6.2

0 0

Fig. 6.10

0 Fig. 6.11

Figs 6.9-6.11

Tumours may secrete substances into the blood which can be used both diagnostically (to detect the presence of disease) and in therapy, to monitor the effects of treatment. A variety of biochemical disturbances may be induced in the body by a tumour, such as the rise in alkaline phosphatase caused by liver or bone metastases, but these are crude and insensitive guides to the effect of treatment. The markers in common clinical use (Table 6.7) are oncofetal antigens, placental products, isoenzymes, ectopic hormones and some cellular antigens, usually derived from the cell membrane. Ideally, the tumour marker should always be produced by the cancer in question. In choriocarcinoma, human chorionic gonadotrophin (hCG) is produced in over 95% of cases; and 75% of germ cell tumours secrete either oc-fetoprotein (AFP), hCG or both. Carcinoembryonic antigen (CEA) (a glycoprotein produced by colonic mucosa) is found in the serum in 65% of cases of colorectal cancer. With some markers, e.g. hCG, AFP and placental alkaline phosphatase (PLAP), the serum level of the marker is a reasonably accurate and sensitive reflection

6

FIG. 6.13 Scanning 0 5] Isotope bone scan showing multiple areas of increased uptake in a patient with bone metastases from lung cancer. B Ultrasound scan of liver showing several metastases (arrowed) in a patient with colorectal cancer. C CT scan of thorax showing a solitary subpleural metastasis in the left lung. D MR scan of the brain showing a left-sided frontal metastasis with some associated oedema. The patient had non-small cell lung cancer. E MR scan of the thoracic vertebrae. A large paravertebral mass is invading the left lateral T9 vertebra. The tumour passes through the intervertebral foramen and is displacing the spinal cord to the right. This was a round cell sarcoma of the chest wall. F MR scan showing a large intraosseous tumour arising in the right tibia. The associated soft tissue mass is clearly shown. This was an osteosarcoma.

of the tumour mass in an individual patient, and can therefore be used as a guide to therapy. Prostate-specific antigen (PSA) is a sensitive marker of early prostate cancer which is now widely used in screening. This has led to considerable controversy over the treatment of early, screen-detected, prostate cancer. The antigen is a serine protease produced at low levels in non-malignant prostate disease. The serum marker may be a valuable guide to tumour recurrence and, in most cases of germ cell tumours (such as teratoma of the testis), is a more sensitive indicator of tumour recurrence than any other technique. There is, of course, little value in knowing when relapse is occurring if there is no effective treatment. However, in the case of early-stage germ cell tumours, the markers have allowed a policy of observation following initial surgery, as highly effective chemotherapy is available for relapse. An example of the use of tumour markers is shown in Figure 6.14.

The tumour marker may be produced in other conditions: for example, serum AFP is raised in early pregnancy, and CEA may be present in inflammatory bowel disease. AFP is produced in the fetal liver, in the malignant yolksac cells in teratoma, and in hepatoma. It has a plasma halflife of 5-7 days. hCG is produced by the placenta and by choriocarcinoma, and by the trophoblastic elements in teratoma. The plasma half-life is 30 hours.

PRINCIPLES OF CANCER MANAGEMENT Surgery and radiotherapy are the principal methods of treating the primary tumour. For many early-stage cancers - e.g. in cancer of the breast, squamous cancer of the lung, and cancers of the ovary, kidney, testis and gastrointestinal tract - surgery alone has a good chance of cure. Unfortu-

157

FIG. 6.14 Tumour markers in management of testicular cancer A Both hCG and AFP markers are elevated before orchidectomy but fall following operation (hCG more rapidly because of its short half-life). B No evidence of recurrence, followed by a rise in marker levels. C Combination chemotherapy for recurrence is followed by a fall in levels to normal. D, E A further rise in marker levels is treated more intensively and the patient is disease-free at 3 years F.

TABLE 6.7 Tumour markers present in blood Marker Oncofetal antigens oc-fetoprotein

Tumour

Carcinoembryonic antigen Pancreatic oncofetal antigen

Germ cell tumours of ovary and testis Hepatoma Gastrointestinal cancer Pancreatic carcinoma

Placental products Human chorionic gonadotrophin Placental lactogen Placental alkaline phosphatase

Choriocarcinoma Teratoma Choriocarcinoma Seminoma

Isoenzymes Alkaline phosphatase Lactic dehydrogenase Neuron specific enolase

Osteosarcoma Neuroblastoma, many cancers Small cell carcinoma of lung

Ectopic hormones (see Table 6.4) Other cellular antigens (defined by antisera) Ca 125 (aglycoprotein) Ca 19-9 (a carbohydrate) Prostate-specific antigen (pSA)

158

Ovarian cancer (also breast, pancreas and gut) Pancreatic cancer (also gut) Prostate cancer

nately, many patients cannot be cured surgically, either because the tumour is so advanced locally as to make resection impossible, or because investigation has revealed metastases. Advances in radiotherapy technique and the recent developments in combination cytotoxic chemotherapy have meant that, for many patients, a combined approach

to management is needed to achieve the best results. This involves close collaboration between surgeon, radiotherapist and medical oncologist. Before treatment begins, it is essential to decide whether there is a possibility of cure or whether the intention should be palliative. Many factors will influence this decision, including the degree of spread of the tumour locally, the presence of distant metastases, and the patient's age and state of health. It may be very difficult to decide on the best policy if the chances of cure are relatively small and the toxicity of treatment considerable. In a young patient the aim of cure would probably overcome doubts about toxicity, but this might not be the case in the elderly. If the clinical evidence makes it quite clear that there is no hope of cure, then investigations should only be performed if they will alter management. Investigation, as previously outlined, will be necessary in all other patients to determine the degree of spread of the disease before deciding on the treatment plan. The principles of surgical management of cancer are beyond the scope of this chapter, but it is important for all physicians to have an understanding of the basis of management using radiotherapy, chemotherapy and hormone therapy.

RADIOTHERAPY Most radiotherapy treatment involves the use of electromagnetic waves with very high energy and short wavelength - gamma rays and X-rays. Some modern radiotherapy machines use 60Co as a source, which decays to a more stable form with a half-life of 5.3 years. The process of decay (with loss of a neutron) is accompanied by emission of gamma rays with a high energy (1.2MeV). X-rays were first produced by Roentgen, who heated an electrode in a sealed vacuum tube. The electrode produced electrons which bombarded a target, resulting in the emission of

TABLE 6.8 Relative radiosensitivity of tumours

Depth/cm FIG. 6.15 Tissue penetration of different types and energies of radiation

electromagnetic waves which were designated 'X-rays' and whose nature was the same as that of gamma rays. By increasing the voltage, X-rays of shorter wavelength and higher energy are produced which have greater tissuepenetrating power than those produced from low-voltage machines (Fig. 6.15). Modern machines accelerate the electrons almost to the speed of light, resulting in a high-energy beam. These machines (linear accelerators) have become standard equipment in most departments. They provide a sharper beam than that obtained from 60Co sources and greater depth doses. When the target is moved out of the path of the electrons, the high-energy electron beam can be used for the treatment of superficial tumours. The tissue-penetrating power of high-energy X-rays and gamma rays means that normal tissues will be irradiated as well as the tumour. With lower voltage X-rays, the penetrating power is less and more energy is deposited in superficial tissues, especially bone. With electrons, the radiation decays rapidly in tissue, the depth of penetration depending on the energy of the beam (Fig. 6.15). The important site of radiation damage is nuclear DNA. The damage appears to be induced indirectly. The radiation first produces highly reactive radicals, which in turn damage the DNA, impairing the reproductive integrity of the cell. The damaging effects of radiation are less marked in cells which are hypoxic. Some areas of tumour may not be well vascularized and these cells are relatively radioresistant. The mechanisms of resistance are not well understood. The amount of cell death following exposure to irradiation is also proportional to the dose administered. At low doses the damage is sublethal. Nevertheless, radiation administered as a few large fractions may not be as effective as multiple-fraction regimens using smaller daily doses. The latter technique may be more effective because it allows the tumour to shrink and become less hypoxic, may be easier to tolerate, and allows normal tissues to recover at a faster rate than the tumour. In treating tumours, multiple fractions are usually employed and the total dose administered depends on what is known of the radiosensitivity of the neoplasm (Table 6.8). The total dose

Highly radiosensitive Lymphomas Ewing's sarcoma Seminoma Wilms' tumour Myeloma

Relatively resistant Squamous cell lung cancer Hypernephroma Bladder carcinoma Rectal carcinoma Soft tissue sarcoma (adults) Cervical carcinoma

Moderately radiosensitive Breast cancer Small cell lung cancer Ovarian cancer Medulloblastoma Basal cell carcinoma Teratoma

Very resistant Melanoma Osteosarcoma Pancreatic carcinoma

received by the tumour is measured in grays (1 Gy = 1 J/kg) orrad (100 rad = 1Gy). The planning of the radiation field is often a highly complex procedure that has been greatly facilitated in recent years by the use of CT scanning and simulators, which allow the fields to be constructed with precision. In planning the fields, the radiotherapist must therefore consider the localization of the tumour, its probable inherent radiosensitivity, the tolerance of the normal tissues in the path of the beam, and the need to include draining lymph nodes in the treatment. Multiple fields may be necessary to achieve the desired dose in the tumour without damaging normal tissues. The fields may be open (or direct}. Here, the radiation beam is applied directly, usually at rightangles to the skin. Alternatively, if several fields are used, the field may be wedged. These wedges, inserted into the beam, alter the dose distribution so that the tumour dose is homogeneous and problems of irradiation of normal tissue where fields overlap are minimized. An example of a multiple field is shown in Figure 6.16. The treatment area is immobilized and, for some tumours, such as head and neck cancers, moulds and shells are used to ensure accurate and reproducible positioning. The radiation source is mounted in a gantry which moves isocentrically about the patient. Radiotherapy may also be given by placing a sealed source containing a radioactive isotope inside a body cavity or in tissues. This is known as brachytherapy. Examples of such isotopes are 137Cs, which is widely used as an intracavity source of radiation in treating carcinoma of the cervix, and iridium wires, which may be placed in breast tissue to provide interstitial irradiation when radical radiotherapy is used following conservative surgery for breast cancer. Unsealed sources of radioactivity may be taken by mouth, e.g. 131I in the treatment of thyroid cancer, or given intravenously, e.g. 32P in the treatment of polycythaemia rubra vera.

6

159

TABU 6.9 Complications of radiotherapy Complication

FIG. 6.16 Radiation planning for carcinoma of bladder The relationship of the three beams to the tumour and the dose levels (% of maximum) achieved within the tumour and surrounding normal tissue are shown.

The complications of radiotherapy depend on the radiation sensitivity of normal tissues in the path of the beam, and may be immediate or delayed. A summary of some of these is given in Table 6.9.

SYSTEMIC TREATMENTS In the last 40 years there has been a dramatic increase in interest in the use of drugs to kill cancer cells. So many cancers are either disseminated at diagnosis or too advanced locally to resect, that a systemic approach to treatment offers the only chance of remission or cure. The two treatments in general use are cytotoxic chemotherapy and hormone therapy.

Cytotoxic chemotherapy The earliest compounds to be introduced were nitrogen mustards, which were shown to be cytotoxic to lymphoid tissue and then to have activity against lymphomas in humans. In 1948, Farber demonstrated remissions of leukaemia using the antifolate aminopterin, and this led to the development of a wide variety of antipurines and antipyrimidines. Antitumour antibiotics were first developed in 1940 (actinomycin A), with a significant step forward in 1963 with the isolation of anthracyclines. The platinum drugs were introduced in the 1970s and proved to have great activity in germ cell tumours. The epipodophyllotoxins were also introduced at that time. Recently the taxoids have been introduced. There are now about 35 cytotoxic drugs of proven value in routine use, and the

1

160

MCQ 6.3

Tissue

Immediate

Delayed

Skin

Erythema, desquamation

Oral cavity Gut Bone Kidney

Mucosal ulceration Nausea, diarrhoea Bone necrosis Acute nephritis

CNS

Radiation myelitis and encephalitis

Eye

Conjunctivitis

Gonads Bone marrow

Sterility Leukopenia

Fibrosis Telangiectasia Squamous carcinoma Loss of saliva Fibrosis and stricture Loss of bone growth in children Chronic nephritis Hypertension Demyelination Possible alteration of personality and intellect Dry eye Cataract formation Sterility Suppression of haemopoiesis in area irradiated

RECENT ADVANCES IN NEW METHODS OF ADMINISTERING RADIOTHERAPY The advent of CT scanning has meant that precise three-dimensional images of the tumour can now be obtained. The radiation can now be channeled with considerable accuracy using new methods of directing the beam ('beam's eye view'). A further development has come from the knowledge that repopulation of tumours with new cells occurs rapidly. This has led to the development of hyperfractionated schedules of radiation, in which two or more treatments are given each day. This, and conventional fractionation, may now be given at the same time as cytotoxic chemotherapy in an attempt to improve local control and to prevent interruption of systemic treatment. These complex procedures are likely to play an increasing part in modern cancer management.

increasing complexity of medical treatment of cancer has led to the development of medical oncology as an important specialty in cancer management. Mechanisms of action Cytotoxic drugs produce their effect mainly by damaging the capacity of the cell to divide, and also by interfering with synthesis of essential proteins. For many of these agents the most important toxicity concerns DNA, with impairment of DNA synthesis, reaction with the DNA helix or damage to DNA repair mechanisms. During cell division there is a phase (G1) which precedes the phase of DNA synthesis (S). Following the S phase

SUMMARY 1 Common chemotherapeutic agents Alkylating agents Nitrogen mustard Melphalan Cyclophosphamide Ifostamide Busulphan Chlorambucil Nitrosoureas Cisplatin

Plant alkaloids Vinca alkaloids Epipodophyllotoxins Taxoids Antibiotics Anthracyclines Doxorubicin Daunorubicin Bleomycin

Anti-metabolites Methotrexate 5-fluorouracil 6-mercaptopurine 6-thioguanine Cytosine arabinoside

there is a postsynthetic phase (G2), which is followed by mitosis (M). The daughter cells then either enter into the cycle of division or remain in a non-reproductive stage (Go). Because of their action on DNA, most cytotoxic drugs exert their maximum effect on replicating rather than resting cells. If, after damage to DNA (e.g. by alkylation), the cell does not divide, there will be time for DNA repair mechanisms to excise the attached alkyl groups and to repair strand breaks. Rapidly proliferating cells are thus more susceptible to cytotoxic damage, but this is not the only determinant of susceptibility since, as with radiation, different tissues and tumours show inherent differences in susceptibility, the biochemical basis of which is poorly understood. Alkylating agents are very reactive compounds which exert their effect by linkage of the alkyl group to DNA bases, particularly guanine. If the drug is bifunctional (having two alkyl groups), inter- and intrastrand crosslinks will form. Nitrogen mustard, phenylalanine nitrogen mustard (melphalan), cyclophosphamide, busulphan and chlorambucil are commonly used alkylating agents. Cyclophosphamide itself is inactive but is metabolized by the liver to 4-hydroxycyclophosphamide, which is transported to the tissue and there releases the active alkylating compounds. Nitrosoureas and cisplatin probably act mainly as alkylating agents. Methotrexate inhibits dihydrofolate reductase, thus preventing the formation of tetrahydrofolate, which is necessary for thymidine synthesis. Other antimetabolites include 5-fluorouracil (5FU), which also blocks thymidine formation; 6-mercaptopurine (6MP), which interferes with purine nucleotide biosynthesis; and 6-thioguanine, which has a similar mechanism of action to 6MP. Several plant alkaloids have anticancer activity. The vinca alkaloids interfere with the formation of the mitotic spindle by blocking the assembly of tubulin. The taxoids interfere with disassembly of the microtubules that constitute the mitotic spindle. The epipodophyllotoxins damage

DNA through topoisomerase II. This enzyme normally allows the unwinding of DNA during DNA replication. Anthracycline antibiotics intercalate between DNA bases, as does actinomycin, which has the additional action of interfering with RNA synthesis. Bleomycin causes DNA strand breaks.

6

Testing a new drug When a new drug has shown activity in the laboratory, it is tested in patients with cancer. At first, it is given in gradually increasing doses to patients with advanced drugresistant cancer. These trials are called phase I studies: the aim is to monitor toxicity and to study metabolism and distribution of the drug, as well as to note any activity. In phase II studies, patients with a wide range of tumours are then treated in order to determine the spectrum of activity of the drug. Later, phase HI studies are performed in which the drug is tested, usually in previously untreated patients, either alone or in combination with other drugs, and often as part of a randomized comparison. Combination chemotherapy In recent years it has become standard practice to use cytotoxic drugs in combination. The aim is to produce a variety of different biochemical lesions in the tumour, if possible without increasing toxicity. It quickly became apparent that quadruple therapy in, for example, advanced Hodgkin's disease, produced a far higher complete response rate (disappearance of all symptoms and signs of the tumour) than did single-agent therapy. Furthermore, many of these complete responses were sustained, about 40% of patients being disease free between 5 and 10 years later. This had not been observed with single-agent therapy. Similar results were reported with non-Hodgkin's lymphoma and, later, with testicular teratoma, Ewing's sarcoma and other childhood tumours. In the common tumours of adults (such as cancer of the breast and small cell lung cancer), an increased response rate is also obtained with drugs in combination, but most of these responses are partial (>50% reduction in size) and not sustained for long. Most of the regimens used today have been derived empirically by oncologists with knowledge and experience. There are, however, some general principles governing selection of the drugs (Table 6.10). Using these general principles, combination chemotherapy has been introduced into the treatment of many cancers and forms the mainstay of management in an important minority of uncommon tumours. Unfortunately, most of the common adult cancers are only moderately sensitive to these drugs, and some are very resistant (Table 6.11). Cancer is a disease of the elderly and many patients are not fit enough to risk the toxicity of chemotherapy, especially if the intention is to palliate rather than cure. Complications of chemotherapy 1 There are both immediate and long-term complications of chemotherapy. Many patients greatly fear chemotherapy

161

TABLE 6.10 General principles governing the selection of drugs used in combination therapy • The drugs are known to be effective when used as single agents. • Where possible, drugs with differing modes of action are combined. • The major toxicity of each drug should be as different as possible from that of other agents. • Pulsed intermittent treatment is used to allow recovery of the gut and bone marrow. • If possible each drug should be used in its optimum dose and schedule, although in practice some reduction in dose is nearly always necessary. • There should be no known synergistic toxicity.

as they have often been told lurid stories about toxicity. However, with many regimens toxicity is very mild. The drugs are not equally toxic and it is incorrect for inexperienced physicians to give patients the impression that they are. Immediate The most troublesome immediate side-effects of some chemotherapy regimens are nausea and vomiting, mucosal ulceration, bone marrow depression and alopecia. Nausea and vomiting are probably initiated by a centre in the medulla. Intravenous alkylating agents tend to produce nausea at 12-18 hours, and cisplatin and doxorubicin typically produce nausea about 6 hours after administration. Many other drugs produce very little nausea. Nausea is often accompanied by vomiting, and the symptoms usually last 12-24 hours. These symptoms can be very troublesome but are tolerated by the majority of patients. Skilled, meticulous and flexible use of antiemetics is essential. 5HT3 antagonists have been a big advance in controlling nausea and vomiting, and are now the mainstay of antiemetic treatment. The most widely used agents are ondansetron and granisetron. These are the antiemetics of first choice for chemotherapy which regularly produces troublesome vomiting,

SUMMARY 2 Complications of chemotherapy Long-term Impairment of fertility Secondary cancers Pulmonary fibrosis Cardiomyopathy Nerve damage Loss of hearing Renal impairment

Immediate Nausea and vomiting Mucosal ulceration Bone marrow depression Immune suppression (opportunistic infection) Alopecia

1

162

Case 6.2

2

Figs 6.12-6.15

3

MCQ 6.4

TABLE 6.11 Relative chemosensitivity of tumours Highly sensitive tumours (which may be cured by chemotherapy) Teratoma of testis Hodgkin's disease High grade non-Hodgkin's lymphoma Wilms' tumour Embryonal rhabdomyosarcoma Choriocarcinoma Acute lymphoblastic leukaemia in children Ewing's sarcoma Moderately sensitive tumours (in which chemotherapy may sometimes contribute to cure and often palliates) Small cell carcinoma of lung Breast carcinoma Low grade non-Hodgkin's lymphoma Acute myeloid leukaemia Ovarian cancer Myeloma Relatively insensitive tumours (in which chemotherapy may sometimes produce palliation) Gastric carcinoma Bladder carcinoma Squamous carcinoma of head and neck Soft tissue sarcoma Cervical carcinoma Resistant tumours Melanoma Squamous carcinoma of lung Large bowel cancer

such as cisplatin (Table 6.12). Piperazine phenothiazines (e.g. prochlorperazine) are of some help when given in full dose, but aliphatic phenothiazines (e.g. chlorpromazine) have less antiemetic effect. Metoclopramide is of some value, and its effect may be increased by giving it in high dosage. Benzodiazepines enable the patient to sleep during the period of maximum nausea. Mucosal ulceration in the mouth and gut is a complication of chemotherapy which is particularly likely to occur following methotrexate and anthracyclines. It is dose related, but some individuals are especially susceptible and the dose may then need to be lower than usual. Alopecia is very common with many drugs especially doxorubicin and cyclophosphamide, and is caused by damage to the proliferating cells of the hair follicle. Although the hair regrows when treatment is finally stopped, many patients find this a distressing side-effect. With anthracycline the hair loss can be minimized by scalp cooling during drug administration. The patient must be warned of the likelihood of hair loss, and a wig ordered if desired. Bone marrow depression results in acute leukopenia

TABLE 6.12 Commonly used antiemetic agents Drug

Usual dose

Side-effects

5HT3 antagonists Ondansetron Granisetron

0.15mg/kg 3mg i.v.

Constipation, headache Constipation, headache

12.5mg i.m. or as suppository

Drowsiness Dyskinesia

Benzodiazepine Lorazepam

1-2 mg p.o, or i.v.

Sedation, dizziness, hypotension

Steroids Dexamethasone

6-8 mg i.v.

Restlessness

Phenothiazines Prochlorperazine

10mg p.o. or i.v., Metoclopramide or 1-2mg/kg i.v. (5HT3 and dopamine antagonist)

Dyskinesias, diarrhoea

and thrombocytopenia and the slower onset of anaemia (because of the greater lifespan of red cells). The nadir of the leukocyte count often comes at 7-9 days after intravenous chemotherapy, and it is at this time that the patient is most susceptible to serious infection. 1 Increasingly, oncologists are adjusting the dosage in each cycle of very myelosuppressive regimens according to nadir counts, even though this involves an extra blood count between visits. Patients must be forewarned of their susceptibility to infection and bleeding, so that they can report symptoms promptly. The commonest infections are of the respiratory and gastrointestinal tracts. Immune suppression is cumulative. Patients become increasingly susceptible to opportunistic infection. These are typically pulmonary (aspergillosis, Pneumocystis), oral and oesophageal Candida, and herpes zoster 2 and simplex, which can be severe, confluent and disseminated. Long-term The long-term sequelae of cancer chemotherapy are now becoming better defined as more children and adults are surviving previously incurable cancers. One in 1000 adults now has been cured of cancer in childhood or adolescence. Although complications may be an inevitable risk of regimens that are potentially curative, they are a reminder of the risks attached to increasing the intensity of treatment. Cytotoxic drugs impair fertility in adults. In men, complete and irreversible loss of fertility occurs with many regimens, particularly those that include an alkylating agent or procarbazine. Prepubertal boys do not become infertile, but Leydig cell function may be impaired and testosterone levels may fall. It is important to inform adolescent boys and men of this risk and to arrange sperm storage before chemotherapy begins. Curiously, in Hodgkin's disease, where there is greatest experience of

this problem, patients are often subfertile even before chemotherapy begins. For women, the problem is more complex. Amenorrhoea is common during chemotherapy, especially if an alkylating agent is used. Menstruation will usually return when treatment is finished, although subfertility is common. Chemotherapy may, however, induce the onset of menopause; this is more likely to occur in older women who are nearer the natural menopause. Even in younger women the period of reproductive life may be reduced. It is essential to discuss this potential problem. In both sexes, patients who are receiving chemotherapy should be advised of possible short-term teratogenic effects and to use a contraceptive during treatment. Second cancers are increasingly recognized as a longterm consequence of chemotherapy, particularly in ovarian cancer or Hodgkin's disease. In both cases the risk is especially of acute myelomonocytic leukaemia. Second cancers are also more common in patients receiving immunosuppressive therapy with cytotoxic drugs, and in these patients there is an increased incidence of brain lymphoma, cervical carcinoma and skin cancer. Pulmonary fibrosis is a complication of treatment with busulfan (e.g. for chronic myeloid leukaemia), and acute pneumonitis leading to fibrosis is a dose-related complication of bleomycin. Other cytotoxic agents rarely produce lung damage, although pulmonary fibrosis is a long-term complication of cyclophosphamide and nitrosourea treatment. Anthracyclines such as doxorubicin cause a cardiomyopathy, the risk being dose related. Above a total dose of 400mg/m2 of doxorubicin about 50% of patients will have measurable impairment of cardiac function, but only 5% will develop cardiac symptoms. Established cardiomyopathy leads to cardiac failure, which is irreversible on stopping the drug. Vinca alkaloids cause damage to peripheral and autonomic nerves. The earliest symptoms are of tingling paraesthesiae. There is loss of tendon reflexes and, later, sensory loss and motor weakness. Autonomic neuropathy produces ileus and postural hypotension. Vincristine and vindesine are particularly likely to produce neuropathy, and vinblastine less so. The neuropathy is usually reversible, is totaldose related and is more common and severe in the elderly. Cisplatin may cause irreversible loss of hearing, particularly of high frequencies, and peripheral neuropathy. Renal impairment is also common and irreversible. These toxic effects are a reminder that the treatment of cancer with cytotoxic drugs is a matter requiring skill and judgement if the benefits are not to be outweighed by the disadvantages. Nevertheless, the improvement in prognosis in some cancers as a result of cytotoxic drug treatment fully justifies the acute and long-term risks involved.

6

Hormone therapy3 The growth of many normal tissues is under hormonal regulation, and the cancers that arise in them often retain

163

sensitivity to changes in the hormone environment. Hormone therapy is an essential part of management of cancers of the prostate, breast and endometrium. In these cases, the steroid hormone binds to a hormone-binding molecule (receptor) in the cytoplasm and the receptor/ hormone complex is transported to the nucleus, where it modifies the activity of DNA. There is an increase in RNA polymerase activity, resulting in increasing production of cytoplasmic proteins followed by cell division (Fig. 6.17). In some cases the hormone provokes the increased synthesis of receptors for other hormones; oestrogen, for example, increases the synthesis of progesterone receptor in breast epithelium. Response to hormone therapy is strongly associated with the presence of cytoplasmic hormone receptors. In advanced breast cancer, the presence of oestrogen receptors is associated with a 60% response to hormone therapy but only a 5% response if the cancer contains very low levels of receptor. Response is also strongly associated with the presence of progesterone receptors. The following general approaches are taken in hormonesensitive tumours. Blocking the action of circulating hormones In breast cancer, the use of tamoxifen has been a major advance. It appears to exert most of its effect by binding to the oestrogen receptor, thereby preventing the activity of circulating oestradiol. Paradoxically, the drug is effective in postmenopausal women whose tumours are receptor positive, but in whom circulating levels of oestrogen are likely to be low. A direct growth-regulatory effect of the drug may be responsible. Reduction in concentration of circulating hormones Until recently, this was the mainstay of hormone therapy and was usually accomplished surgically. In breast cancer, oophorectomy was performed, or in premenopausal

164

women a radiation menopause induced. In those who showed an initial response but then relapsed, further responses could be obtained by adrenalectomy, which prevented sex hormone synthesis in the adrenal, and by hypophysectomy which produced a similar effect. The advent of aromatase inhibitors has led to a decline in frequency of these operations. Androstanedione is converted in peripheral tissues (and in the breast itself) to oestradiol by enzymes known as aromatases, and several powerful aromatase inhibitors can now be used to prevent oestradiol formation. In prostate cancer, luteinizing hormone-releasing hormone (LHRH) agonists cause a spurt of LH release followed by a profound fall, which reduces plasma testosterone levels to those seen after castration. Prostate cancer is usually very sensitive to a fall in testosterone levels. LHRH agonists are therefore a widely used medical alternative to orchidectomy.

Addition of hormones The hormonal environment can be altered by the administration of hormones which are judged to provide an unfavourable environment for the tumour. Stilboestrol has for many years been used to produce regression in carcinoma of the prostate, and androgens, glucocorticoids and progestogens may all slow the growth of the tumour in breast cancer. Progesterone derivatives will also produce regression in about 20% of carcinomas of the endometrium. In advanced breast cancer, endocrine therapy has the great advantage of low toxicity and low cost compared to chemotherapy. Tumour responses are often more complete and more durable than those achieved with cytotoxic drugs. On the other hand, the proportion of responding patients is lower (30% with endocrine therapy, 60% with chemotherapy). For these reasons, many physicians prefer to start treatment with hormone manipulation, moving to chemotherapy when the disease becomes unresponsive and progresses.

FIG. 6.17 The mechanism of oestrogen action on cell growth Tamoxifen acts by binding to the oestrogen receptor and blocking its action.

Adjuvant chemotherapy and hormone therapy There are many situations in which a cancer has been removed surgically or treated with radiotherapy, but the risk of local or systemic recurrence is high. Examples include breast cancer, where there is involvement of axillary nodes; gastric and other cancers, where the regional nodes are involved; or Ewing's sarcoma of bone, where the tumour is highly sensitive to radiotherapy but the risk of metastasis is very great. In these situations, advances have been made by the use of cytotoxic drugs or hormones in an attempt to delay or prevent recurrence. Even if these treatments are not very effective against advanced cancer (of the stomach or breast), they may be more so when the residual tumour mass is small, when the penetration of the drugs may be better and the degree of cell kill may lead to cure rather than temporary tumour regression. Many large-scale trials have been undertaken to assess the value of adjuvant cytotoxic or endocrine therapy. For the results to be accepted with confidence, there must be both sufficient numbers of patients in the study, and prospective randomization of the allocation of the treatment policy. (The difficulties involved in these trials are discussed below.) In patients with stage II breast cancer, both adjuvant endocrine therapy (using tamoxifen) and cytotoxic therapy (using drug combinations such as cyclophosphamide, methotrexate and 5-fluorouracil) have been shown to give a survival advantage of about 7% at 5 years and 10% at 10 years (Fig. 6.18). A follow-up overview of all randomized trials now shows that patients with stage I disease also benefit from chemotherapy if they are premenopausal, and from hormone therapy after the menopause. Adjuvant drug therapy after surgery or irradiation has become an essential part of management in most childhood tumours (rhabdomyosarcoma, Wilms' tumour, osteosarcoma, Ewing's sarcoma), and in a number of adult tumours, such as breast cancer, ovarian cancer, and in Duke's C colon cancer. There are many other tumours (gliomas, gastric carcinoma, operable lung cancer) in which adjuvant chemotherapy has not yet been adequately evaluated. As the survival benefit is likely to be small, it will only be demonstrated in very large-scale trials, which are still to be undertaken or in patients data-based metaanalyses. Even a 5 % improvement in survival in colorectal cancer is worthwhile, as the disease affects 1 million people a year in eastern and western Europe.

SUPPORTIVE CARE IN CANCER MANAGEMENT Patients with cancer need supportive care for both the psychological impact of the disease and the physical effects of the tumour and its management. Even if there is no

6

FIG. 6.16 Radiation planning for carcinoma of bladder The relationship of the three beams to the tumour and the dose levels (% of

prospect of cure, the patients' quality of life can be greatly improved by the attitude of the medical staff and by meticulous attention to the management of pain, anaemia, infection, nausea and anorexia.

Psychological support The last 20 years have seen an increasing readiness on the part of doctors to discuss the diagnosis of cancer with their patients and to explain the principles of management. Indeed, the complexities of modern cancer management make it difficult to withhold the facts of the diagnosis, even if it were desirable to do so. Nevertheless, the manner in which the diagnosis is explained and the words used are of great importance, and must be carefully judged in each case. Cancer is the only word that conveys the nature of the complaint accurately; other words used are often misinterpreted by patients as implying a less, or even more, serious condition. The problem is that in the mind of many patients cancer is invariably fatal. Explanation of the diagnosis must therefore be accompanied by an unhurried and easily comprehended account of management and the possibilities of cure. It is nearly always wrong to offer no hope of either cure or a period of normal life. Usually, the physician will need to talk to the patient on several occasions, as their understanding of the problems may be incomplete at the beginning. Time taken in unhurried, private conversation is always well spent and greatly appreciated. One of the advantages of having specialized oncology units in hospitals is that all members of the medical team - nurses, doctors, social workers and counsellors - become familiar with the problems that are associated with malignant disease and skilful in dealing with them. It is important that all members of this team know what has been said to the patient in the way of explanation, in order to avoid conflicting advice. A relapse of the disease or the appearance of a new symptom will often cause a slump in morale and confidence which will require more explanation and reassurance. The emotional demands of this aspect of cancer

165

medicine are considerable, and are one reason why some doctors have been reluctant to spend time in discussion with their patients. Nevertheless, the patient will need the help and support of the doctor at these difficult times. It is an essential aspect of cancer medicine and one of the most rewarding.

PAIN RELIEF This is a very important aspect of management, particularly in patients with widespread disease who are terminally ill. No single analgesic will suit all patients all of the time. A patient will need to change from one to another, depending on the severity of pain. A brief classification of some of the most useful analgesics is given in Table 6.13. When using opiates, the dosage should be adjusted to the patient's requirements. A wide range of dosages is encountered. Long-acting morphine and heroin are especially useful and should be given regularly, rather than 'as required'. The constipating effects of opiates can be partly relieved by regular laxatives. If pain is at a particular site, it may be possible to give relief by local measures, such as radiotherapy to a bone metastasis, a nerve block or a coeliac plexus block. The advice of colleagues specializing in these forms of pain relief is often helpful, and the procedures may allow a reduction in dosage of opiate.

Nutritional support

166

Patients lose weight because of the local effects of the tumour (especially if this is in the alimentary tract), metastases, and as a result of treatment. Successful treatment of the primary tumour (e.g. by surgery) will be accompanied by a return of appetite and weight gain. During radical treatment with surgery, radiotherapy and intensive chemotherapy, weight loss is common. Calorie supplements can be given both in the form of oral supplements and intravenously to patients undergoing cancer treatment. However, the value of nutritional support is difficult to demonstrate. It is usual practice to consider total parenteral (intravenous) nutrition in patients undergoing exceptionally intensive treatment, such as high-dose chemotherapy or total-body irradiation with bone marrow transplantation, or during the intensive phases of chemotherapy for acute leukaemia, where there is prolonged hypoplasia. At these times, mucosal ulceration, nausea and diarrhoea make it difficult to feed the patient by mouth, and nasogastric tubes may produce monilial infection or exacerbate acid reflux and oesophagitis. Parenteral nutrition is also used for patients undergoing extensive surgery, particularly gut resection. The aim is to maintain the patient's weight and nutritional status so that recovery is more rapid, allowing further treatment with chemotherapy or irradiation to begin as soon as possible. There is no evidence that additional enteral or parenteral feeding is of value in the management of advanced

TABLE 6.13 Useful analgesics Drugs Mild analgesics Aspirin Paracetamol Indole derivatives (e.g. indomethacin) Propionic acid derivatives (e.g. ibuprofen)

Duration of action (hours)

4-6 2-4 6-8 4-6

Moderate analgesics Codeine (and dihydrocodeine) Pentazocine Dipipanone Oxycodone

4-6 3-4 6 8

Strong analgesics Morphine sulphate Diamorphine Dextromoramide Pethidine Methadone

4-6 (up to 12 with sustained-release forms) 4-6 4-6 3 12-30

cancer where cure is impossible. It is, however, important to encourage the patient to eat by giving food that appeals (often in the form of smaller, more frequent meals), by controlling nausea with drugs such as metoclopramide, and by the careful use of corticosteroids which improve wellbeing and stimulate the patient's appetite.

Tumour lysis syndrome In some very sensitive untreated cancers, chemotherapy may produce massive tumour breakdown when treatment is first given. This process may be accompanied by severe metabolic disturbances, which constitute the tumour lysis syndrome. The syndrome is usually associated with the early phases of treatment of lymphoma and leukaemia in children and, less frequently, in other chemosensitive childhood tumours. In adults the syndrome is infrequent but may occur with treatment of lymphoma, especially if there is impairment of renal function (such as ureteric obstruction by lymph nodes), which diminishes renal excretion of the tumour products. Tissue destruction releases large amounts of urate, phosphate and potassium. The urate may be deposited in the renal tubule, causing reversible renal failure, which in turn diminishes further urate excretion. Severe hyperuricaemia and renal failure may necessitate dialysis. This complication can be largely avoided by using the xanthine oxidase inhibitor allopurinol (100-200 mg 8-hourly) before and during the early phase of treatment, and by establishing and maintaining a diuresis during the first 24-48 hours.

Hyperkalaemia may be severe, particularly if renal failure occurs, and may need treatment with glucose and insulin (see Ch. 21). Hyperphosphataemia can be partly prevented by hydration and diuresis before and during the initial treatment.

NEW APPROACHES TO MANAGEMENT Progress in cancer treatment is slow, and results usually improve after many small refinements in therapy, rather than by sudden steps forward. Since the late 1970s only modest progress has been made, and the search for new drugs and new approaches continues.

Haemopoietic growth factors (see also Ch. 23) The production of haemopoietic growth factors by recombinant technology has allowed their use in chemotherapy to reduce the period in which the white count is low. Proven advantages are a modest reduction in the period of neutropenia and somewhat diminished antibiotic use. It is possible that the intensity of chemotherapy may be increased by avoiding dose reductions. It has yet to be proved that the routine use of growth factors increases the cure rate of cancer by allowing increasing dose of chemotherapy. They are used if the dose of a curative treatment would otherwise have to be reduced but can be maintained if growth factors are given. Granulocyte colony-stimulating factor (GCSF) is most widely used. It is given for 7-10 days subcutaneously following chemotherapy. It may cause fever and bone pain. It is expensive and should only be given if there is a clear indication, and not if treatment is palliative.

High-dose therapy Allogeneic (from one individual to another) bone marrow transplantation was introduced for the treatment of acute myeloblastic leukaemia in the early 1970s. Allogeneic marrow was infused after an attempt had been made to eradicate the leukaemia by total-body irradiation and high-dose cyclophosphamide. The results (Ch. 23) showed that the leukaemia might be eradicated. This finding led to the consideration of total-body irradiation in the management of other disorders, such as lymphomas and myeloma. More recently, attempts have been made to eradicate leukaemia, lymphoma and some solid tumours with very high-dose chemotherapy (rather than total-body irradiation) using autologous (taken from, and reinfused into, the same individual) bone marrow, harvested just before treatment, as a means of preventing life-threatening myelosuppression. These approaches are based on the assumption that increasing doses of chemotherapy will eradicate residual disease when multiple conventional doses will fail.

This approach has been revolutionized in the last few years by the finding that the haemopoietic progenitor cells appear in large numbers in the peripheral blood following chemotherapy and the use of haemopoietic growth factors. These cells can be 'harvested' by a leukopheresis machine and used, instead of bone marrow, to reconstitute the blood count following high-dose treatment. This is called a peripheral blood stem cell (PBSC) transplant. High-dose therapy and PBSC transplant has now become a safe and much cheaper treatment allowing further exploration of the value of high-dose therapy.

6

Monoclonal antibodies to cell surface antigens Many tumours express oncofetal or differentiation antigens not usually found in the normal tissue from which the cancer is derived. An example are the antigens found on the surface of both acute lymphoblastic leukaemia cells, lymphomas and normal lymphatic progenitor cells. The development of monoclonal antibodies of defined specificity has resulted in the production of monoclonal reagents, which bind preferentially to antigens on the surface of some tumour cells (and to those normal cells that express the antigen). Cytotoxic drugs, radioactive isotopes, and enzymes capable of converting an inactive cytotoxic agent to an active one, can be attached to such antibodies with the aim of selectively destroying the tumour. There are considerable difficulties with this approach: heterogeneity of antigen expression within the tumour; variable uptake and distribution of antibody in normal and malignant tissue; and the production of antibodies by the patient against the foreign (usually mouse) protein. Nevertheless, there are interesting future possibilities for therapy and tumour localization using this approach.

TRIALS OF TREATMENT Many cancers have a long and unpredictable natural history: cancer of the breast, for example, is associated with an excess mortality for over 25 years after diagnosis. If survival is the endpoint of treatment, assessment of therapy in this disease will mean prolonged periods of observation. In small cell lung cancer or acute myeloblastic leukaemia, on the other hand, a majority of patients are dead within 2 years, and relapse in those who survive is less frequent beyond that point. Treatment trials will therefore give answers quickly. In assessment of therapy of a cancer (such as breast cancer), there may be more than one endpoint to be considered: does the treatment give effective local control: does the treatment influence the onset of metastasis; is the proportion of survivors at 5 and 10 years or longer increased? Comparison of a treatment with historical controls is very unreliable, as diagnostic and staging criteria are changing constantly and selection of cases is thereby

167

altered. Randomized prospective studies are now regarded as the only reliable way of making a comparison between two treatments. Such studies must include sufficient numbers of patients to allow the questions to be answered with confidence statistically. The studies require careful documentation and prolonged observation of the patient groups. The treatments being compared must be justifiable ethically and the design of the study capable of giving a clear answer to the central question. These considerations mean that treatment trials in cancer are complex and usually demand collaboration between many different centres: even, in the case of rare tumours, between different countries. To detect a 5 % difference in survival with confidence may require many hundreds or even thousands of patients in a study. Nevertheless, a 5% difference in long-term survival in breast cancer, if achieved by a non-toxic hormonal means, would mean 1000 lives saved a year in the UK alone - more than the number of patients cured of Hodgkin's disease. However, a 5% difference achieved by means of toxic chemotherapy might be regarded with less enthusiasm. At present, because progress in cancer management is slow and achieved only by small improvements, large-scale clinical trials are the only reliable way of validating new treatments.

CARE OF THE DYING

168

Many patients with cancer die from their disease; it is the responsibility of the oncologist to be sure that aggressive management of the tumour is not thoughtlessly continued beyond the point at which cure is possible. This does not mean that treatment is discontinued, but that the aim of management becomes palliative rather than curative. In skilled hands palliative treatment can often give the patient a period of happy life free from distressing symptoms. Palliation may require radiotherapy, e.g. for painful bone metastases or bronchial obstruction; chemotherapy at low dose to suppress constitutional symptoms in lymphoma; aspiration of ascites or pleural effusion; or surgical pinning of a bone weakened by metastasis. There is more to palliative management than pain relief. Nevertheless, relief of pain, control of nausea and treatment of depression become increasingly important as the disease progresses. The most important component of management at this stage is the psychological support of the patient and his or her family. Support comes from the nature of the relationship that the patient has established with the medical team throughout the illness. Extra help may be needed from domiciliary nursing teams, which are increasingly providing expert and readily available advice on symptom control to patients in their own homes. Many patients prefer to remain at home in the last few weeks of life, and the family doctor and specialist nurses can often make this possible. Alternatively, the patient's wishes or family circumstances may make admission to a hospital or hospice necessary.

The nursing practices and the approach to symptom control that have been developed in hospices are now being incorporated into hospital practice. In some hospitals, nurse specialists have been appointed to advise on these aspects of management. The gradual awareness of death produces many different reactions. The patient may become angry and direct this at the medical or nursing staff. Feelings of apathy, hopelessness and fatigue are common, and depression may be severe enough to need treatment with antidepressants. The support of relatives and friends is essential. Meticulous attention to detail and a willingness to take time to talk to the patient are just as important at this stage in management as during the early stages when the intention may have been curative. Even though active management of the cancer has finally failed, this does not mean it was not worthwhile, nor is it a reflection on the abilities of the medical or nursing staff. The fact of treatment failure leads some doctors to avoid seeing their patients at this stage. This is bad practice and increases the patient's feeling of isolation and helplessness. Many non-medical agencies are able to help the patient and his or her family. Patients with religious beliefs may gain great comfort from their clergyman and other members of their local religious community. Counselling and support groups have been set up in many towns to provide help during the illness and to the bereaved relatives afterwards.

ALTERNATIVE THERAPY A multitude of unorthodox approaches to management of cancer have been offered to patients over the years. These usually have echoes of current medical practice. Serotherapy was used after antitoxins were introduced for infectious disease; tooth extraction was practised, based on the interwar notions that dental sepsis was responsible for chronic inflammation. Nowadays, psychological ideas are translated into visualization therapy, and concern over pollution and diet are transmuted into strict vegetarian or fatfree diets. The uncritical assumption that answers will be found in nature has led to a wide range of herbal remedies, some of which (e.g. laetrile) are somewhat dangerous. As with many chronic diseases, the practitioners of homeopathy and faith healing offer their remedies. Most of these approaches share certain recognizable characteristics. These include a tendency to borrow the jargon of medicine without its science; a belief that all tumours can be treated by the same or similar remedies; a tacit or explicit assumption that there exists a medical conspiracy to conceal the truth about the remedy in question; and a complete failure to adopt any of the proven methods of validation of the results of treatment. What these approaches offer distressed and anxious patients is a feeling that they can themselves do something to combat their disease, a glimmer of hope that they might

be cured, and relief from the side-effects of conventional treatment if this is abandoned. These benefits are often a reflection of inadequacy in the relationship of the patient and doctor, rather than a clearly held view by the patient and family. The damaging aspects of these approaches are that they may induce the patient to abandon a worthwhile treatment, weaken the relationship between doctor and patient, and impose unpleasant constraints on the patient's life, such as rigid dietary control. These disadvantages should be discussed with the patient and, if patients are thinking of abandoning a potentially curative treatment, every attempt should be made to persuade them not to do so. Nevertheless, the doctor's role in management is to offer help, advice and expertise; the patient has the right to choose what to do. In many cases no great harm will be

done, and an experienced doctor will make it clear that the patient is welcome to continue under his or her care now and in the future.

6

FURTHER READING DeVita V T, Hellam S, Rosenberg S 1997 Cancer: principles and practice of oncology, 3rd edn. Lippincott, Philadelphia. A comprehensive cancer textbook. Souhami R L, Tannock I, Hohenberger P, Horiot J C 2001 Oxford textbook of oncology, 2nd edn. Blackwell, Oxford. A comprehensive textbook of oncology. Souhami R L, Tobias J S 1997 Cancer and its management, 2nd edn. Blackwell, Oxford. A medium-sized text covering most aspects of cancer treatment.

169

This page intentionally left blank

7 Ageing and Disease Peter W Overstall

Theories of ageing 172

Altered reactions to disease 189

Sociological aspects 173

Clinical approach 191

Special features of disease in old age 174

Investigation in the elderly 193

Autonomic dysfunction 178

Assessment of fitness for operation 196

Nutrition and ageing 179

Prescribing for the elderly 197

Balance and falls 180

Cholesterol lowering in stroke 200

Urinary incontinence 185 Faecal incontinence 187 Sleep and insomnia 187

Rehabilitation 201 Community care and long-stay care 202

Sexual function 189

Most ill health in developed countries now occurs in elderly persons. Not only are the elderly more prone to illness, but the proportion of elderly people has increased considerably in the last 100 years. The response of an individual to growing old depends more upon the previous pattern of life than upon anything in the ageing process itself, so that one of the features of old age is increased differences between individuals.

LIFE EXPECTANCY Life expectancy is the average observed years of life from birth, or from any stated age. Since 1900, life expectancy at birth in the UK has increased from 49 to 75 years for men and from 52 to 80 years for women (Fig. 7.1). This is mainly the result of reductions in infant mortality from the major

1 MCQ 7.1

infectious diseases. Tuberculosis, acute rheumatic fever, smallpox, diphtheria, tetanus and poliomyelitis now account for less than 2% of the ill health they caused in 1900. These changes are largely due to improvements in nutrition and to public health measures which have reduced water- and food-borne diseases. The influence of immunization and treatment on mortality is comparatively much smaller. Life expectancy may also deteriorate sharply. In the former Soviet Union and eastern bloc countries male life expectancy has fallen since 1980 from 62 to 58 years. This has been blamed on the disruption caused by the transition to market economies and the resulting poverty, unemployment, homelessness, excessive drinking and smoking. Life expectancy in sub-Saharan Africa has also fallen since the 1980s because of HIV/AIDS. Much interest was aroused by early studies in industrialized countries which suggested that there was a connection between income inequality and life expectancy, implying that income inequality is bad for the whole population. It has now been shown that there is no association at the population level, although household income has a powerful effect on mortality. In the UK life expectancy at birth for daughters of semiskilled and unskilled manual workers is 3 years less than for those born to professional and managerial classes. For their brothers there is a 5-year class divide.

CHANGES IN THE ELDERLY POPULATION The age structure of a population is affected by infant mortality and by fertility rates, late-age mortality and migration. Demographic ageing, i.e. a shift towards a larger proportion of elderly people in the population, began in Europe in the early 20th century, as a result of falling fertility. This change has already occurred in the UK, so that during the 1990s there was very little overall change in the population aged 60 years and over, but a rapid expansion of the over-85s. After 2001, the cohort of the post-1945 'baby boom' reaches retirement and there will then be a further rise. The proportion of elderly in the population will not increase thereafter unless there are further substantial falls in both fertility and late-age mortality. Fertility rates are declining worldwide and we can anticipate the most rapid changes in age structure of the population to be in the developing world (Fig. 7.2). Note the narrowing of the base of the pyramids during the 21st century due to lower fertility rates. The population pyramids in Europe and the western Pacific are starting to invert. Already over 60% of older people live in developing countries, and by 2020 this will reach 70%. Absolute numbers of elderly in the world will increase from 385 million now to 1.5 billion in 2050. Most of this increase will occur between 2025 and 2050, as children bom in the 1960s reach the oldest population cohorts. 1

171

FIG. 7.1 Life expectancy at birth (both sexes) in WHO regions (weighted by population)

SUMMARY 1 Epidemiology of ageing • In the UK, the increase in life expectancy at birth seen in the early part of the 20th century was due largely to a decrease in infant mortality. • The rising percentage of elderly in the population is due to falling fertility and lower mortality rates, particularly in the very old. • During the 1990s there was a pause in the overall growth of the elderly population in the UK, but the over 85s increased by onethird. • Developing countries face a growing number of elderly in the first half of the 21st century.

FURTHER READING ON LIFE EXPECTANCY Mackenbach JP 2002 Income inequality and population health. BMJ 324: 1-2. FIG. 7.2 Population pyramids of WHO regions

THEORIES OF AGEING

172

Ageing is a complex process resulting from intrinsic and extrinsic damage to the organism at a rate dependent on each particular individual's genetic background and environment. The resulting disability is not fixed or inevitable. For example, high-tone deafness and high blood pressure are common in elderly Britons, but are absent among

elderly persons in the Easter Islands. Osteoporosis is common in western Europe and the USA but rare in China. Thus descriptions of physical decline are too variable to be useful for defining ageing. A better approach, which allows populations to be compared and so highlights differences in environmental or intrinsic ageing factors, is to define ageing

as a process which brings about an ever-increasing age-specific mortality (i.e. the older the individual becomes the more likely he or she is to die). In the UK mortality is at its lowest around puberty, and ageing can be said to begin at this point, rising progressively throughout adult life. There is general agreement that ageing is at least partly genetically determined. Each animal species has a specific lifespan: the maximum recorded longevity of Marion's tortoise, for example, is over 152 years, for man 122 years, and for golden hamsters 2 or 3 years. In species of Drosophila, the Fl hybrid has a greater longevity than either parental strain. Female sex also generally confers increased longevity in a number of different species. Individuals whose parents live to the age of 75 years or more live longer than those whose parents died before the age of 60, and siblings of centenarians have a four times greater chance than average of surviving to their early 90s. The mean difference in longevity in dizygotic twins is twice that of monozygotes. Current theory sees the ageing process as non-adaptive, with ageing evolving as a late by-product of processes that benefit the individual during its earlier, more fecund lifespan. With limited energy resources, an organism has to strike the right balance between investing in bodily maintenance and repair, and producing and rearing its young. Oxidative metabolism plays a key role in cellular ageing. Over the years free radicals cause increasing damage to enzymes, structural proteins, and nuclear and mitochondrial DNA. Mitochondria are the main intracellular source of free radicals, and with increasing age more free radicals are generated while at the same time antioxidant production declines. This suggests that defective mitochondria are central to the process of ageing. The only proven way to delay ageing in higher animals is to eat a low-calorie diet. A 50% calorie diet restriction increases the lifespan of mice from 38 to 56 months. Whether humans would show a similar gain is unknown, but restricting calories by onethird results in lower blood pressure, cholesterol, glucose and white blood cell counts.

FURTHER READING ON THEORIES OF AGEING Kirkwood T 1999 Time of our lives: the science of human ageing. London: Orion Publishing.

SOCIOLOGICAL ASPECTS Retirement age is an administrative measure without relation to the capacity of individuals. Most people at retirement can expect good health for several years, and for developed countries the World Health Organization has introduced a new demographic indicator of 'life expectancy without incapacity'. At age 65 this is estimated to be between 8 and 11 years for men, and between 9 and 12 years for women. People in developed countries are main-

taining better health in later life than ever before, and severe disability in older people is declining at 1.5% per annum. For example, it is anticipated that in the USA the number of severely disabled elderly will halve between 2000 and 2050. The notion that the older you get the sicker you get has been challenged by the finding that centenarians appear to have lived most of their lives in good health, with a relatively rapid terminal decline. This suggests that the older you get the healthier you have been, and promises a potential reduction in morbidity and disability as people approach the limits of their lifespan. None the less, medical costs are four times greater in the first decade of retirement than during working life, and by the second decade of retirement they are nine times as great. The top diagnostic categories for healthcare costs for the over-70s are dementia, stroke, musculoskeletal disorders and falls.

7

ATTITUDES TO THE ELDERLY In Europe the falling birth rate, increased longevity and trend to early retirement has reduced the workerpensioner ratio. In 1901 4.7% of British citizens were aged 65 or over; today this figure is 15.6%. This striking change had occurred by the 1980s (since when it has levelled off) and was not accompanied by undue economic strain or intergenerational strife, despite concerns in the 1960s and 1970s about 'dependency ratios' (the ratio of non-working people to those of working age). Discrimination against elderly people springs from three main impulses: a fear of death and dying, a social belief that falsely associates success with productivity, and misconceptions about the inevitability of senility. There is, in fact, little or no sign of a significant decline in happiness or life satisfaction with age. Over 80% of elderly people maintain independent house-holds, three-quarters of those over 85 can wash all over unaided, 95% can go to the toilet alone, and only 5% of people over 75 have significant dementia. It is true that institutionalization rates for the very old have increased considerably in the last 30 years owing to a decline in the amount of family support available to the elderly. Women aged between 45 and 60 have carried the brunt of community care, and since 1900, when there were 83 women in this age group for every 100 people over 65, there are now only 45. None the less, families do still care for their elderly relatives and indeed provide far more support than is available from health or social services. Family support ratios compare the relative sizes of older and younger generations and indicate the availability of family support for the elderly (Fig. 7.3).

RETIREMENT Retirement is now a well established part of life in developed countries and is starting to appear in developing countries among government employees and urban

173

FIG. 7.3 Parent support ratio in WHO regions (weighted by population) The increase in the parent support ratio (population aged over 80 divided by the population aged 50-64) is most marked in developed countries, particularly Japan (increase in ratio from 8 to 41 between 1975 and 2025).

174

workers (Fig. 7.4). Most people no longer see it as a traumatic event marked by major losses to the individual, but instead prepare for it as a normal and expected part of the life-cycle and welcome it as a time of opportunity and increased leisure. The rate of early retirement is, therefore, rising. Since 1975 there has been a decline in the UK of about one-third in the number of men aged 60-64 in fulltime employment. Of course, not all of these early retirements are voluntary, and there has been a noticeable trend for businesses in economic difficulties to make their older rather than their younger workers redundant. The period of active retirement, in which many take up part-time or voluntary work, ends when the person becomes disabled by declining physical or mental health. The difference between these two stages of old age depends on functional capacity, not on chronological age. Some people do find that retirement produces troublesome stresses. The most frequent problem is adjustment to a lower level of income. Some suffer from a loss of selfesteem, others fear the onset of ill health, or are distressed by their idleness. Most of these problems fade as the new pattern of activity is taken up, and serious depression is not a particular feature of retirement. Claims of large numbers of super-centenarians living in certain geographical regions - such as Vilcabamba in Ecuador, Hunza in Pakistan and Abkhasia in Georgia have been shown to be exaggerated. However, it is true that the inhabitants of each of the above areas are relatively long-lived, and in each case the esteem and merit enjoyed by a person increases with age. There is no retirement age

FIG. 7.4 Percentage of the population aged 65 years and over that is economically active worldwide The developed countries have still to find a satisfactory role for their unemployed elderly.

and old people continue to play an active role in the economic and social life of their communities.

FURTHER READING ON SOCIOLOGICAL ASPECTS Grundy E 1996 Age, 'dependency' and intergenerational relationships. Rev Clin Gerontol 6:303-304.

SPECIAL FEATURES OF DISEASE IN OLD AGE Measurement of physiological functions in healthy young adults gives fairly uniform results, with a narrow range between the upper and lower limits of normal. From about 30 years of age, a functional decline in performance can be detected in most organs and body systems (Fig. 7.5). However, the organs and systems age differentially, so that the variation both within and between individuals increases with age. The chronological age of an old person may therefore bear little relation to his or her functional age (Fig. 7.6), and the response of an old person to the stress of illness will be more unpredictable than that of a young adult. Degenerative changes occur throughout the body with increasing age, but these may become sufficiently marked to constitute a pathological process. The distinction between this and normal physiological ageing is often difficult to make, and there is increasing recognition that so-called normal ageing is the result of occult pathology. These changes have an important bearing on the management of the patient, as will be seen in the examples that follow.

TABLE 7.1 Good practice guidelines in osteoporosis 1. Frail housebound elderly should be prescribed calcium and vitamin D, assessed and managed for falls risk and considered for hip protectors.

7

2. Patients with osteoporosis risk factors should have their bone mineral density measured (DXA at the hip). If their T-score is above -1 they can be reassured. If between -1 and -2.5 they should be given lifestyle advice and osteoporosis treatment if there is a history of a previous fracture. If the T-score is below -2.5 they should have lifestyle advice and osteoporosis treatment.

30

40 Age (years)

3. Patients with a previous fragility fracture should have routine investigations (blood count, ESR, bone and liver function tests, serum creatinine, serum TSH) and, because they are at high risk for a further fracture, should be started on osteoporosis treatment and given lifestyle advice.

FIG. 7.5 Age decrements in physiological functions in males A Fasting blood glucose; B resting cardiac index; C maximum breathing capacity; [6] maximum work rate.

cause of visual impairment is age-related macular degeneration. This is untreatable, but about 80% of patients can improve their reading performance with simple magnifying devices and brighter illumination.

THE AGEING EAR

FIG. 7.6 The relation between resting cardiac output and age in men without circulatory disorders Note that several men over the age of 70 have values similar to men in their 30s.

Impaired hearing increases with age and can contribute to poor health, social isolation, depression and (probably) paranoid psychosis. About 30% of the elderly regard themselves as having a hearing impairment, but audiometric testing reveals a much higher prevalence (60%) in those aged over 70. Presbyacusis is age-related loss of high-frequency hearing, but this cannot be reliably distinguished from the effect of ototoxic drugs, diuretics or vascular disorders. Some of the hearing loss usually regarded as an inevitable part of old age may be due to our noisy environment.

OSTEOPOROSIS (See Table 7.1)

THE AGEING EYE The force required to focus the eye increases considerably at about 40 years of age and, by 45 or 50, accommodation will be so poor that for most persons small print can be read only at arm's length (presbyopia). This is a result of both a reduction in elasticity of the lens capsule and a thickening of the lens, which prevents it adapting its shape to focus on near objects. Opacities and general yellowing of the lens, together with an age-related meiosis, cut down the amount of light entering the eye and reduce visual aculity. A loss of cells in all regions of the visual pathway affects perception and contrast sensitivity. About 20% of those aged 75 and over have a visual acuity of less than 6/12. With continuing advances in the management of cataracts, glaucoma and diabetic retinopathy the main

Postmenopausal (type 1) osteoporosis results from accelerated bone loss due to oestrogen deficiency. There is mainly a loss of trabecular bone, typically resulting in fractures of vertebral bodies and the distal forearm in women in their 60s and 70s. Senile (type 2) osteoporosis is a slower, age-related bone loss that occurs in both sexes. There is cortical as well as trabecular bone loss, which typically results in femoral neck fractures in men and women in their 70s and 80s. These fractures are a serious threat to older people and are associated with a 20% increased risk of either death or institutional care over the following year. Secondary causes of osteoporosis include endocrine (thyrotoxicosis, hypogonadism, hyperadrenocorticism) and gastrointestinal (malabsorption, primary biliary cirrhosis) disorders, rheumatoid arthritis, malignancy, hypertension, and drugs such as corticosteroids and heparin. Other risk

175

factors include a strong family history, inactivity, excessive alcohol and smoking. Unfortunately, clinical risk factor profiles correlate poorly with both bone density and fracture risk, and the most useful way of assessing increased osteoporosis risk is to measure bone density by dual energy X-ray absorptiometry. Ultrasound assessment of the skeleton is a promising alternative method. The predictive value of bone density is similar to that of blood pressure in determining the risk of stroke. The relative risk of fracture increases two to three times for each standard deviation decrease in bone density. Definite indications for bone density measurements include fracture after minimal trauma in previously fit individuals, past history of early menopause (<45 years) in women up to the age of 70 years, underlying secondary causes of osteoporosis and apparent osteopenia on X-ray. There is currently no evidence to indicate that routine population screening is valuable, and instead a case-finding strategy is used. Bone density measurements are not needed in patients at greatest risk, such as those over 80 years who are frail and housebound or institutionalized, those with a previous fragility fracture and patients taking prolonged oral steroid therapy (>7.5 mg prednisolone/day), as they should be treated regardless of their bone density. Prevention of osteoporosis depends on achieving the maximum peak bone mass (regular weight-bearing exercise and dietary calcium intake of 1000-1500 mg/day, between the ages of 15 and 25) and reducing the rate of bone loss by encouraging elderly people to keep exercising, maintain a daily calcium intake of 1000 mg, avoid falls (see p. 175), stop smoking and moderate alcohol intake. These are all worthwhile interventions at the population level. For the individual postmenopausal patient at risk, particularly if she has climacteric symptoms such as hot flushes, the single most effective intervention is hormone replacement therapy (HRT). The greatest benefit is gained by starting treatment soon after the menopause. For current users of HRT there is a 65% reduction in hip fracture risk but 5 years after stopping HRT most of the benefit is lost. For optimal protection HRT should be continued for long periods. Long-term use increases the risk of breast cancer; for women aged 50-70 the risk of developing breast cancer increases from 51 cases per 1000 women to 57. HRT reduces the risk of coronary artery disease but the benefit is not as great as early studies suggested. The risks of developing Alzheimer's disease appears to be reduced. Raloxifene is an alternative to HRT for the postmenopausal patient without climacteric symptoms, which can be exacerbated by raloxifene. It appears to be associated with a lower incidence of breast cancer and does not cause

1

176

MCQ 7.2

2

Fig. 7.1

endometrial bleeding, but, like HRT, it can cause deep vein thrombosis. For the elderly patient HRT is rarely indicated and bisphosphonates or calcium and vitamin D are used instead. The best evidence for efficacy in reducing spinal, non-vertebral and hip fractures is for alendronate and risedronate. Calcitonin has proven efficacy for preventing spinal fractures and is also effective in relieving the pain associated with vertebral fractures. Calcium and vitamin D have proven efficacy in preventing non-vertebral and hip fractures, and it has been argued that these supplements should routinely be prescribed for frail housebound elderly persons. Corticosteroid treatment doubles the risk of fractures of the hip and distal radius and quadruples the risk of vertebral fracture. All patients over 65 years taking prednisolone 7.5 mg a day or more should have treatment to prevent osteoporosis: the first choice is a bisphosphonate1212.

FURTHER READING ON OSTEOPOROSIS Royal College of Physicians, 2000. Osteoporosis: clinical guidelines for prevention and treatment. RCP, London.

DEMENTIA AND NORMAL AGEING Cross-sectional and longitudinal studies consistently demonstrate a decline in cognitive function with increasing age. A decline in cognitive ability is not confined to old age, and tests that measure the ability to solve novel problems rapidly (fluid intelligence) show a fall in performance by the late third and early fourth decades. Age-related changes in cognitive function appear to be related to a general decline in the speed at which information is processed. Thus, highly practised skills, such as vocabulary test scores, alter very little with age, but timed novel problem solving, reaction time tests and the ability to learn new material decline with increasing age. Ill health undoubtedly affects cognitive performance, particularly cardiovascular disease. Congestive cardiac failure doubles the risk of cognitive impairment, and systolic hypertension in middle age increases the likelihood in old age of developing white matter hyperintensities (on MRI) and dementia. Other predictors of cognitive decline are increasing age, 'old age forgetfulness', low educational status, lack of social and physical activities and possession of the ApoE-4 allele. The major cause of intellectual loss in old age is Alzheimer's disease (AD), followed by vascular dementia and Lewy body dementia. There is increasing recognition that vascular risk factors are important predictors of AD as well as of vascular dementia. Severe atherosclerosis carries a threefold risk of developing AD, and the Syst-Eur study showed that treating systolic hypertension halves the likelihood of developing dementia. AD is a distinct entity rather than a mere exaggeration of normal ageing, and prevention strategies

17

pathological decrements in SUMMARY 2 Normal vs physiological performance in old age Normal age-related functional decline

pathological process

Impaired sight (presbyopia)

Cataracts, macular degeneration

Impaired hearing (presbyacusis)

Effects of ototoxic drugs, diuretics or vascular disorders

Bone loss

'Accelerated' osteoporosis

Decline in intellectual function

White matter lesions due to hypertension; Alzheimer's disease

Impaired glucose tolerance

Diabetes

Minor gait slowing and balance impairment (increased sway)

Parkinson's, Alzheimer's or cerebrovascular disease; dementia; vestibular lesions; cervical spondylosis; visual problems

Altered sleep pattern

Insomnia due to nocturia (due to detrusor instability); pain, depression, etc.

need to concentrate on 'cognitive impairment, not dementia' in people aged 65 and over. This has twice the prevalence of all types of dementia combined and is associated with functional disability and the need for institutional care, yet little is known of its natural history.

IMPAIRMENT OF BIOCHEMICAL HOMEOSTASIS Most normal old people can, under resting conditions, maintain their internal environment at levels similar to those in the young. When stressed, however, it is apparent that homeostatic mechanisms in the elderly are impaired. Blood glucose control and osmoregulation are two good examples of this.

The control of blood glucose The average fasting blood glucose level increases only slightly with advancing age. When a glucose load is imposed, however, the rise in maximum blood glucose levels is clearly greater, and the return to resting levels slower, in old people (Fig. 7.7). The upward drift of blood glucose with age produces diagnostic difficulties. If criteria used to diagnose diabetes in young populations were applied to the elderly, there would be a considerable increase in the number of those diagnosed as diabetic from the glucose tolerance test. Thus, elderly patients with an impaired glucose tolerance test (i.e. a 2-hour post-glucose blood sugar of 7-llmmol/L) should probably not be regarded as diabetic if they are asymptomatic, but they do carry a

Minutes FIG. 7.7 Mean blood sugar levels after a 50 g oral glucose load in a cross-sectional population study Known diabetics have been excluded.

higher risk of developing frank diabetes. Recent American Diabetic Association (ADA) criteria propose the use of fasting blood sugar, with 6.1-6.9 mmol/L indicating impairment and 7.0 mmol/L as the new diagnostic cut-off (instead of the WHO fasting criterion of 7.8 mmol/L). The fasting ADA criteria are less predictive than the WHO criteria of cardiovascular risk in the elderly, and for older adults a fasting glucose cut-off of 7.8mmol/L is preferable.

Osmoregulation On average, young adults maintain their blood osmolality in the range 280-295 mmol/kg. Although resting plasma osmolality is unchanged in the elderly, the osmoregulatory response to 24-hour water deprivation is impaired, even in good health. Despite developing higher plasma osmolality and sodium concentrations in response to dehydration, elderly subjects do not respond with appropriate thirst and water intake. They do not drink sufficient water to replenish their body water deficit and return plasma sodium levels to normal. Not only is there reduced thirst and water intake following dehydration, but the old person's kidneys are less able to retain water, due to reduced renal responsiveness to arginine vasopressin (AVP). These changes can predispose even healthy old people to fluid and electrolyte disturbances when there is excessive salt and water loss. They are also at risk of overhydration and hyponatraemia because they are unable to excrete excess water, mainly because of a decline in glomerular filtration rate with age. Particular care is needed with intravenous infusion to avoid overhydration and hyponatraemia. Postoperative patients are at high risk, especially if they have been taking a thiazide diuretic and as a result have mild hyponatraemia. Surgical stress causes inappropriate ADH secretion, which increases the hypona-

177

traemia. Sympathetic blockade resulting from spinal anaesthesia lowers the blood pressure, and if in an attempt to counteract this large volumes of i.v. 5% dextrose/saline are given, serious hyponatraemia ensues, which carries a 20% risk of brain damage or death. Hypotonic fluids must not be given to elderly postoperative patients.

FURTHER READING ON SPECIAL FEATURES OF DISEASE IN OLD AGE Mosekilde L 1998 Aging of bone. Rev Clin Gerontol 8:281-296. O'Neill P A 1997 Aging homeostasis. Rev Clin Gerontol 7:199-211. Sorlie P, Gordan T, Kannel W B 1980 Body build and mortality: the Framingham Study. JAMA 243:1828-1831. Staessen J A, Fagard R,Tuijs L et al 1997 for the Systolic Hypertension in Europe (Syst-Eur) trial investigators. Randomised double-blind comparison of placebo and active treatment for older patients with isolated systolic hypertension. Lancet 350:757-764.

AUTONOMIC DYSFUNCTION This results from a number of factors: a reduction in nerve density and neurotransmitter concentrations in the autonomic nerves; reduced responsiveness of effector organs, such as blood vessels and stretch receptors in arterial walls; and a decline in cellular p-adrenergic sensitivity. Dysfunction may follow central lesions such as cerebrovascular disease, Parkinson's disease, Shy-Drager syndrome, Wernicke's encephalopathy (due to thiamine deficiency associated with alcoholism) and Alzheimer's disease. Peripheral autonomic neuropathy may occur in diabetes mellitus, chronic alcoholism and malignancy. Autonomic dysfunction may also be caused by drugs such as phenothiazines, tricyclic antidepressants and haloperidol. The important clinical consequences are orthostatic hypotension (Table 7.2), bowel and bladder disturbances, impotence and impaired thermoregulation.

AGEING AND THERMOREGULATION

178

Under normal circumstances the body core temperature is about 37°C with, in temperate climates, a skin temperature of 33°C, giving a core-skin temperature gradient of 4°C. Thermoregulatory responses are impaired in old age due to disordered autonomic function. This usually passes unnoticed, but in times of environmental stress the old person is less able to maintain a constant deep body temperature. Surveys of old people living at home in winter show that only 0.5% had frank hypothermia (deep body temperature less than 35°C) but that 10% had a deep body temperature below 35.5°C. The latter had impaired

TABLE 7.2 Orthostatic hypotension • Fall in systolic blood pressure of at least 20mmHg or a fall of 10 mmHg or more in diastolic pressure on standing • Prevalence in elderly between 10 and 24% • Age-related aetiology Impaired baroreceptor response Failure of cerebral autoregulation Increased arterial and arteriolar rigidity Autonomic dysfunction • Cardiovascular causes Low cardiac output (e.g. heart failure) Restricted cardiac output (e.g. aortic stenosis) Absolute hypovolaemia (e.g. blood loss) Relative hypovolaemia (e.g. venous pooling) • Drug causes Centrally acting (e.g. L-dopa, MAOIs, sedatives) Vasodilators (e.g. nitrates, alcohol) Diuretics, calcium channel blockers • Symptoms Typically these consist of faintness, giddiness and blurring of vision, and are precipitated by standing and exacerbated by exertion or a heavy meal. However, if there is a lesion of the parasympathetic system these symptoms may be absent, so that despite a fall in blood pressure on standing there is no warning giddiness. The patient remains upright and experiences symptoms such as profound weakness, falls, confusion and pallor • Management Raise head of bed 15-20°, leg and abdominal compression, 9afludrocortisone 0.1 mg daily initially, increasing by 0.4mg each week. Other drug treatments include mididrone, dihydroergotamine, indomethacin, domperidone (in Parkinson's disease) and octreotide (for postprandial hypotension)

thermoregulation, as shown by a core-skin temperature gradient only half that of normal individuals. The elderly may also suffer impairment of thermal perception. Young persons can detect temperature differences as small as 0.8°C, but old people only perceive differences of 2-5°C. Furthermore, the elderly shiver less efficiently in response to cooling, with little more than half the metabolic heat production seen in young subjects. Old people are thus not only less able to maintain their deep body temperature when subjected to cold stress, but may also be less aware of cold conditions and therefore fail to take the appropriate steps. This may explain why some old people can tolerate cold living conditions without discomfort. Although most British homes now have central heating, hypothermic patients are less likely than controls to have the heating on at the time of collapse. Typically, hypothermia occurring indoors results from a collapse due to illness, when the patient is alone, lightly clothed and not in bed. These conditions cause progressive core cooling. About 80% of the variation in mortality rates throughout the year is associated with changes in temperature

a

FIG. 7.8 Observed and expected deaths in the UK during the cold winter of 1984/5 Note the rise in death rate during the winter months and the excess deaths during the two exceptionally cold periods in early 1985.

FIG. 7.9 Relation of obesity to mortality found in the Framingham study The obesity index is the ratio of the subject's weight to a reference weight obtained from height and frame charts compiled by large insurance studies. Note that the lowest mortality occurs at an obesity index rather above 1.0.

FURTHER READING ON AUTONOMIC DYSFUNCTION (Fig. 7.8). For every degree change in the average winter temperature, the number of annual winter deaths rises or falls by about 8000. Thus, in the UK in the mild winter of 1983/4, about 30000 more elderly people died than during the summer months (a 14% increase). In the cold winter of 1984/5 this figure rose to 46000 (an increase of 20%). However, the majority of these 'excess' deaths are not directly due to hypothermia, but to coronary and cerebral thrombosis resulting from haemoconcentration and hypertension after cold exposure. Over the last 20 years there has been a decline in this excess winter mortality. About half of this decline is the result of non-seasonal factors, such as improved general medical care. The rest is due to improvements in home heating and greater car ownership, which reduces outdoor exposure to cold. The diagnosis and treatment of hypothermia are discussed in Chapter 2, p. 46. The elderly are also less able to cope with excessive heat, and the mortality rate during heatwaves rises with increasing age. In Chicago during the benign summers of 1992, 1993 and 1994 the heat-related mortality was 1, 3 and 9, respectively, but during the exceptionally hot summer of 1995 it rose to 515, with 162 deaths on a single day in July. It has been estimated that on the hottest day of a heatwave there are six extra deaths per 100000 population. There is much unrecognized heat illness and large variations in the mortality rates, as there are no international or even national agreements on the definition of heat-related death. Various criteria have been introduced, such as a core body temperature above 40.6°C at the time of death. Most models of global warming predict an increase in the frequency and intensity of heatwaves. The impact is likely to be greatest in mid-latitude cities with infrequent but extreme heatwaves, so that in cities such as Shanghai and New York there will be several thousand extra heat-related deaths annually by the middle of the 21st century.

Collin K J 1997 Aging, disease and the autonomic nervous system. Rev Clin Gerontol 7:119-126.

NUTRITION AND AGEING At retirement age in developed countries the average person is more likely to be obese than undernourished, but subsequently body weight and nutrient intake fall and, for the very elderly, the main concern is undernutrition. Moderate overweight is not associated with a high mortality in old age (Fig. 7.9), but low body weight is associated with a higher mortality in both fit old people and those admitted to hospital. Nutritional surveys have shown a marked decrease in energy requirements and nutrient intake with increasing age. However, this is due mainly to physical illness and reduced physical activity, which reduce appetite and energy expenditure. With increasing age there is a gradual decline in metabolism at rest, but this is relatively unimportant. For healthy active persons over 60 years of age the energy and nutrient requirements are similar to those of a 30-year-old. Longitudinal studies have shown that nutrient intakes are maintained in old people who remain in good health. Housebound elderly women consume 15% less carbohydrate and 46% less vitamin C than do active age-matched controls.

NUTRITIONAL DEFICIENCIES A survey of old people living at home found that 3% were suffering from malnutrition, including protein-calorie malnutrition, iron deficiency and specific vitamin deficiencies.

179

supplement of 800 IU vitamin D and calcium reduces fracture rates and should be widely prescribed to at-risk patients. 1

TABLE 7.3 Risk factors for malnutrition Unexpected weight change of more than 3 kg Physical disability Lack of sunlight Gastrectomy Mental confusion Chewing or swallowing difficulties Fewer than eight main meals, hot or cold, in a week Absence of fruit or vegetables in diet Food wastage Alcoholism Poverty (receiving financial support from state) Depression or loneliness

BALANCE AND FALLS

In most cases this was the result of some underlying medical problem and was only rarely due primarily to economic or social factors (see Table 7.3). Social factors are, however, likely to be important in cases of subclinical malnutrition. For example, 31% of those living alone had leukocyte ascorbic acid levels below 7.0ug/108wbc, compared to 8% of those living with a spouse. Scurvy remains rare even in the most isolated patients. Almost 10% of acute hospital admissions are seriously malnourished and this is associated with a slower rate of recovery, longer hospital stay and increased treatment costs. Patients should be weighed on admission and questioned about their diet. Megaloblastic anaemia due to a folate-poor diet is rare in old people in the UK. However, low serum folate levels are found in about 8% of elderly hospital patients, and often in patients with physical or mental disorders that interfere with shopping and cooking. Osteomalacia may contribute significantly to the skeletal rarefaction seen in old age (although this is more commonly due to osteoporosis). There are many causes of osteomalacia, but two of the most important are reduced vitamin D intake and low exposure to sunlight - two factors often present together in the elderly. More than half of elderly admissions to a general medical ward have 25-OHD concentrations below 37.5nmol/L and nearly a quarter of these have a level below 20nmol/L indicating severe deficiency. The current UK recommended intake of vitamin D for people aged 65 years and over is 400IU (10 ug) daily, and this is not easily achieved without fortified foods or adequate levels of sunlight. A daily

180

1

MCQ 7.2

0 Case 7.1 3

4

Fig. 7.3

0 Fig. 7.4 6

0 Fig. 7.5

Fig. 7.2 MCQ 7.3

During normal quiet standing balance is maintained when the centre of gravity is kept within the support base provided by the feet. Maintenance of this upright position is associated with body sway, mainly in the anteriorposterior (AP) direction, and this sway may be measured either as degrees of angular movement or, using force platforms, as changes in the pressure over the soles of the feet. Threats to balance are detected principally by vision, lower limb cutaneous sensation and proprioception, and to a lesser extent by the vestibular system. Following a pertubation sensory information is centrally processed and corrective responses, such as muscle stiffening and cocontraction, or protective responses, where one or more steps are taken or an arm is thrown out to catch hold of an external support, are triggered. Disease in any part of this system will impair balance. Loss of a single sensory input is rarely completely disabling because of good overall reserve, but closing the eyes always causes an increase in postural sway and it appears that peripheral vision in particular is necessary for normal balance.

CHANGES IN BALANCE IN OLD AGE Both AP sway velocity and area increase in normal elderly subjects (Fig. 7.10), (i.e. those who report that their balance is normal and who are functionally independent). Further increases in AP sway correlate with spontaneous falls, but a better predictor of falls is increased mediolateral sway. Age-related changes occur in all parts of the postural control system. There is reduced visual acuity, contrast sensitivity, depth perception and dark adaptation. There is decreased cutaneous sensitivity in the feet and proprioceptor loss in the ankles and apophyseal joints in the cervical spine. In the vestibular system there is loss of labyrinthine hair cells and vestibular ganglion cells. Centrally there is an age-related loss of neurons, resulting in slowing of information processing and depletion of neurotransmitters such as dopamine in the basal ganglia. The older person has to pay much more attention to their balance, and loss of concentration contributes significantly to the risk of falling. On the motor side there is an agerelated reduction in muscle strength due to decreases in the size and number of muscle fibres and motor neurons. Loss of strength of the ankle dorsiflexors and knee extensors (i.e. quadriceps) is particularly associated with an increased risk ofgn f There has been a long-running debate on what constitutes normal gait in old age and the cause of senile gait

TABLE 7.4 Risk factors for falls

a

More than one previous fall Lower limb disability (including muscle weakness) © Impaired balance or gait Use of sedative drugs Use of more than four drugs Cognitive impairment (mini mental state examination score <26/30) Impaired activities of daily living 0 Visual deficit

FIG. 7.10 Postural sway (measured as movement in the anterior-posterior plane with the subject standing) increases with age At all ages, women sway significantly more than men. 1 unit = 3.5° of angular movement.

disorder. There are undoubtedly some elderly people who maintain a normal gait even in extreme old age. However, gait laboratory studies show subtle changes even in carefully screened healthy elderly subjects. Gait slowing is due to a decreased stride length and cadence (steps per minute) remains unaltered. Stance time and double support time increase and there is a less vigorous push-off. Whether this represents an early degeneration of the balance control system or an adaptation to make the gait safer is unclear. However, it is now apparent that the senile gait disorder, which is characterized by caution and shorter and more frequent strides and is said to occur in about 20% of the elderly, is not due to age but to underlying neurodegenerative syndromes and subclinical cerebrovascular disease. These patients have a twofold increased risk of cardiovascular death, compared to age-matched subjects with a normal gait. 4 FALLS Although falls are the leading cause of death from injury in people over the age of 75, serious injury only occurs in about 5 % of falls in community-living elderly, and a much more common consequence is loss of confidence, a reluctance to go out of doors and reduced quality of life. Thirty per cent of people over the age of 65 give a history of having had a fall in the previous year, and 50% of fallers do so repeatedly. Incidence rates in nursing homes and hospitals are nearly three times the community rate and the likelihood of an injury is more than doubled. Falls increase with age and fall rates in women are twice those in men. The more risk factors a person has (Table 7.4) the more likely they are to fall. Over 1 year 65-100% of people with three risk factors will fall, compared to 8-12% for those with none. The typical frequent faller usually has multiple defects affecting the postural control system.

Clinical or subclinical disease is a far more important cause of falls than environmental factors such as trailing wires and loose rugs. It is therefore no surprise to find that falls, incontinence and functional dependence all share the same risk factors. Increased fall rates in women probably reflect greater muscle weakness and possibly lower levels of fitness: for example, over 50% of women over the age of 55 are unable to walk at 3 miles per hour on the level, and the average healthy 80-year-old woman has to make a maximum quadriceps contraction to rise from a low armless chair or the lavatory. 7

HISTORY AND EXAMINATION It is important to spend time with the patient (and ideally a reliable witness) getting a precise account of the circumstances surrounding the fall. Was it a simple trip or accident, which carries a relatively good prognosis? Or was it a more ominous spontaneous fall (' I don't know what happened, I just went down') (Table 7.5). Some may say that they felt dizzy (particularly if prompted by the doctor). However, only very rarely does the patient experience true vertigo (a sensation of the surroundings or the inside of the head spinning round). They do not so much feel dizzy as unsteady when standing or walking, and are frightened that they are going to fall over. Some of these patients suffer drop attacks. These are falls which occur without warning: the patient suddenly crumples to the ground and there are no neurological sequelae. As with other spontaneous falls their aetiology is multifactorial, although sometimes a specific cause will be found, such as reflex-mediated syncope, normal-pressure hydrocephalus or Meniere's disease. Specific causes of falls such as these are less common than widespread postural defects (e.g. cerebrovascular disease and a diabetic peripheral neuropathy), which result either in spontaneous falls or a fall following a trivial pertubation because of slowing or even loss of the normal corrective responses. These patients are usually older and more frail, their gait and balance are noticeably impaired and their general health poor. Their mortality is about five times that of non-fallers.

181

TABLE 7.5 Causes of spontaneous falling

TABLE 7.6 Causes of syncope

Patient usually impaired in more than one of the systems concerned with balance control:

Pathophysiology of syncope is complex in older patients. Orthostatic hypotension, carotid sinus hypersensitivity and vasovagal syncope overlap in 30% of elderly with syncope.

• • • • • •

Vision, e.g. impaired depth perception Impaired cutaneous sensation (e.g. peripheral neuropathy) Impaired proprioception, e.g. cervical spondylosis, peripheral neuropathy Vestibular lesions, e.g. Meniere's disease Impaired central coordination, e.g. cerebrovascular disease Unstable support base, e.g. turning quickly, walking downstairs or arising from a chair

If balance is already impaired, falls are precipitated by: • Acute illness, e.g. chest infection, worsening of heart failure • Drugs, e.g. tranquillizers • Occult ill health, e.g. anaemia, hyponatraemia

Reflex mediated (neurovascular instability) Vasovagal Carotid sinus hypersensitivity Orthostatic hypotension Situational syncope (cough, micturition, etc) Cardiac Arrhythmias Obstruction (aortic stenosis, pericardial effusion etc) Neurological Vertebrobasilar insufficiency Subclavian steel syndrome

Rare

Points to note in the history are: • Circumstances of the fall; • If the fall occurred soon after standing up (raising the possibility of orthostatic hypotension) • Loss of consciousness (can the patient remember hitting the floor?). There are only two common causes syncope (Table 7.6) and epilepsy; • Syncope accounts for up to 6% of medical admissions and 20% of elderly patients presenting to A & E with unexplained falls, such as drop attacks. Patients will sometimes deny loss of consciousness, despite unequivocal evidence that it occurred; • Vasovagal syncope may be preceded by the classic symptoms of lightheadedness, sweating and nausea, but in the 'malignant' variety there is no warning, the patient falls unexpectedly, the attack is prolonged with slow recovery and injury, and convulsions and incontinence are common; • Palpitations before the fall, a sense of dizziness or faintness and dimming of vision are important pointers to syncope; • Loss of consciousness, dizziness and faintness are not features of transient ischaemic attacks. These depend for their diagnosis on focal neurological signs (e.g. motor or sensory loss, dysphasia); • The patient's drugs will need to be checked. Sedatives or hypnotics frequently slow the patient down and may precipitate a fall. B-Blockers, certain calcium channel blockers and diuretics may cause orthostatic hypotension or vasovagal syncope; • Acute or chronic medical problems

1

182

Fig. 7.6

SUMMARY 3 Typical features in patients with recurrent falls • • • • •

Have several risk factors, especially lower limb muscle weakness Significantly impaired gait and balance Difficulty in accounting for how they fell Show fear of falling Generally have multiple defects both in the sensory systems concerned with balance control (vision, proprioceptors and the vestibular system) and in central coordination centres

• Mental state should be assessed if there is a suspicion of dementia or depression; • A history of tinnitus, deafness or vertigo should be sought. If present, this may indicate a vestibular lesion, either peripheral (e.g. Meniere's disease) or central (vascular disease in the brainstem or cerebellum). • A sudden onset of severe vertigo suggests a posterior circulation stroke. Episodic vertigo lasting less than 1 minute and provoked by rapid positional change is probably benign positional vertigo, which is confirmed by a positive Hallpike test (for description see p. 183); In the general examination: • The lying and standing blood pressure should be measured (a drop in systolic pressure of more than 20mmHg indicates postural hypotension). • Carefully examine the heart. • The range of neck movements should be assessed. • Watch the patient rise from a chair, walk several paces and return. • Normal old people should not have a positive Romberg test (i.e. closing the eyes when standing should not markedly impair balance). • Check for impaired righting reflexes by tapping the

RECENT ADVANCES IN THE PREVENTION OF FALLS 1. Intervention studies have demonstrated significant reductions in falls. Because falls' are multifactorial broad-based interventions are best. Programmes such as the following reduce the risk of falling by 30%: • Review and adjust drugs • Balance and gait training • Muscle strengthening exercises • Practise transfer skills from bed/chair/toilet • Make home safer. 2. Assessment and appropriate treatment and referral of patients attending an A & E department lowers the risk of recurrent falls, reduces fracture rate by 50% and decreases the odds of hospital admission. 3. The Frailty and Injuries: Cooperative Studies of Intervention Techniques (FICSIT) programme demonstrated that exercise alone could produce a 25% reduction in falls. In particular, Tai Chi appeared beneficial and led to a 47% reduction in falls incidence in elderly women. A home visit by a physiotherapist and the prescription of 30 minutes of strength and balance retraining exercises and an outdoor walk three times a week reduces fall rate and remains effective for at least 2 years. 4. Hip protector pads reduce the risk of hip fracture. They are indicated for frail at-risk patients who are sufficiently well motivated to wear them. 5. Patients with parkinsonism can be taught how to manoeuvre more safely in the home and to walk with a normal stride length using markers on the floor or attentional strategies.

patient lightly on the sternum or giving the shoulders a tug from behind. Be ready to catch them if they fall. • In severe cases there will be manifest anxiety and fear of falling, with the patient reluctant to move without clutching on to furniture or onlookers for support. • Check muscle strength (especially quadriceps), lower limb joints, vision, neurological function (especially lower limb peripheral nerves). In addition to routine investigations: • A 24-hour ambulatory ECG should be obtained for patients suspected of having episodes of cardiac syncope due to an arrhythmia. • Patients with a normal 24-hour tape, but who have symptoms suggesting vasovagal syncope or carotid sinus syncope, require head-up tilt testing and carotid sinus massage (a positive response produces asystole exceeding 3 seconds, a decrease in systolic blood pressure exceeding 50mmHg, or a combination of the two). 1 • An EEG is occasionally helpful, although the diagnosis of epilepsy is essentially clinical (see Ch. 24 p. 1321).

Early morning falls in patients taking a long-acting hypoglycaemic drug (e.g. chlorpropamide) raise the possibility of hypoglycaemia. Patients suspected of having a vestibular lesion can have this accurately localized only by specialist examinations, which include electronystagmography and caloric tests. The Hallpike manoeuvre can be carried out in the clinic and tests for the presence of positional vertigo. The nystagmus of benign positional vertigo occurs after a latent period of 2-20 seconds when the head is reclined and turned to the affected side. To carry out the test the patient is seated at the top end of a flat examination couch and warned that they will experience vertigo. The examiner holds the patient's head firmly and turned 45° to the left or the right; the patient is then carried rapidly backwards until the head is dependent. The patient is asked to keep his eyes open all the time and look at the examiner's nose. The nystagmus of benign positional vertigo is rotatory (torsional) and is directed towards the affected ear when it is undermost (i.e. when the left ear is affected the nystagmus is to the left when the head is hanging to the left.) The nystagmus lasts for up to a minute and coincides with a sense of vertigo, which may be severe. When the patient sits up the nystagmus may reappear beating to the opposite side. With repeat testing the symptoms and signs adapt and fatigue. Central types of positional nystagmus may also be recognized by this test. In contrast to benign positional vertigo the central type of nystagmus appears after no latent period, it persists for longer than a minute, it does not fatigue, vertigo is usually absent, and the direction is variable. The patient's vision, both with and without spectacles, should be noted (depth perception, low contrast, etc.). Low-contrast visual acuity and contrast sensitivity (the detection of large visual stimuli under low-contrast conditions) are better predictors of falls than impaired high-contrast visual acuity.

a

MANAGEMENT Many patients with recurrent falls have multiple postural defects and their balance is precarious. All require a multidisciplinary assessment. Sometimes a minor illness such as a chest infection or a newly prescribed sedative drug can simply prove too much for a frail patient. Identification of precipitating factors enables the patient to be appropriately advised. Culprit drugs should be withdrawn in patients with syncope. Vasovagal syncope may be treated with B-blockers, SSRIs or pacing. Cardioinhibitory carotid sinus syndrome usually needs a permanent pacemaker. Cervical spondylosis should be treated with neck exercises, not a cervical collar, which often makes the patient feel more unsteady. Faulty spectacles should be replaced and everything possible done to improve the patient's vision, including better illumination in the home.

183

CASE STUDY 7.1

FALLS AND 'FY T

An 83-year-old woman, who looks after her 90-year-old husband, gives a 2-year history of increasing difficulty with walking. When standing or walking she feels unsteady, as if she might fall, but there is no sense of vertigo. She goes out for a walk every day with her daughter, who lives nearby, but lacks confidence and does not go out of doors on her own. She has had several spontaneous falls, sometimes going over backwards. She says that her walking has become difficult and that her legs feel 'as if they are in concrete', but they are not painful. Sometimes when she starts walking her legs will not move - 'they seem to be nailed to the floor' - but after a few seconds she is able to get going. She finds a similar problem when she turns: her legs lag behind her upper body, with the feet seemingly stuck to the floor, and she has fallen over twice as a result. She also describes 'funny turns'. These occur about once a week without warning. She has an unpleasant sick feeling in her stomach that rises rapidly to her head. Her daughter says that during these episodes her mother looks vacant. Her eyes are open, but she does not respond to questioning and sometimes makes a funny grunting noise. The episodes last just a minute or two, but she often feels unwell for some hours afterwards. The daughter has also noticed that her mother's

memory has deteriorated. On direct questioning the patient admits to urinary urge incontinence and daytime frequency, sometimes having to pass urine every hour. At night she gets up two or three times. There is no dysuria.

Questions 1, What gait abnormality has the patient described? 2. What is the cause of her 'funny turns'? 3, What is the likeliest explanation for her urinary incontinence?

The patient's description of her feet sticking to the ground when she tries to walk or turn is known as freezing. It is a feature (in order of frequency) of progressive supranuclear palsy, arteriosclerotic parkinsonism, normal pressure hydrocephalus and idiopathic Parkinson's disease. It is rare in multisystem atrophy and very rare in drug-induced parkinsonism. The commonest type of freezing is start hesitation (sometimes called gait ignition failure), but it can also be observed in speech, the eyelids, in handwriting and in tasks such as shaving. Her 'funny turns' are typical of complex partial fits (elderly patients

Peripheral vestibular lesions may be improved with cinnarizine (15 mg t.d.s.) or betahistine (8mg t.d.s.), but these drugs may have troublesome sedative effects. Central vestibular lesions are rarely helped by drugs and these should therefore be avoided if this diagnosis is definitely made. Phenothiazines should not be prescribed for vague complaints of dizziness, because of the risk of drug induced parkinsonism. Benign positional vertigo can be treated by a physiotherapist with the Epley manoeuvre or Cooksey1

184

Case 7.2

0 Figs 7.8-7.9

0 Fig. 7.7

do not have petit mal). There is a steep rise in the incidence of fits in old age and the commonest cause is cerebrovascular disease. She has described urge incontinence and frequency, and the likeliest explanation for this is detrusor instability. Examination of the patient The patient scored 24/30 on the minimental state examination, confirming mild cognitive impairment. She was unable to stand unsupported with her feet together, even with her eyes open. When asked to walk there was start hesitation before she got into her stride. She walked with a broadbased, short-paced gait. Her posture was upright and she swung both arms normally. There was no tremor and tone in her arms was normal. In her legs tone was increased and there was also bradykinesia, shown by marked slowness and frequent hesitations when asked to rapidly tap her heel on the ground. Reflexes were brisk in her legs, but plantars were flexor. The rest of the examination, including PV and PR, was normal. Is this idiopathic Parkinson's disease? Probably not. Despite having three of the four cardinal signs (rigidity, bradykinesia and gait disturbance) the signs are all confined to the legs and this strongly suggests arteriosclerotic parkinsonism

Cawthorne exercises. The Epley manoeuvre is based on the premise that debris moving in the endolymph of the posterior semicircular canal stimulates the hair cells and causes vertigo. The manoeuvre aims to direct the debris out of the canal into the utricle. About 80% of patients experience immediate relief of vertigo and nystagmus after a single session. Cooksey-Cawthorne exercises are suitable for any patient with chronic dizziness. They are performed twice a day and help the patient to develop tolerance to the vertigo and increase confidence. The patient's chair should be at the right height (the seat level with the top of the knee) so that getting out of it is not difficult. A walking frame may initially help both mobility and confidence, but should be removed once the

7

CASE STUDY 7.1 CONTINUED (sometimes known as lower-half parkinsonism). Further support for this diagnosis and against idiopathic Parkinson's disease is the symmetry of the findings, the initial presentation with a gait abnormality, normal arm swing and upright posture, the early appearance of cognitive impairment and the absence of tremor. Management Patients with arteriosclerotic parkinsonism (unlike idiopathic Parkinson's disease) only rarely improve with levodopa. Treatment was started with co-careldopa 12.5/50 one tablet b.d., and gradually increased until she was taking the equivalent of 400 mg levodopa a day. The drug made her feel sick and

shaky and there was no improvement in her walking, and so it was stopped. Gait, mobility and confidence improved with physiotherapy. After an MSU was checked to exclude a urinary tract infection she was started on bladder retraining; after 6 weeks she was no longer incontinent and was managing to hold for 2-3 hours between voids during the day. A CT brain scan confirmed small vessel disease. (Case Fig. 7.1.1) She was checked for vascular risk factors and started on 75 mg aspirin daily. Although hypertension is associated with small vessel disease there is no evidence that lowering blood pressure or taking aspirin slows the rate of progression of the disease.

patient's walking is better. Balance and muscle strengthening exercises and measures to improve confidence and wellbeing are of proven benefit. If the falls have been such as to warrant hospital admission, then follow-up for a few weeks in a day hospital is helpful. In case of another fall, the patient should be taught how to get up off the floor 3 and advised of the various alarm systems that can be installed in the home. An occupational therapist should also visit the home and correct obvious environmental hazards and teach the patient safe transfers. Preventing fractures is important. Patients should be assessed for osteoporosis and treated accordingly (see p. 175). FURTHER READING ON BALANCE AND FALLS AGS/BGS/AAOS Panel on Falls Prevention 2001 Guidelines for the prevention of falls in older patients. J Am Geriatr Soc 49:664-672. Campbell A J, Robertson M C, Gardner M M 1999 Falls prevention over 2 years: a randomised controlled trial in women 80 years and older. Age Ageing 28:513-518. Close J, Ellis M, Hooper R et al 1999 Prevention of falls in the elderly trial (PROFET): a randomised controlled trial. Lancet 353:93-97. Davies A J, Kenny R A 1999 Syncope in older patients. Rev Clin Gerontol 9:117-126. Lempert T, Gresty M A, Bronstein A M 1995 Benign positional vertigo: recognition and treatment. BMJ 311:489-491. Maki B E, Mcllroy W E 1996 Postural control in the older adult. Clin Geriatr Med 12:635-658. Nutt J G, Marden C D, Thompson P D 1993 Human walking and higher level gait disorders, particularly in the elderly. Neurology 43:268-279.

CASE FIG. 7.1.1 CT of the head showing leukaraiosis. This is a radiological diagnosis indicating periventricular white matter hypodensity. It is commonly caused by small vessel disease and is associated with gait disturbance, falls and dementia it may also be seen in normal elderly people.

Province M, Hadley E, Hornbrook M et al 1995 For the FICSIT Group. The effects of exercise on falls in elderly patients. A preplanned meta analysis of the FICSIT trials. J Am Med Assoc 273:1341-1347

URINARY INCONTINENCE Although this is more common in the elderly, the causes and management are similar to those in the young (see Ch. 20). The major difference is that elderly patients are more likely to suffer in silence or even to deny the problem. Doctors must be proactive and should routinely ask at-risk patients whether they have 'any bladder problems'. After a pertinent general assessment (diuretics, mental state, mobility, etc.), examination to exclude prolapse, pelvic mass and prostatic carcinoma and urine tests for infection, blood, glucose and protein, the patient is referred to the continence nurse specialist for initial management. Age has little effect on prognosis and the majority of patients can expect to become dry following treatment.

DETRUSOR INSTABILITY The majority of patients are incontinent because of idiopathic detrusor instability. The cortical centre for this willed control over the hindbrain-sacral reflex lies in the frontal premotor area. In old age these tracts are vulnerable to cerebrovascular disease and Alzheimer's disease. An early sign of detrusor instability in old age is nocturia,

185

TABLE 7,7 Principles of bladder retraining 1. 2. 3. 4. 5. 6.

Explanation of condition Establish baseline micturition pattern using chart Select appropriate time for patient to wait between urinations, e.g. 1-2 hours Provide patient with a suitable pad to wear in case of incontinence Gradually extend interval as patient improves Concomitant relaxation therapy for stressed or anxious patients

which may progress to overactive symptoms of urgency, urge incontinence and daytime frequency. Stress incontinence is usually due to urethral sphincter weakness, but similar stress leakage may be seen in patients with detrusor instability and is due to cough or sneeze raising intraabdominal pressure and triggering an unstable contraction. A cystometrogram is needed to distinguish between the two causes, and patients with genuine stress incontinence can expect good results from pelvic floor exercises and surgery. Whether the detrusor instability is idiopathic or due to cortical damage (e.g. from a stroke), the response to bladder retraining is good provided that the patient can comprehend the instructions and is well motivated. About 80% of patients will be dry after 6 weeks' treatment. Bladder retraining (see Table 7.7) depends on asking the patient to urinate only after a certain fixed time, the underlying assumption being that 'holding' in this way strengthens cortical inhibition over the sacral reflex. If patients are struggling with retraining they may benefit from the addition of tolterodine m/r 2mg o.d. or oxybutyrin m/r 5-30 mg o.d. The modified release formulation is less likely to cause dryness of the mouth. If despite these measures patients continue to wet the bed or have very troublesome nocturnal polyuria, they can be helped by desmopressin tablets 0.1-0.2 mg at bedtime. This reduces urinary output in patients with a high nocturnal diuresis (>0.9mL/min) and is ideal for occasional use, such as a stay in a hotel. Up to 3 weeks treatment is safe; over longer periods blood pressure and serum sodium must be monitored. 12

VOIDING DIFFICULTIES These increase with age and also occur in women. Typical symptoms, such as hesitancy, poor stream, straining to void and a feeling of incomplete bladder emptying, do not reliably distinguish bladder outlet obstruction from detrusor hypocontractility, which may be idiopathic or sec-

1

186

Case 7.3

2

ondary to, say, a cauda equina or peripheral nerve lesion. Neurological and rectal examination and pressure-flow studies are required for an accurate diagnosis. Hypocontractility causing a large residual urine volume carries the risk of recurrent urinary tract infections, and is managed initially by advising the patient how to empty the bladder completely and, if necessary, with prophylactic antibiotics. Long-term management is with intermittent self-catheterization. Patients with benign prostatic obstruction have a choice of treatments: watchful waiting if symptoms are causing little trouble, drugs if symptoms are troublesome and surgery is declined (a-blocker or a 5-a-reductase inhibitor), TURP, or alternatives such as laser prostatectomy or thermotherapy.

POSTOPERATIVE URINARY RETENTION This may happen to any elderly man following surgery. It is commonly assumed to be due to prostatic outflow obstruction, but is much more likely to result from a combination of postoperative pain, preoperative opiates, anticholinergic anaesthetic agents, a bladder overdistended with intravenous fluids, and lying immobilized in bed. If the patient does not respond to an oc-blocker, such as alfuzosin 2.5mg b.d. (beware hypotension and give the first dose at bedtime), then the bladder should be emptied with intermittent catheterization until normal detrusor activity returns.

CATHETERS A permanent indwelling catheter is the last resort for urinary incontinence, but is not necessarily a bad solution. Some patients are very relieved to be clean, dry and comfortable for the first time in years. The gauge should be 14 or 16 Ch and the retaining balloon should contain no more than 5 mL of water to reduce the risk of provoking unstable contractions, which will cause urine to leak out around the catheter. If leakage persists tolterodine or oxybutyrin b.d. should be tried; if this is unsuccessful the best solution is to change to a suprapubic catheter. Catheters always become infected within 2 weeks of insertion. This is rarely a problem if drainage is satisfactory and fluid intake is l-1.5L/day. Catheter specimens of urine should only be sent for culture if there is an unexplained deterioration in the patient's condition or symptoms of system infection. Bladder washouts and prophylactic antibiotics are ineffective (see Table 7.8).

FURTHER READING ON URINARY INCONTINENCE Fig. 7.10

Incontinence. Causes, management and provision of services. London: Royal College of Physicians, 1995.

TABLE 7.8 Good practice guidelines: Urinary incontinence

FAECAL LOADING

Summary of recommendations from Good Practice in Continence Services (Dept. of Health 2000) and US Department of Health and Human Services: Clinical Practice Guidelines: Urinary Incontinence in Adults (1996). The strength of evidence supporting each recommendation is based on the following criteria: A Scientific evidence from properly designed and implemented controlled trials providing statistical results that consistently support the guideline statement; B Supported by scientific evidence from properly designed and implemented clinical series; C Supported by expert opinion.

This is the commonest cause of faecal soiling. Rectal sensation is impaired, probably aggravated by chronic rectal distension, and the patient often leaks before experiencing a call to stool. Problems result not only from a large hard stool in the rectum, but also from a mass of soft stool, which is not readily recognized. The rectum and lower colon should be cleared either with repeated enemas, which are uncomfortable and time-consuming, or a kinder alternative such as oral polyethylene glycol (Movicol). To clear faecal impaction, eight sachets in a litre of water are taken over at least 6 hours on 3 consecutive days. Bran must be used with caution, as it may make the stool consistency too soft and can actually cause faecal incontinence.

In primary and secondary care • All patients require initial assessment, to include a review of symptoms, examination of abdomen and perineum (including pelvic floor contraction) to identify prolapse and excoriation, and rectal examination to exclude faecal impaction and urinalysis (grade B). • Detrusor instability should be treated with bladder training (grade A).and, if necessary, anticholinergic drugs (grade A). Symptoms of stress incontinence are reduced by pelvic floor muscle exercises (grade A). • There is no evidence that oestrogen supplementation improves incontinence in postmenopausal women (grade B). • Colposuspension and sling procedure are more effective than anterior colporrhaphy or needle suspension (grade B). • Appropriate continence aids such as pads need to be provided (grade C). • Existing medications (e.g. diuretics) need to be reviewed (grade C). • Refer immediately to a urologist patients with acute retention of urine, renal impairment, carcinoma of the prostate, haematuria or recurrent infection (grade C). • Refer immediately to a gynaecologist patients with a pelvic mass, vaginal bleeding or a history of previous gynaecological surgery (grade C). • If, after initial assessment and 6 weeks of bladder retraining or pelvic floor exercises patients are not improving, they should be referred for specialist review (grade C).

SLEEP AND INSOMNIA Insomnia is a common complaint in the elderly and there is great pressure on doctors to prescribe hypnotics. The elderly receive more hypnotics than any other age group. It is often assumed that hypnotics are both harmless and effective, but for the elderly neither assumption is true. Elderly patients taking hypnotics regularly rate the quality of their sleep as worse than those not on hypnotics, and more than one in 10 persons aged 70 or over taking nitrazepam (10 mg) at night have troublesome daytime sedation, with confusion unsteadiness and an increased risk of falling. Any patient complaining of insomnia - but particularly an elderly one - needs careful assessment and not just a prescription for a hypnotic.

TABLE 7.9 Causes of faecal incontinence

NORMAL SLEEP

• Secondary to colorectal disease, e.g. infective diarrhoea, carcinoma of the rectum • Idiopathic weakness of the anal sphincter and pelvic floor muscles • Faecal loading • Neurological causes, e.g. dementia, impaired consciousness and rectal instability

A person falling asleep 'descends' rapidly through progressively deeper stages of sleep. Slow-wave sleep - which occupies about 20% of the night in the young - is reduced to only 5% by 70 years of age, and stage four is almost completely absent (Fig. 7.11). The elderly wake more frequently and take longer to fall asleep, so that even for the normal old person sleep becomes increasingly fragmented. Rapid sleep onset is promoted by dilatation of the blood vessels of the hands and feet, which increases distal heat loss. A hot water bottle at the feet rapidly induces vasodilatation and the resulting heat loss is most effective when the ambient temperature is cool. Difficulty falling asleep in old age may reflect an inability to vasodilate owing to impaired autonomic responses. Thus, the doctor needs to decide whether an elderly patient's complaint of insomnia is an unrealistic expectation of what constitutes a good night's sleep. Less physical activity and daytime naps will further reduce the ability to sleep well at night.

FAECAL INCONTINENCE Faecal incontinence is distressing for both patient and carer and is a major reason for care breaking down in the home. The prevalence in the community is less than 1%, rising to 10% in residential homes and 25-30% in nursing homes or on long-stay wards. The causes are given in Table 7.9.

a

187

TABLE 7.10 Advice to patients with insomnia • • • • • •

Avoid daytime naps. Avoid caffeine or alcohol at bedtime (a hot milky drink is better). Ensure that the bed is warm and comfortable. Only go to bed when tired. Do not read or watch television in bed. If not asleep in 20 minutes get out of bed and sit in another room until feeling tired again. • Get up at the same time every morning.

A similar effect can be seen after hypnotics are withdrawn, when the patient complains of anxiety, tension and insomnia. If these withdrawal effects are not understood by the doctor and patient then a hypnotic may be needlessly represcribed.

CHOOSING A HYPNOTIC

FIG. 7.11 Typical hypnograms for young adults and elderly people The elderly wake more frequently and have less deep sleep.

SUMMARY 4 Causes of insomnia • • • • • •

Age-related changes in sleep pattern Nocturia (probably due to detrusor instability) Pain Anxiety and depression Drug induced, e.g. alcohol, withdrawal from hypnotics Specific problems, e.g. sleep apnoea, myoclonus

CAUSES OF INSOMNIA

188

Insomnia may be age related, as discussed above. Nocturia is another common reason for an old person waking up; the condition may be caused by heart failure, renal failure or diabetes, but is most likely due to detrusor instability (especially if associated with some urgency). Pain is also an important cause of insomnia, and the akinetic parkinsonian patient who cannot turn over in bed will find that sleep is disturbed. Anxiety and depression commonly cause sleep disturbances, and are best managed by psychiatric treatment directed at the primary condition. Insomnia may be drug induced. The most familiar example is alcohol which, if taken in the evening, will quickly induce sleep but because of its rapid metabolism results in rebound insomnia with awakening later the same night.

Hypnotics should only be used when insomnia is impairing the patient's daytime performance, and after advice on non-pharmacological measures (see Table 7.10). There should be a clear indication, such as a recent bereavement or impending major surgery, and the prescription should be limited to 7 or 10 days wherever possible. The drug should have a rapid onset of action, particularly if the patient's main complaint is difficulty in falling asleep. There should be no daytime sedation, although this may occur because of the accumulation of long-acting hypnotics and their active metabolites, or because of an age-related alteration in sensitivity to the drug.

Benzodiazepines Diazepam has a half-life in hours approximately equivalent to the person's age, but the constant clearance rate means that its average steady-state plasma level is unaffected by ageing. However, the half-life of its major active metabolite, desmethyldiazepam, is prolonged as a result of reduced plasma clearance. The dose necessary to produce sedation is therefore inversely correlated with age, owing partly to an increased sensitivity to the drug in old age and also to accumulation of the active metabolite. Diazepam (2 mg or 5 mg) may be a suitable hypnotic for anxious old people where a daytime anxiolytic effect is required. Daytime drowsiness, confusion and falls, however, may be a problem. Nitrazepam also produces more sedation in the elderly, probably as a result of increased receptor sensitivity (as the pharmacokinetics remain unaltered). The unwanted daytime sedation is largely dose dependent (Fig. 7.12); 2.5 mg is the maximum dose that should be given to an old person.

Morgan K 2001 Mental health factors in late life insomnia. Rev Clin Gerontol 11:71-81.

7

SEXUAL FUNCTION

FIG. 7.12 Relationship of daily dose (5.0-9.9mg/day and 10.0 or more mg/day) to the frequency of unwanted daytime CNS depression attributed to nitrazepam, with the population stratified according to age

Temazepam has a half-life of only 5-10 hours in young people but may be as long as 30 hours in the elderly. Accumulation therefore occurs and, after a week of continuous dosing at 20mg each night subjects show a significant slowing of reaction times. The effect is likely to be dose related; 10 mg at night rarely causes serious problems. Lormetazepam has a good pharmacokinetic profile for use in the elderly. Although the elimination half-life is prolonged to about 14 hours in elderly subjects, accumulation does not occur and there is no impairment of daytime performance with a nightly dose of either 0.5 mg or 1.0 mg.

Other hypnotics Zopiclone, a cyclopyrrolone, has a half-life of about 8 hours in the elderly; it does not accumulate over 7 days' treatment and psychomotor performance is unimpaired. Treatment should start with 3.75mg. Clomethiazole (chlormethiazole) is valuable in the elderly. Its pharmaco-kinetics are unaltered by age and its short plasma half-life of 3-4 hours prevents accumulation. It is metabolized by the liver, and should be used with caution in patients with cirrhosis of the liver or in alcoholics. One capsule (192mg of chlormethiazole base) may be sufficient but two capsules is still safe. Restless and confused patients are often best sedated with haloperidol (0.5-3mg nocte). Long-term use carries a risk of extrapyramidal side-effects and should be avoided in parkinsonian patients.

In western society sexuality among adolescents and young adults is openly studied and discussed. No such liberal attitude exists towards sex in old age and, unfortunately, the stereotype of the sexless old persists. Although sexual activity does decline with advancing age, much of this is due to ill health or the death of the marriage partner. Old age itself causes very little impairment of sexual enjoyment: 81% of married men and women aged 70-79 are sexually active and find it enjoyable. Overall, it would appear that continued sexual activity in old age depends on the frequency and enjoyment it afforded in middle age, the availability of a suitable partner and (particularly in men) continuing good health. The main physical changes in women result from a lack of oestrogen, which may cause vaginal and clitoral soreness, with reduced lubrication and painful intercourse. Length of orgasm may be reduced but many postmenopausal women report increased desire and arousability. Male fertility is usually maintained into extreme old age, although the volume of prostatic secretions reduces, it takes longer to achieve an erection, orgasm is briefer and the following refractory period is longer. Many doctors are reluctant to discuss sexual practice with their elderly patients and thereby lose the opportunity for helpful counselling. Vaginitis due to oestrogen deficiency responds to topical oestrogen cream and lack of lubrication is helped by over-the-counter products such as K-Y jelly or Senselle. Complaints of impotence in the male should prompt a thorough physical examination, including a blood sugar estimation. Sildenafil is generally safe and effective, but is contraindicated in patients taking nitrates. Many drugs interfere with sexual function and the elderly should be warned about this when they are prescribed. Propranolol and thiazide diuretics may produce impotence. All of the phenothiazines may cause impotence; thioridazine may also cause lack of ejaculation. Haloperidol is free of these problems. Sexuality - or the lack of it - in old age should not be seen in isolation. The need for affection, tenderness and caressing may be more important than intercourse, and counselling as to what is normal and appropriate may dispel fears of inadequacy or embarrassment.

ALTERED REACTIONS TO DISEASE FURTHER READING ON SLEEP AND INSOMNIA Espie C A 1993 Practical management of insomnia: behavioural and cognitive techniques. Br Med J 306:509-511.

In the elderly acute illness frequently presents in a non-specific way, such as with confusion, falls and incontinence.

189

TABLE 7.11 Differential diagnosis of confusion and dementia Differential diagnosis Acute confusional state

Dementia

Onset Clouding of consciousness

Abrupt Yes

Course

Fluctuates, lucid intervals during day. Worse at night Yes (may be intermittent)

Gradual, months or years No (but there may be drowsiness due to sedation) Stable over course of day

Disorientation and cognitive impairment Alertness Hallucinations Duration

Increased or impaired Yes (common) Hours or days

After treatment of any underlying cause, management of the confused patient relies mainly on the correct environment, and sedation should be kept to a minimum. The patient is best looked after in a side-ward so as to avoid disturbing other patients. A relative or nurse should stay with the patient at all times to reassure, comfort and explain what is happening. The light should be kept on in the room at night. Sedation is best avoided but may be needed if the patient is very restless, e.g. haloperidol 1-3 mg q.d.s. by mouth, or 2.5-5mg i.m. (NB the oral and i.m. doses are not equivalent).

Yes

Usually normal Yes (uncommon) Steadily progressive

N.B, None of the clinical features of the acute confusional state is reliable and a confident diagnosis depends on the history.

CONFUSIONAL STATES A toxic confusional state is the ageing brain's response to a number of insults: the stress of a physical illness, a variety of powerful drugs, and alterations in the old person's familiar environment. It tends to occur more commonly in those whose mental state is already impaired, for example by Alzheimer's disease, or who have communication difficulties because of defective hearing and vision. Typically, the onset is abrupt over hours or days. This and the clouding of consciousness are the main points of differentiation from dementia (see Table 7.11). Typically, the acutely confused patient is drowsy; it is difficult to get their attention and there are fluctuations in the level of consciousness. The patient's agitation and confusion are usually worse at night. About 15% of elderly patients admitted to acute medical wards are delirious. One-third die within a month; of the survivors, 80% will have recovered mentally within a month.

Investigations

190

Management

Proper assessment and diagnosis are essential, and require a careful history from a reliable witness. Physical illness, coexisting dementia, recent emotional stress, alcohol abuse, drugs (especially benzodiazepines, antidepressants, antiparkinsonian drugs, analgesics, NSAIDs and steroids), or head injury are important precipitating factors. In hospital, surgery is an important cause of confusional states. Physical examination should concentrate on identifying possible causes. The commonest are cardiac failure, Gram-negative septicaemia and pneumonia.

ABSENCE OF PAIN For reasons not clearly understood, pain due to illness or trauma, which would be expected in a younger individual, is often minimal or absent in the elderly. Thus, myocardial infarction is commonly pain free and the patient presents instead with dyspnoea or non-specific illness. Similarly, abdominal pathology such as gastric ulcers or mesenteric thrombosis may be silent; however, acute appendicitis often has a fulminating course in older patients, with the rapid development of gangrene and peritonitis. Even a fracture of the neck of femur may cause the patient only slight discomfort, evidenced by a disinclination to walk.

MULTIPLE PATHOLOGICAL PROCESSES The explanation of all symptoms, signs and abnormal investigations by one diagnosis is seldom possible in older patients, the usual pattern being one of multiple disorders. Thus, a momentary cardiac arrhythmia which may produce only a slight feeling of dizziness or palpitations in a young person may cause loss of consciousness or even a stroke with hemiplegia in an old patient. Similarly, a chest infection may precipitate heart failure because the cardiac reserve is already impaired by ischaemic heart disease. Common conditions such as obesity, osteoarthritis, diverticular disease, hypertension, diabetes, dementia and presbyacusis may all be present in a single individual, and the aim then is to decide which are causing important functional disability.

UNREPORTED ILLNESS The early identification of disease or disordered function allows treatment to be initiated or social support organized before things have gone too far. Studies in general practice show that an elderly person has, on average, at least three disabilities, and for every one known to the GP there is another that is not. Typically, cardiac, chest and neurological disabilities are known to the GP, but incontinence,

difficulty in walking, painful feet, dementia and depression often go unreported.

FURTHER READING ON ALTERED REACTION TO DISEASE Treloar A 1998 Delirium: prevalence, prognosis and management. Rev Clin Gerontol 8:241-249.

CLINICAL APPROACH When assessing an elderly patient it is generally more useful to think in terms of functional ability than to emphasize the underlying pathology. For example, the distinction between rheumatoid and osteoarthritis in a patient with arthritis of the knees may be largely academic: the fact that the patient has poor mobility is much more important. Emphasis is placed on physiotherapy, walking exercises and possibly surgery to improve mobility and independence.

HISTORY A full history remains the primary investigation. In a confused patient it is necessary to obtain additional information from a relative, community nurse, or someone who knows the patient well. The tempo with which the symptoms have developed is particularly important. Forget fulness gradually increasing over 2 or 3 years suggests Alzheimer's disease (AD), whereas a sudden onset of confusion over a day or two indicates a toxic confusional state and requires a search for underlying physical problems, such as an infection or myocardial infarct. A sudden onset of headaches or a recent change in bowel habit is never normal in old age, whereas gradually failing hearing and vision may be. The history must include a full list of the patient's medications, including details of dose and frequency; details of their functional ability (shopping, cooking, cleaning and personal toilet); and social circumstances. The main carer at home should be identified, as well as additional support that might be provided by family, neighbours and social services.

ASSESSMENT The chances of successfully discharging an elderly, frail person, if his or her homelessness, anxiety and impaired psychological state are ignored, are remote. Seeing the patient as a whole is thus not only good for the patient, it

is essential if limited health service resources are to be used effectively. Recent legislation and guidance place a responsibility on local authorities, health authorities and health providers to work together to assess, plan and meet the needs of people leaving hospital. An overall view of the patient includes assessment of medical, functional, social and psychological status, and usually brings together the professional views from several disciplines. The doctor attends to the physical and mental diagnosis and appropriate treatments. The nurse notes to what extent the patient can wash and feed him or herself or use the lavatory, and also particular problems such as incontinence, pressure sores or nocturnal confusion. The physiotherapist is concerned with whether the patient can get in and out of bed, rise from a chair, balance and walk; and as the patient improves, the occupational therapist assesses how well they can dress, bathe, cook and care for themselves. Discharge planning begins as soon as possible after admission. The social worker or discharge liaison nurse enquires into the patient's social background to find out what circumstances are like at home - whether they are already receiving support from relatives, neighbours, community nurse or social services, and whether this support will be continued or need to be increased once the patient returns home. Whether the patient's needs are simple or complex the social worker, together with the rest of the multidisciplinary team, draws up a care plan before discharge and later reviews this plan as an essential part of the care management process. The aim is to ensure that patients are speedily and effectively treated and returned to their own home in a planned and thoughtful manner, so that the gains resulting from treatment are likely to be maintained. A similar process of assessment can be initiated by family doctors using community staff, for patients living at home.

7

PHYSICAL EXAMINATION An extensive and detailed examination of every system can exhaust a frail patient. However, with practice, the

SUMMARY 5 Assessment of an elderly patient • Main stress is on functional disability, rather than underlying pathology. • Overall view essential: Medical status - manner of onset of symptoms is important. Full drug history required. Functional ability - washing, cooking, shopping etc. Psychological status - depression, anxiety, confusion etc. Social circumstances - support available at home from family and friends. • Involves a multidisciplinary team of doctors, nurses, occupational and physiotherapists and social workers.

191

examination can be thorough, informative and quick if the doctor concentrates on identifying disabilities and assessing loss of function.

General appearance The examination should begin with an observation of the patient's general appearance and clothes, noting any signs of self-neglect. Facial pallor is frequently present and so is a poor indicator of anaemia. Pupil size is often small, with reduced response, and upward gaze is often limited. Angular stomatitis - if not due to ill-fitting dentures - suggests multivitamin deficiency or iron deficiency anaemia, but a black hairy tongue or sublingual varicosities are of no significance.

Skin changes Loose, dry, scaly skin is a common finding, as is senile purpura owing to increased capillary fragility on the dorsum of the hands and extensor surface of the arms. Sunexposed skin should be searched for basal cell carcinomas or actinic keratoses, which are potentially malignant. The incidence of both conditions increases with years of exposure to the sun, and both are common in the elderly. Intertrigo - sore, inflamed patches of skin under the breasts or in the abdominal folds of obese patients - is often the result of poor personal hygiene. A keratinized transverse crease across the abdomen is the consequence of long-standing kyphosis, following osteoporotic spinal wedging. The breasts should be examined for lumps.

of unwanted drug effects (see p. 199). Systolic murmurs usually of the ejection type and present at the left sternal edge and aortic area - are found in up to 60% of the elderly. They are commonly associated with significant disease such as cardiac failure, ischaemic heart disease, arrhythmias and hypertension, and should not therefore be regarded as always benign.

Feet

Bowels and bladder

The ankles and feet should be inspected for signs of peripheral vascular disease (especially in diabetics), oedema and the need for chiropody. 1

A rectal examination is essential: carcinoma of the rectum or prostate is common in the elderly, and there may be loading of the rectum with faeces - a common cause of faecal incontinence. The presence of urinary incontinence should be noted and, particularly in bedridden patients, the pressure areas inspected for sores.

Cardiovascular Abnormalities of cardiac rhythm are frequently present: occasional ectopic beats occur in about one-sixth of elderly persons, but any other arrhythmia should be regarded as abnormal and investigated by electrocardiography. Systolic and diastolic blood pressure increases with age, although there is little change over the age of 65; values up to 160/90 can be regarded as normal. Overviews confirm that treating systolic or diastolic levels above this reduces total mortality as well as coronary and stroke mortality and morbidity in persons up to the age of 85 years (Fig. 7.13), but the value of treatment needs to be set against the risk

FIG. 7.13 Distribution by age of deaths from stroke in England and Wales The single most important risk factor for stroke is hypertension. Lowering the diastolic pressure at any level by 7mmHg halves the relative risk of stroke.

Eyesight and hearing The patient should be asked whether their vision is satisfactory; if it is not, it should be tested with eye charts (Fig. 7.14). Hearing is probably adequate if the patient can hear a normal conversational voice. Examination of the eardrums for wax should be routine, particularly where a hearing aid is worn, as the earpiece may compress wax into a hard, impenetrable lump.

Neurological changes 1

192

Fig. 7.11

Ankle jerks are absent in about 10% of elderly patients. Superficial abdominal reflexes are also often absent, but the presence of extensor-plantar responses is always pathological.

ing the benefits and risks of investigation and treatment. For example, a patient may be having repeated falls associated with loss of consciousness which do not bother them unduly but cause the relatives great stress and anxiety, because they dare not leave them alone in the house. Tilt table testing could be used to detect whether the falls are due to cardioinhibitory carotid sinus hypersensitivity. If they are, then a pacemaker can be safely inserted - even in the very elderly - with the potential for a considerable improvement in quality of life.

7

FIG. 7.14 Causes of visual impairment and blindness in the elderly

Vibration sense in the lower limbs is reduced, and may be absent at the ankles, in up to a half of patients over 75. However, joint position sense is usually normal. Appreciation of pain, temperature and light touch is very variable, but at least one of these sensations is impaired in about a quarter of the elderly. Both the glabellar tap, and primitive reflexes such as the pout reflex (tap the upper lip and the lips pout), grasp reflex and palmomental reflex (the palm is scratched and the chin wrinkles on the same side), can often be elicited in the elderly, particularly in those who have dementia.

Gait and balance Healthy old people show very little change in gait. There may be slight slowing, but a speed greater than 1m per second is still maintained. Nearly a fifth of the very elderly have a normal gait. A variety of pathologies, especially cerebrovascular disease, produce the senile (or cautious) gait typified by a normal to mildly widened base, shortened stride, slowing of walking and en bloc turns. There is mild disequilibrium in response to a light tap on the chest or a flick on the shoulders from behind, and difficulty in balancing on one foot (see section on falls).

WILL THE INVESTIGATION INFLUENCE MANAGEMENT? Before ordering an investigation it is necessary to consider whether the result is likely to usefully influence management. For example, a severely demented patient with an abdominal mass and positive faecal occult bloods could have the presumptive diagnosis of carcinoma of the colon confirmed by a barium enema. However, if surgery is not being considered (because of the patient's poor general condition and inability to cope with a colostomy), there is little point in the barium enema. On the other hand, if the same patient is found to be unable to walk following a fall, the hips should certainly be X-rayed. Repair of a fractured neck of femur will not only relieve the patient's pain: it is also much easier to nurse an ambulant patient than one who is bedridden.

WILL THE DIAGNOSIS AFFECT PROGNOSIS? It is important to consider whether the investigation will establish a diagnosis which significantly affects prognosis. A very frail patient with severe dyspepsia can be treated easily with an H2 antagonist. But if symptoms persist it may be important for the patient and relatives to know if the diagnosis is peptic ulceration or carcinoma of the stomach.

INVESTIGATION IN THE ELDERLY Potentially remediable conditions cannot be ignored simply because the patient is elderly, but it is not good medicine to order invasive, uncomfortable and possibly expensive tests when the potential benefit to the patient is minimal.

OVERALL STATE OF HEALTH Does the patient have the physical and mental capacity to benefit from a proposed treatment? If surgery is an option, would the patient be fit for operation? If the patient is debilitated and being looked after by family at home, then the views of these relatives must be included when assess-

ROUTINE SCREENING Most patients seen in clinic or admitted to hospital will have a full blood count, ESR (or viscosity), urea and electrolytes, glucose and TSH. Urine should be analysed for protein, glucose and blood. All of these tests can usefully detect treatable disease where the clinical diagnosis is unsuspected or doubtful. This is particularly true of thyroid disease, which is found on screening in about 4% of elderly inpatients in whom the classic symptoms and signs of hypoor hyperthyroidism may be absent. The interpretation of thyroid function tests can be difficult in the elderly because of the effects of severe systemic illness, which cause a reduction in circulating T3 often accompanied by a rise in reverse T3. As the illness becomes more severe T4 levels

193

fall, partly due to interference with binding to plasma proteins. TSH immunoradiometric assays can accurately measure the subnormal range and are now widely used as the single first line test of thyroid function. Although TSH measurements are usually reliable, in very severe illness this may not be true with subnormal levels being seen during the acute stage and sometimes a rebound to slightly above the normal range during recovery. A number of drugs can interfere with thyroid function, both lithium and amiodarone commonly cause goitre and either hypothyroidism or hyperthyroidism. In the case of amiodarone the total and free T4 levels are raised and cannot be used to diagnose thyrotoxicosis. A routine chest X-ray is usually taken, as about 5% of patients over 65 admitted to hospital are found to have an unexpected lesion on chest X-ray (such as a tumour, cardiomegaly or TB). It is particularly valuable in patients presenting with falls, immobility, incontinence or deteriorating mental function; nearly a quarter of these have cardiopulmonary disease without classic features. An electrocardiogram is not routine, but is required for patients with cardiopulmonary disease, strokes (quite commonly associated with a myocardial infarct) and falls (which might be caused by a cardiac arrhythmia).

In addition to the routine blood tests, the patient should have the serum B12 level checked and a VDRL test to exclude syphilis. Although it cannot positively diagnose dementia or Alzheimer's disease, a CT scan is a valuable screening test as it can identify multiple infarctions, spaceoccupying lesions and hydrocephalus. However, only 2% of all demented patients could expect an improvement in their mental state from treatment following a CT scan. Thus, CT is only indicated in patients with mild to moderate dementia (i.e. a mini mental state examination score of 10-26/30) of less than 1 year who are fit enough for surgery if a remediable lesion is found. 1

INVESTIGATION OF DEMENTIA

INVESTIGATION OF GASTROINTESTINAL LESIONS

The main concern in the investigation of dementia is not to miss a cause for which there is specific treatment. Although rare among elderly patients referred to a geriatrician or general physician, such reversible dementias are

In the investigation of upper gastrointestinal lesions, both barium meal and endoscopy give accurate results. However, endoscopy has the advantage of providing a tissue diagnosis where a tumour is suspected, and tends to be

relatively common among referrals to specialist neurological centres. Conditions which may respond to treatment include: • • • • • • • •

depression and other psychiatric disorders normal pressure (communicating) hydrocephalus resectable tumours subdural haematoma toxic drug effects thyroid disease pernicious anaemia syphilis.

RECENT ADVANCES IN THE CARE OF THE ELDERLY 1. Comprehensive geriatric assessment with immediate and ongoing specialist multidisciplinary interventions improves functional status and reduces the length of the initial hospital stay and the likelihood of subsequent readmissions. It also reduces the rate of immediate nursing home admissions and delays permanent nursing home placement. 2. Older adults who adopt or maintain a lifestyle with increased moderate daily physical activity derive considerable health benefits. The range of activities of daily living is increased, walking speed and balance improve, and there is a reduced risk of morbidity, institutionalization and mortality. Even social and productive activities involving little or no physical exertion improve survival.

1

194

MCQ 7.4

2

Fig. 7.12

3. The benefit of thrombolysis following myocardial infarction increases with age and is most cost effective in patients 75 years and older. The development of left ventricular failure is a more important determinant of outlook than age. 4. A daily supplement of 200 mg of vitamin E improves T-cell-mediated immune response in healthy elderly people. Vitamin E may also delay the progress of Alzheimer's disease and reduce cardiovascular events. Randomized controlled trials on the use of vitamin E have given conflicting results and its use as a regular dietary supplement is not currently recommended. 5. Hypertension as a risk factor for coronary artery disease, stroke and heart failure has a greater impact on elderly than on middle-aged patients.

RECENT ADVANCES IN THE CARE OF THE ELDERLY (CONTINUED) Lowering blood pressure, particularly systolic pressure, has a marked effect on reducing the incidence of stroke. Isolated systolic pressure greater than 140 mmHg or a diastolic greater than 85 mmHg should be treated, in patients up to the age of 80. No trial has looked specifically at those aged 80 and over, although the Hypertension in the Very Elderly Trial is underway. A meta-analysis of other trials which included very elderly subjects suggests that antihypertensive treatment reduces stroke, cardiovascular events and heart failure, but does not reduce mortality or major coronary events. At present no firm clinical guidelines can be given, but patients over the age of 80 should stay on existing treatment. Treatment decisions in newly diagnosed patients should be based on biological rather than chronological age. Treatment should begin with a low-dose thiazide, calcium antagonist or ACE inhibitor. 6. Non-rheumatic atrial fibrillation is associated with a fivefold increase in the risk of stroke or transient ischaemic attacks. Its prevalence increases with age to 11% in those aged over 75. The absolute annual risk of stroke in patients with atrial fibrillation is about 5% and this is increased in the presence of other stroke risk factors (see below). For primary prevention warfarin prevents 31 strokes for every 1000 patients treated per year, but at a cost of three haemorrhages. In secondary prevention warfarin prevents 80 strokes for every 1000 patients treated per year but at a cost of 21 major haemorrhages. Aspirin is less effective than warfarin in the secondary prevention of strokes in patients with atrial fibrillation, and prevents 40 strokes per 1000 patients treated each year at a cost of seven haemorrhages. Patients with non-rheumatic atrial fibrillation should be considered for warfarin if they are aged 65-75 and have one stroke risk factor, and over the age of 75 if they have two risk factors. Patients with no risk factors should be given aspirin. Paroxysmal atrial fibrillation carries the same risk of stroke as continuous atrial fibrillation, and patients should be considered for warfarin if they have one or more risk factors present. For patients on warfarin the INR should be kept around 2.0 to minimize the risk of bleeding. Aspirin or

better tolerated by a frail elderly patient. It is the investigation of choice but may need to be preceded by a barium swallow (to check if a pharyngeal pouch is present) when investigating dysphagia. For lower bowel disorders the first steps are usually rectal examination, sigmoidoscopy, stool culture, examination of the stool for occult blood and, finally, barium enema and possibly colonoscopy. Colonoscopy is the investigation

7

warfarin is best begun within a few days of the stroke, provided that a CT scan has excluded haemorrhage. If the cerebral infarct is very large there is a risk of haemorrhagic transformation and warfarin should not be started for 14 days. Contraindications to warfarin include cerebral haemorrhage, active peptic ulceration, uncontrolled hypertension, falls, moderate to severe cerebral periventricular small vessel disease and concerns over compliance. It is particularly important for elderly patients taking these drugs to have annual review to reassess the risk-benefit ratio. Treatment with aspirin is simpler and safer than warfarin and should always be considered for patients in atrial fibrillation when warfarin is difficult to control or is contraindicated.

Stroke risk factors • Previous stroke, transient ischaemic attack, myocardial infarct or peripheral arterial thromboembolism • Hypertension • Ischaemic heart disease • Recent heart failure • Diabetes mellitus • Thyrotoxicosis • Echocardiographic abnormalities, e.g. impaired left ventricular function, enlarged left atrium, enlarged left ventricle, intracavity thrombus. 7. Advances in surgical techniques have meant that major cardiac and vascular surgery is now routinely carried out on elderly patients. Thrombolysis and angioplasty allow successful treatment of peripheral ischaemia, and carotid endarterectomy should now be considered for a patient of any age who has a stenosis of greater than 70% and is either having transient ischaemic attacks or has made a good recovery from a cerebral infarct. Patients over the age of 70 years undergoing aortic valve replacement have a survival rate of 95% at 1 year and 75% at 5 years, which is no different from age- and sex-matched controls from the general population. Even in octogenarians undergoing aortic valve replacement the 5-year survival is 65%. By comparison aortic balloon valvuloplasty results are poor. 0

of choice for colonic adenoma, carcinoma or angiodysplasia, as it is more sensitive than barium enema and often better tolerated. Treatment of non-metastatic colonic malignancies gives a good prognosis; vigorous investigation of fit and active elderly patients is therefore justified. In the frailer patient, however, more caution is necessary before ordering a barium enema. Both the preparation (laxatives and sodium citrate enema) and the barium

195

enema itself are uncomfortable and difficult procedures for elderly patients to undergo. Ultrasound scans have the advantages of being non-invasive, quick and relatively cheap, and are particularly useful for investigating jaundice, abdominal masses, pancreatic tumours and possible liver metastasis. In some centres CT scanning of the colon is now established as an alternative to barium enema in frail patients. Investigation of faecal incontinence is described on page 187.

INVESTIGATION OF BACTERIURIA

196

A routine midstream urine (MSU) test in elderly patients admitted to hospital will show significant bacteriuria (bacterial count of more than 100000 organisms per mL) in 20-30% of men and women. Provided the patient is asymptomatic this bacteriuria is best ignored, as it is not associated with excess morbidity or mortality. A better guide to infection is pyuria, and in the presence of lower urinary tract symptoms even a sterile pyuria may be relevant. This is most conveniently detected by the finding of leukocytes on urine dipstick (e.g. Multistix 8 SG), with an MSU being sent for culture and sensitivity if dipstick testing is positive. The cardinal feature of a urinary tract infection is dysuria, but there is increasing recognition that some patients with urinary incontinence have no dysuria yet have a low-grade infection indicated by pyuria, and when this is treated with antibiotics the symptoms improve. Similarly, some patients complain of suprapubic or perineal pain during bladder filling, which may not be attributed to infection if the MSU fails to show a significant bacteriuria (i.e. more than 105 organisms per mL). However, it is possible that this criterion is inappropriate in these patients and bacterial counts of 102 per mL may be relevant. Urine should also be cultured for Chlamydia trachomatis, Trichomonas vaginalis and yeast. In summary, patients with a significant pyuria (positive urine dipstick or more than 10 leukocytes per highpowered film) who have lower urinary tract symptoms should be given a course of an appropriate antibiotic, such as trimethoprim or nitrofurantoin. Some patients require several weeks' treatment before the urine is clear. Patients with recurrent urinary tract infections associated with lower urinary tract symptoms should be considered for continuous low-dose antibiotic prophylaxis. A number of different regimens have been used, but all appear equally effective. Patients with recurrent urinary tract infections should have their postmicturition residual urine checked. If this is increased it is usually the result of detrusor hypocontractility or bladder outflow obstruction, and patients need to be advised on the importance of emptying the bladder completely as a chronic residual urine increases the risk of further infection. Surgery or intermittent self-catheterization may be needed. A group of patients often difficult to manage are those with recurrent episodes of suprapubic or perineal pain

during bladder filling, which is relieved by emptying and associated with frequency and urgency, but no actual urinary incontinence. Repeat urine testing is always unremarkable and antibiotics do not relieve symptoms. These patients usually have idiopathic detrusor hypersensitivity and some improve with bladder retraining. Patients with severe pain or with red cells in their urine should undergo cystoscopy to exclude interstitial cystitis or bladder cancer.

FURTHER READING ON INVESTIGATIONS IN THE ELDERLY Kafetz K 1998 Blood tests in elderly people and their interpretation. Rev Clin Gerontol. 8:305-318.

ASSESSMENT OF FITNESS FOR OPERATION Advanced age is no longer a contraindication to surgery as over the last 40 years the surgical mortality rate for patients over 80 has fallen from 25% to about 5%. In those over 90 mortality is around 10%. This compares with less than 2% for younger patients. This increased risk is only justified if surgery offers a reasonable prospect for improving the quality of the patient's life, or preventing loss of function. Average life expectancy for a 70-year-old male is 11 years and for a female 14 years, and many patients feel that this potential outweighs high surgical risks. Surgery for the relief of severe pain or complete bowel obstruction would probably be indicated even in the most severely disabled patient. However, a patient would need to be relatively fit and active before removal of a non-obstructive malignant bowel tumour, repair of an aortic aneurysm or a coronary artery bypass would be considered. Ageism should not be allowed to cloud the issue and each patient requires careful individual assessment. This is especially true of cancer in old age, which is less extensively investigated and receives less treatment than in younger patients, a difference not wholly explained by appropriate adjustment for comorbidity or frailty. Patients should always have a specialist opinion, as chemotherapy and radiotherapy protocols can be modified in less fit patients. The 1999 report of the National Confidential Enquiry into Perioperative Deaths highlighted the need for senior surgeons, anaesthetists and physicians to be closely involved in the care of elderly patients who have poor physical status and a high operative risk. Patients needing an urgent operation should be given a high surgical priority. The increased surgical mortality in the elderly is due to several factors, which are shown in Table 7.12. The mortality risk must be weighed against the likely final outcome of surgery. Total hip replacement has trans-

TABLE 7.12 Selecting patients for surgery or intensive care Chronological age alone is a poor indicator of outcome: • Good results in the elderly reflect good selection criteria • Consider the patient's previous health, level of physical independence, exercise tolerance and mental state • Patients' and families' views must be considered; they may be much less enthusiastic than the doctor about major surgery Increased mortality is associated with: • The presence of unsuspected heart disease • Reduced pulmonary and cardiac reserve (making a postoperative chest infection or myocardial infarct more likely) • Greater susceptibility to anaesthetic drugs • Greater risk of developing pressure sores postoperatively

is practicable). Many patients will have a minor ECG abnormality which provides a useful baseline measurement for comparing postoperative changes. Preoperative ECG abnormalities do not reliably predict postoperative cardiovascular complications, which can be expected in about one-fifth of elderly patients and consist mainly of heart failure with some cases of myocardial infarction. Cardiovascular disease is the major cause of increased mortality following general surgery in the elderly. Impaired mental function and poor motivation are also barriers to postoperative rehabilitation. If concentration and short-term memory are poor the patient may be unable to grasp the physiotherapist's instructions; episodes of mental confusion and agitation may also occur.

7

FURTHER READING ON SURGERY formed the lives of many old people and is now a routine operation. However, if chronic lung disease or an old stroke has reduced the patient's exercise tolerance to a few yards, the operation is unlikely to be worthwhile unless for the relief of severe pain. Before surgery, heart failure and diabetes should be controlled, dehydration and electrolyte disturbance corrected, and severe anaemia remedied by blood transfusion. A routine preoperative ECG is worthwhile. Its major use is in detecting a recent myocardial infarct, which would mean postponement of surgery for 6 months (if this

Seymour D G 1999 The ageing surgical patient - an update. Rev Clin Gerontol 9:221-233.

PRESCRIBING FOR THE ELDERLY A general rule when prescribing drugs for the elderly is 'not too many and not too much'. Although the elderly make up only 18% of the total population, they receive almost half of all prescriptions issued.

CASE STUDY 7.2 SURGERY IN OLDER PATIENTS A 90-year-old woman was admitted to hospital for rehabilitation following an accidental fall at home. She had a past history of left ventricular failure due to ischaemic heart disease, which was well controlled with drugs. She also suffered chronic osteoarthritis of her hips and 2 years previously she underwent an elective total hip replacement on the left. She lived alone in a house with downstairs bed and toilet. She was fairly independent and mobile around the house with the help of a walking frame. She had home care visits five times a week. She was taking furosemide (frusemide) 40mg o.d., captopril 25 mg b.d., diclofenac 50mg t.d.s., aspirin 75 mg o.d. and co-codamol 2 tablets q.d.s. She weighed 47.6kg.

A few days after admission she had an episode of haematemesis and melaena. Although frail she was alert and haemodynamically stable. Pulse 90 per minute, BP 140/100. Abdominal examination was normal, haemoglobin 8.8g, serum albumin 27 g. The rest of her blood results, chest X-ray and ECG were normal.

Question

1. How would you manage this patient?

She needs a blood transfusion (2 units were given) and an urgent oesophagogastroduodenoscopy (OGD). This showed a duodenal ulcer with a deformed pylorus. CLO

test was negative. The aspirin and diclofenac were stopped and she was given omeprazole 20 mg daily. Over the next few days she made good progress and was eating and drinking well, but she then had a second episode of haematemesis and melaena. Haemoglobin was 10.4 g. The omeprazole was increased to 40 mg daily. Three days later the nurses reported faecal vomiting. The abdomen was distended and tender, but bowel sounds were normal and the rectum was empty. Plain abdominal X-rays showed only gaseous distension with no signs of obstruction. She was treated with intravenous fluids, given nil by mouth, and had a phosphate enema. The nausea and vomiting persisted intermittently for the next few days. 197

CASE STUDY 7.2 CONTINUED A barium enema showed only diverticular disease of the colon. She was referred to the surgeons, who recommended the continuation of conservative management as she was considered unfit for surgical intervention.

Question 2. Do you think that this was the right decision?

The view of the physicians at the time was that this patient should have had surgery. Prior to admission she had been in relatively good health. Her heart failure was well controlled and she had successfully undergone a total hip replacement. All of these factors would point to a relatively good prognosis. The 1999 report of the National Confidential Enquiry into Perioperative Deaths stressed the importance of a team of senior surgeons, anaesthetists and physicians being closely involved in the care of elderly patients who have poor physical status and high operative risk. Furthermore, the experience of the surgeon and anaesthetist need to be matched to the physical status of the elderly patients as well as to the technical demands of the procedure. Following the decision not to operate the patient continued to have

198

nausea and vomiting and steadily deteriorated. Four weeks after the second episode of haematemesis her haemoglobin had dropped to 8.6g and her weight was now 38kg. She declined a repeat OGD and a barium meal showed complete pyloric obstruction. Again she was referred to the surgeons, who remained very reluctant to operate. The opinion of the anaesthetist was sought and he felt that surgery might be considered under epidural anaesthesia. Despite her general frailty the patient was fully alert, and after the available options and surgical risks were explained to her she opted for surgery. She was transferred to the intensive therapy unit where total parenteral nutrition was given for 3 days; by this stage her weight was 32 kg. She then underwent gastrojejunostomy under epidural anaesthesia, which was successful. She had an uneventful postoperative recovery. Three days later nasogastric tube feeding was started and 2 days after that she was swallowing liquids. After another 10 days she was able to start rehabilitation, and 2 months following surgery, by which time she weighed 38 kg, she successfully returned home. Discussion This patient had haematemesis and melaena as a result of peptic ulcer disease, presumably secondary to her intake of non-steroidal anti-

About 75% of those over the age of 65 and living at home are taking some kind of medication, with 15% receiving four drugs or more. However, multiple prescriptions increase the risk of drug interactions and adverse effects, and compliance is likely to be poor. Over 90% of people in residential and nursing homes receive prescribed medicines: many of these are unnecessary. Adverse reactions are two or three times more common in the elderly than in those under the age of 60. This is due primarily to slower rates of inactivation and excretion but also, with some drugs, to increased sensitivity

inflammatory drugs. She developed gastric outlet obstruction due to a deformed pylorus. Conservative management was initially adopted, but without success. Despite the high surgical risk this patient survived, but with hindsight it is apparent that the surgical risk would have been considerably less had the operation been carried out electively 2 or even 3 weeks previously. Emergency surgery for whatever reason in older people increases the surgical risk. Mortality in patients over the age of 70 undergoing emergency surgery is 20%, compared to 1.9% in patients undergoing elective operations. Postoperative complications and morbidity are also higher. Over the last few years surgery has become much safer owing to improvements in anaesthesia and surgical expertise. Age is generally a poor predictor of surgical morbidity or mortality, and certainly a much poorer predictor of outcome than severity of illness. Careful preoperative assessment of a patient's co-morbidities is more important than age in determining risk. This patient's major preoperative risk was associated with her poor nutritional status: weight loss of more than 20% or a serum albumin less than 35 g/L increases mortality. Giving preoperative total parenteral nutrition for severely malnourished patients probably reduces morbidity and mortality, although the evidence is inconclusive.

of the target organ. Thus, many drugs need to be given in lower dosage.

ALTERED PHARMACOKINETICS Drug absorption Changes in gastrointestinal function with increasing age include a rise in gastric pH, decreased gastric emptying time, decreased splanchnic blood flow, and loss of small

bowel mucosal surface area. Although these changes are theoretically rate-limiting for drug absorption, none appears to be important for passively absorbed drugs.

young patients. The drug plasma concentration may have to be measured.

Distribution

ALTERED PHARMACODYNAMICS

There are a number of theoretically important changes that affect drug distribution in the elderly: • A decrease in lean body mass and total body water reduces the distribution of water-soluble drugs. • Increase in body fat results in an increased volume of distribution of lipid-soluble drugs such as lidocaine (lignocaine), chlordiazepoxide and diazepam. • The small decrease in plasma albumin that occurs in the elderly will reduce drug binding and may thus result in an increased distribution of the drug. As it is the unbound fraction that is clinically active, there may be an increased pharmacological effect. This may be important in sick or malnourished patients, who have a greater reduction in plasma albumin concentration. • Conversely, if the drug binds predominantly to a1-acid glycoprotein (e.g. lidocaine (lignocaine) and disopyramide) then protein binding is likely to be normal, or even slightly increased, in the elderly.

Metabolism There is a reduced rate of hepatic metabolism with increasing age. This reduces presystemic metabolism and causes a marked increase in the bioavailability of drugs metabolized by the liver, such as propranolol, labetalol, lidocaine (lignocaine), clomethiazole (chlormethiazole) and verapamil. In the case of propranolol the age differences in plasma concentrations are large, with the mean concentration in the elderly being about twice that found in the young. However, because of reduced sensitivity, the dosage need not be lowered. Age-related effects on drug metabolism are variable, so that even some extensively metabolized drugs, e.g. warfarin or diazepam, may show no change in clearance.

Excretion Advancing age is accompanied by a reduction in both glomerular filtration rate and tubular function. At 60 years of age renal function is about half that at age 30, and will be further impaired by dehydration, cardiac failure and urinary retention, as well as by diseases such as diabetic nephropathy. Normal elderly people may have significant reductions in glomerular filtration rate and tubular function yet have a normal blood urea and serum creatinine. For drugs excreted by glomerular filtration the rate of excretion correlates with glomerular filtration rate. Some of these drugs, e.g. digoxin, gentamicin, lithium, cimetidine and chlorpropamide, have a low therapeutic index and their dosage therefore needs to be carefully chosen in the elderly. In general, a lower dose will be needed than in

7

The most clinically important result of altered pharmacodynamics occurs with drugs acting on the central nervous system. Sedatives and tranquillizers are more likely to produce drowsiness and hangover effects in the elderly. Although nitrazepam and diazepam have similar plasma concentrations in both young and old, psychological testing shows that the elderly make more mistakes and have longer reaction times for 36 hours after a single dose. They also show increased sensitivity to anaesthetic agents such as halothane. A reduced sensitivity to adrenergic agonists is seen in the elderly heart. Sensitivity to propranolol also decreases, so that despite the higher plasma propranolol concentrations seen in elderly subjects the degree of B-blockade is reduced. The dose of warfarin needed to produce anticoagulation in the elderly is reduced, but this is not due to an agerelated rise in plasma warfarin concentration. Warfarin acts by inhibiting the synthesis of vitamin K-dependent clotting factors, and this inhibitory effect is greater in the elderly, resulting in an increased sensitivity.

ADVERSE REACTIONS Whenever a reduction in drug elimination or increase in receptor sensitivity occurs, there is the likelihood of an excessive drug effect and hence an adverse reaction. The risk of these adverse reactions increases with age (Fig. 7.15A). About 3% of all admissions to geriatric units in the UK are due solely to unwanted drug effects; in a further 8% of cases an adverse reaction is a contributory cause. Two-thirds of all reactions are caused by two groups of drugs: • Cardiovascular drugs: diuretics, digoxin and hypotensive drugs • Drugs acting on the central nervous system: antiparkinsonian drugs, antidepressants, hypnotics and tranquilizers. Although more than 70% of patients taking cardiovascular drugs make a full recovery from the adverse reaction, only 46% recover from the ill effects of antiparkinsonian drugs. The risk of an adverse reaction increases with the number of drugs taken (Fig. 7.15B). Elderly people are more likely to have multiple diseases than the young, and the likelihood of polypharmacy is therefore high. Not only is the patient apt to be muddled by a large number of drugs and inadvertently take the wrong dose, but drug

199

FIG. 7.15 Adverse drug reactions A The increase in adverse drug reactions with age; B Prevalence of adverse reactions related to number of prescribed drugs. The risk in a patient taking only one drug is about 11%, but is 27% if six drugs are taken.

interactions lead directly to an increased adverse reaction rate. This is particularly likely to occur when drugs with a low therapeutic index, e.g. anticoagulants, antidepressants, anticonvulsants and antihypertensives, are used in combination. There are a number of important causes of adverse drug reaction in the elderly other than altered pharmacokinetics and pharmacodynamics. Inadequate clinical assessment is perhaps the most common. Dependent leg oedema, for example, is more often due to immobility than to congestive cardiac failure, and a diuretic is thus not an appropriate treatment. Similarly, old people who complain of dizziness are often expressing a sensation of unsteadiness and fear of falling, rather than true vertigo. Prochlorperazine - an effective vestibular sedative - is frequently prescribed in this situation but is quite useless and also carries a serious risk of parkinsonian side-effects. Loneliness and social problems often present as minor aches and pains, unhappiness and insomnia. Although both difficult and time-consuming, it is far better to deal with the underlying problems than simply to issue a number of drug prescriptions.

COMPLIANCE There is no evidence that the elderly are less compliant than the young, but they do make more medication errors. These occur in up to three-quarters of old people taking

1

200

MCQ 7.5

2

Fig. 7.13

drugs, but only about a quarter are potentially serious, usually because of underdosage. Medication errors are more commonly made by patients over the age of 75 who are living alone and are confused. Poor visual acuity often prevents an old person from reading the instructions correctly, and impaired manual dexterity may make it difficult or even impossible to unscrew the cap on the bottle (particularly if it is one of the child-resistant types). Compliance can be improved, and errors in medication reduced, by careful counselling and an explanation of when and how to take the drug. After a single lesson in inhaler technique up to 30% of elderly patients are still unable to use a metered dose inhaler. Further instruction may be needed, or patients can be given a breath-activated device or a large-volume spacer, which they find easier to use and thus prefer. Written instructions or a calendar pack can be helpful. Dosette boxes (devices which organize a week's supply of tablets) can be useful for patients who are motivated to take their drugs but are absentminded. They can be prepared by the pharmacy or by a carer. 1

CHOLESTEROL LOWERING IN STROKE Total and low-density lipoprotein (LDL) peak at around age 65 in males and 75 in females, and thereafter decline. There is probably also a decline in high-density lipoprotein (HDL) after age 65. Thus cholesterol levels measured in old age do not necessarily reflect lifetime levels. There is an inverse relationship between serum cholesterol and haemorrhagic stroke, and a direct relationship with ischaemic stroke. In general the association between cholesterol and stroke is considerably weaker than those reported for blood pressure, diabetes and atrial fibrillation. Systematic reviews indicate that for people with known CHD and cholesterol levels in the mid to elevated range statins significantly reduce stroke risk by approximately 32%. This is probably greater than the relative stroke risk reduction achieved with aspirin therapy (approximately 25%) and compares with the 40% resulting from antihypertensive treatment. The number needed to treat to prevent one stroke over 5 years using statins in patients with CHD is approximately 80-150, compared with 40-70 for treating elderly hypertensives. The best advice at present is that patients up to the age of 75 years with known CHD and total cholesterol >5 mmol/L will benefit from a statin to reduce the risk of a first stroke, and also to reduce the risk of further CHD events. Patients up to the age of 70 with cholesterol >5 mmol/L and no history of CHD, but at high cardiovascular risk, benefit from a statin because of reduced CHD risk. The cardiovascular risk can be assessed using various tables

RECENT ADVANCES IN STROKE REHABILITATION

RECENT ADVANCES IN DYSPHAGIA AFTER STROKE

• The accuracy and completeness of clinical notetaking is improved by using a clerking proforma. Although there is no reliable evidence it is likely that improved assessment leads to improved outcome. Certainly audit is facilitated and, where the process being audited is of proven effectiveness (e.g. the use of aspirin as secondary prevention of ischaemic stroke), it is reasonable to assume that improved performance (i.e. ensuring that all ischaemic stroke patients leave hospital taking aspirin) will lead to improved outcome. • An overview of controlled trials confirms that stroke units reduce early mortality by 28%, disability (measured as an 'activities of daily living' score) is significantly reduced and there is a substantial reduction in the odds of being dead or requiring institutional care 1 year after a stroke. 2 • The characteristics of a stroke unit are: coordinated multidisciplinary rehabilitation with involvement of carers; a specialist interest in stroke disease and rehabilitation; and training and education programmes for both staff and carers. • There is early evidence to suggest that, after discharge from hospital, stroke patients have better outcomes if physiotherapy is continued in their own homes rather than at geriatric day hospitals.

Patients should be routinely tested by asking them to sip some water and observing any tendency to choke, make a gargling sound when asked to say 'aah' or have a 'wet voice'. The gag reflex is unreliable. Patients with dysphagia should take nothing by mouth, be given intravenous fluid, and be assessed by a speech and language therapist. After 48 hours many clinicians start tube feeding, but there is no evidence that early feeding increases the proportion of patients surviving without severe disability, or indeed whether a nasogastric tube or percutaneous endoscopic/gastrostomy (PEG) is preferable. There is an indication from a small trial that a PEG is better tolerated and produces a greater improvement in nutritional state, but the evidence is not strong and there is considerable uncertainty regarding when tube feeding should begin, which is the best method, and, for patients who are able to swallow, whether nutritional supplements by mouth improve stroke outcome. A large international trial (FOOD) is under way to try to resolve these issues.

(e.g. Sheffield table or the New Zealand guidelines) that predict risk based as the ratio of total: HDL cholesterol, age, sex and the presence of hypertension, smoking and diabetes. • Patients who have had a non-haemorrhagic stroke or TIA are at high risk of a myocardial infarction, and a statin is justifiable for the secondary prevention of CHD if total cholesterol is >5mmol/L up to the age of 75.

REHABILITATION The aim of the rehabilitation team is to help restore maximum capability within the limits of the elderly patient's needs and disability.

PRINCIPLES OF REHABILITATION When setting a realistic rehabilitation goal it is important to consider the patient's needs. For a very elderly and frail

amputee wheelchair independence may be perfectly satisfactory, but this would not be acceptable for a fit 65-yearold amputee who is a keen golfer. Although rehabilitation is a core activity of geriatricians it has suffered in recent years as services have had to focus on the increasing number of acute admissions. As a result, rehabilitation goals are often set merely at achieving independence in activities of daily living (ADL), so that patients can be discharged from hospital as quickly as possible, but goals directed towards quality of life, such as driving the car or using public transport, may be just as important to the patient. Rehabilitation does not stop once the patient leaves hospital, and there is increasing interest in the role of rehabilitation as a means to prevent hospital readmission. Programmes of chronic disease management agreed between primary care and the relevant specialist teams improve outcomes and are cost effective. Hospital admissions can be reduced for patients with chronic obstructive pulmonary disease by comprehensive pulmonary rehabilitation (coordinating medical treatment, nutrition and exercise) and for heart failure by nurse-led monitoring and advice. For patients with Parkinson's disease home-based rehabilitation can be particularly effective. Patients can be taught to break up long movement sequences into separate steps and do only one movement in the sequence at a time. They can be advised to avoid doing two things at once, such as carrying a tray of drinks when walking, and to consciously think about each movement before starting to do it. Movement around the home can be improved by the occupational therapist providing visual cues, which are par-

7

201

ticularly useful in overcoming freezing episodes. Cue cards can be placed around the house with instructions such as 'go' or 'step', and strips of adhesive tape on the floor, particularly in confined spaces such as bathrooms or kitchens, can encourage more normal walking patterns. For patients who have had falls it might seem obvious that a home visit by an occupational therapist should concentrate on removing environmental hazards. However, apart from correcting obvious dangers (such as a broken light bulb at the top of the stairs) most studies on environmental manipulation have shown little benefit in the form of reduced rate of falls. However, what the occupational therapist can do is to teach the patient how to cope safely with various everyday activities and to help them regain their confidence. There is some evidence that this type of intervention reduces the risk of falling not only in the home, but outside it as well. The principles of rehabilitation in the elderly are essentially the same as for young patients, but some special features of old age should be noted. Multiple pathology is likely to be present, so that the speed with which a stroke patient can be mobilized is often hampered by shortness of breath due to chronic lung or heart disease. The mental barriers to recovery may also be greater, whether due to dementia or to the increased feelings of passivity and dependence sometimes shown by older people. There may be difficult social problems owing to lack of family support or poor housing. Rehabilitation is concerned with helping the patient to learn to do things for him or herself, and there may be times when it is better to let a stroke patient struggle to feed himself rather than to always rely on feeding by a carer. Patients enjoy exercising their painfully regained skills and it is essential to regain independence before leaving hospital and returning home.

THE MULTIDISCIPLINARY TEAM

202

The most effective approach to rehabilitation is through a multidisciplinary team, with each member contributing their special skills and working together with the patient and family towards an agreed goal. The team meets regularly to discuss the patient's progress and review future plans. With a good team the patient is undergoing rehabilitation all the time and not just during the half an hour a day spent with the physiotherapist or occupational therapist. Although elderly patients can be successfully rehabilitated on a busy general medical ward, there are definite advantages to grouping them together in one ward: the nurses develop greater expertise, the pace of treatment can be better geared to the patients' needs, and it is easier to create an atmosphere of 24-hour-a-day rehabilitation. Correct positioning of a stroke patient is something all nurses and doctors should be aware of (Fig. 7.16). Before discharge, the occupational therapist will take

FIG. 7.16 Positioning the stroke patient A Incorrect; B Correct. The forearm on the affected side should be supported and should point straight forwards with wrist and fingers extended and thumb abducted. The arm must not be allowed to hang from the shoulder in internal rotation. Choosing a suitable chair - with unrestricted view and with arm rests - is important3, C Transferring the stroke patient to a chair. Note the support given to the affected arm.

the patient on a home visit to see how well they cope in familiar surroundings. This also allows assessment of the need for a specific aid, e.g. a grab handle on the wall next to the lavatory, and the opportunity to have it fitted before the patient arrives home. The social worker not only organizes the care package at home or arranges funding for patients going into residential or nursing homes, but also plays a valuable role in providing support and advice. Increasing disability and loss of independence can cause considerable distress to both patient and family, which can be helped by skilled counselling.

FURTHER READING ON REHABILITATION Department of Health 2001 National Service Framework for Older People. Standard 5: Stroke, www.doh.gov.uk/nsf/olderpeople.htm Young J, Brown A, Forster A, Clare J 1999 An overview of rehabilitation for older people. Rev Clin Gerontol 9:183-196.

COMMUNITY CARE AND LONG-STAY CARE The demand for community and long-stay care rises rapidly with age. Of those aged 65-74 only 1 % are living in institutional care and only 6% are receiving visits from a home help. For those aged 85 and over the figures rise to 26% and 36%. In the UK the principal demographic change in recent years has been a one-third rise in the number of those aged 85 and over. The escalating cost of caring for this frail and dependent section of the population (Fig. 7.17) has caused governments of all industrialized countries to think hard about how this need might be met. In the UK the NHS and Community Care Act of

7

FIG. 7.17 Resources required to keep pace with demographic change Demographic change has the most marked effect on spending on long-term care C 1991 = £8 billion: this includes nursing and residential homes as well as NHS long-stay beds. A Hospital and community health services expenditure 1991 = £16 billion B General practitioners and prescription medicine expenditure 1991 =£5.5 billion.

1990 (implemented 1993) coordinates domiciliary, day care and respite services so that the old person can live in their own home for as long as possible; when this becomes no longer feasible they are cared for in private residential or nursing homes. The cost of this is borne by the individual until capital has dwindled to a threshold level (currently £18500), when the bill is then picked up by the local authority. Local social services departments and health authorities have to agree joint policies for providing both residential and nursing home care and arrangements for discharge from hospital to the community. An old person living at home in need of support has a care worker who identifies his or her needs and plans a package of care. This may involve home carers, district nurse, bathing attendant and 'meals on wheels'. Respite care to give the carers a break may be arranged in either a residential or nursing home or the local community hospital. Day care is available at local authority centres, where the needs are mainly social, or at the geriatric day hospital if the needs are medical. Patients in hospital may not be sent home until there is an agreed discharge plan based on a multidisciplinary assessment. If this assessment, whether carried out in hospital or in the community, determines that the patient is no longer able to cope at home then a decision will be made whether to admit them to a residential or nursing home (paid for as outlined above). Patients with complex conditions or multiple problems who need continuing weekly assessments by the geriatric medical team are cared for free of charge in an NHS long-stay hospital bed. These patients now make up only a tiny minority and the

FIG. 7.18 Vicious circle

NHS withdrawal from the long-stay sector has raised important issues about standards of medical care in private institutions. The Community Care Act has brought improvements in the form of a better coordinated, patient-orientated service, but the time patients spend in hospital waiting for transfer to institutional care has increased because of funding difficulties, and there are continuing uncertainties about the eligibility criteria for an NHS long-stay bed. Increasing acute hospital admissions and a declining number of available beds are putting unprecedented pressure on acute medical services. An important response has been to try to find ways of preventing inappropriate admissions of elderly patients to district general hospitals and to speed up the discharge of those whose investigations and treatment is completed. Initiatives include improved community rehabilitation services; hospital at home; rapid response teams to provide short-term support for patients sent home from A & E departments; and increasing use of community hospitals. This range of services is now described as intermediate care. Despite these developments, community services have not been expanded sufficiently to cope with the increased demand resulting from hospital bed closures. Elderly people are being discharged from hospital 'quicker and sicker', and this has two predictable results. The increased demand on social services departments from a few people with a high level of personal care needs means that very many people who require less intense service packages receive no service at all. Secondly, the pressure for earlier hospital discharge results in

203

insufficient rehabilitation, creating a vicious circle (Fig. 7.18). This will only by broken by joint working between the NHS and social service departments to develop better services. It should not be forgotten that in the UK, as in other countries, the unpaid support provided by relatives, friends and neighbours easily outweighs that provided by statutory services. Advising and helping these carers is an important part of the doctor's role. It is also important to remember that the quality of care in long-stay institutions, private or NHS, is crucially dependent on the visiting doctor's vigilance and enthusiasm. This applies not only to ensuring that high-quality nursing standards are maintained, but that the environment for the patients is made homely, pleasant and comfortable. The emphasis should be on educating patients to do as much for themselves as possible, and physiotherapy and occupa-

204

tional therapy may be entirely appropriate for certain patients.

FURTHER READING ON COMMUNITY CARE AND LONG-STAY CARE Black D, Bowman C 1997 Community institutional care for frail elderly people - time to structure professional responsibility. Br Med J 315:441-442. Department of Health 2001 National Service Framework for Older People. Standard 3: Intermediate care. www.doh.gov.uk/nsf/olderpeople.htm Nikolaus T, Specht-Leible N, Bach M et al 1999 A randomised trial of comprehensive geriatric assessment and home intervention in the care of hospitalised patients. Age Ageing 28:543-550. The Coming of Age - improving care services for older people. London: Audit Commission, 1997.

8

Psychological Medicine Jeremy Holmes

Classification and overview 205

Hysteria 235

The causes of psychiatric illness 210

Histrionic personality and related conditions 238

The psychiatric examination 211

Eating disorders 239

Treatments in psychological medicine 212 Anxiety 212

Drug dependence 243 Alcoholism 247

Loss, grief and bereavement 217

Mental handicap or learning difficulties 252

Depression 218

Psychiatry and the law 253

Suicide and parasuicide 223

Mental health legislation 253

Mania and hypomania 226

Medicine and the emotions 255

Schizophrenia 228

Sex and medicine 256

Paranoid states 231

Death and dying 257

psychiatry are psychology and psychoanalysis, child development and sociology. Neurochemistry and neuroimaging, and the new genetics and neo-Darwinism, are making increasingly important contributions to the subject. The central meeting ground for medicine and psychiatry is the doctor-patient relationship. All patients, whatever their illness, have feelings and fears that the doctor needs to understand and know how to handle. The meaning to the patient of a particular symptom - chest pain, for example, whether due to angina or to anxiety in a 40-yearold man whose father died suddenly of a heart attack in middle age - needs to be understood, in order to appreciate its impact on the patient, as well as diagnosed and investigated. Psychiatry is concerned with meaning as well as explanation, and with methods of communication with the patient, as well as the technology of treatment. Despite the inseparability of psyche and soma, many doctors find psychiatry difficult and confusing. This is partly because the subject is inherently complex and multifaceted, but also because psychiatry involves the doctor's own feelings and attitudes in a way that can often be avoided in other branches of medicine. In approaching psychiatry, doctors have to learn to deal not just with the patient, but also with themselves.

CLASSIFICATION AND OVERVIEW A simple classification of psychiatric illness is summarized in Table 8.1. Some aspects of psychiatry of old age are also considered in Chapter 7.

Somatoform disorders 233

MINDS AND BRAINS

Psychiatric disease constitutes a large proportion of disability and illness in society (Fig. 8.1 and Summary 1). The World Health Organization has identified depression as the main cause of disability throughout the world; suicide is one of the commonest cause of death in young people, especially males, in developed countries. Somatization disorders are among the most frequent cause of consultation in general practice. Throughout the western world, most of the large mental hospitals have now closed and acute psychiatric illness is treated in general hospital units and the community, alongside the other medical disciplines. The boundary between general medicine and psychiatry is much less rigid: patients and their illnesses are no longer seen by physicians purely as scientific puzzles to be solved: psychological aspects of physical illness are, through liaison psychiatry, now recognized to be of great importance. Medical science is, in turn, making an increasingly important impact on psychiatric knowledge and practice. The basic disciplines underlying

An eminent neurologist once stated that 'psychiatry is neurology without physical signs'. This would be true if psychiatry were simply concerned with the brain and its disorders; the problem of psychiatric diagnosis and classification would then be no different from that of any other system of the body. A significant part of psychiatry is indeed concerned with diseases of the brain - the organic psychoses, such as confusional states, for example, or the affective symptoms (i.e. disorder of mood) caused by adrenal disorders. The 19th century discovery that the psychosis known as general paralysis of the insane (GPI) was caused by syphilis stimulated one of the founding fathers of psychiatry, Emil Kraepelin, to search for other physical causes of insanity, with the expectation that all mental illness would soon be found to be caused by diseases of the brain. Psychiatry continues, rightly, to try to understand the structural and biochemical correlates of psychiatric disorders, but the Kraeplinian project can never be fully realized, partly because brain research is difficult and at a rather primitive stage, but also because psychiatry is concerned with minds as well as brains.

205

TABLE 8.1 Classification of psychiatric disorders Organic syndromes Delirium Dementia Intoxication and withdrawal syndromes Functional psychoses Schizophrenia Manic-depressive psychosis Paranoid states Neuroses Anxiety Phobias Depression Obsessive-compulsive neurosis Somatoform disorders Conversion Somatization Hypochondriasis

Others Addictions Alcohol Drugs Eating disorders Sexual dysfunction and deviation Personality disorders (Axis II in DSM) (see p. 180) A. Paranoid Schizoid B. Antisocial (psychopathic) Histrionic Borderline C. Avoidant Dependent Obsessive-compulsive

Prevalence Prevalence of psychiatric ill health in the community Prevalence of psychiatric disorder in general practice Prevalence of psychiatric disorder among schoolchildren Psychiatric referral rate from general practice Psychiatric beds as a proportion of all NHS beds Major disorders Severe depression Suicide attempts Suicide Chronic mental illness Severe mental handicap

SUMMARY 1 The prevalence of psychiatric morbidity among adults aged 16-64, living in private households, in Great Britain • 14% of adults have a neurotic health problem. Women are far more likely to suffer this type of problem than men. • The most common neurotic symptoms are fatigue (27%), sleep problems (25%), irritability (22%), and worry (20%). • The most prevalent neurotic disorder is mixed anxiety and depressive disorder (7.7%), followed by generalized anxiety disorder (3.1%). The prevalence of all neurotic disorders is higher among women than men. • Functional psychosis has a prevalence of 0.4%. • The overall rates of alcohol and drug dependence are 4.7% and 2.2%, respectively. Men are three times more likely than women to have alcohol dependence and twice as likely to be drug dependent. Alcohol and drug dependence are most prevalent among young adults, particularly young men aged 16-24.

Mind has three main properties: • The ability to experience emotions • Intentionality (i.e. having desires and goals) • Self-awareness.

206

Although psychiatric diagnosis approaches the mind as if it were an organ, this is often merely an analogy. When we speak of psychiatric illness we may be speaking about a brain disease, e.g. GPI, Alzheimer's disease or (possibly) some forms of schizophrenia, in the same way we speak of pneumonia as a disease of the lungs. On the other hand, although conditions such as personality disorders or agoraphobia are in some ways like illnesses, they also differ in many important respects because they affect the mind. They are best seen as disorders, rather than diseases. This important distinction between mind and brain may

Neurotic disorders Social pathology Chronic alcoholism (identified cases) Chronic alcohol ism (unidentified cases) Juvenile delinquency Problem families One-parent families with children under 15 FIG. 8.1 Epidemiology of psychiatric disorders A Prevalence of, and provision for, psychiatric ill health in the community. B Cases per annum of various psychiatric disorders in a typical British general practice population of 2500.

help to explain the discomfort some feel when confronted with the essentially rather simple issue of psychiatric diagnosis. The categories of mental illness are few and not difficult to memorize. The main problem is philosophical, rather than one of classification. A number of other difficulties may also be encountered on first approaching psychiatry. First, psychiatric diagnoses tend to be at a rather general level. They are symptomatic/syndromic rather than pathological/specific. The diagnosis of anxiety neurosis or schizophrenia is analogous to that of 'heart failure' rather than 'mitral stenosis'. Secondly, with the exception of the organic psychoses, the underlying pathology of psychiatric disorders is usually complex and multifactorial. This follows from the nature of mental phenomena. Just as there is usually no simple reason why an individual chooses medicine as a career, so

there is rarely a single 'cause' of depression. There may be many explanations for a particular symptom or problem. A third difficulty encountered in psychiatric diagnosis is that of deciding what is normal or abnormal, which is sometimes more a political or ethical question than a technical one. The doctor's own standards and feelings are involved in assessing a psychiatric patient. This also leads to two errors commonly made by newcomers to psychiatry: either the student identifies the patient with himself, and so decides there is nothing much wrong, or he identifies himself with the patient and worries that he must be psychiatrically ill! (Note that throughout this chapter 'he' is used to mean 'he or she'.) Two systems of classification are widely used in psychiatry: ICD (International Classification of Diseases), which is favoured in Europe, and DSM (Diagnostic and Statistical Manual), which has been developed in the USA by the American Psychiatric Association. The latter is interesting in that it is a multiaxial classification: Axis I diagnoses classify the psychiatric disorder; Axis II classifies any personality disorder that may or may not accompany the psychiatric illness; and Axis III allows for the diagnosis of any concomitant physical disorder.

FURTHER READING ON MINDS AND BRAINS Fulford K W 1989 Moral theory and medical practice. Cambridge University Press, Cambridge Kendell R 1975 The concept of disease and its implications for psychiatry. Br J Psychiatry 127: 305-315

PSYCHOSIS AND NEUROSIS The distinction between psychosis and neurosis has been and remains fundamental to psychiatric thinking, although neither is easily defined on its own, and some classifications have attempted to do away with the rather vague and blanket term neurosis. The difference turns on four main issues: • • • •

Relationship to reality Severity Relationship to normality Insight.

Relationship to reality The essence of psychosis is that the patient loses contact with everyday reality. This may happen as a result of either hallucinations or delusions, or both. Delusions and hallucinations are not a feature of neurosis. The neurotic's feelings and beliefs about himself may be distorted, but perception and understanding of the external world remains intact.

SUMMARY 2 The distinction between psychosis and neurosis Feature Relationship to reality

Psychosis Loss of contact with reality Delusions and/or hallucinations

Neurosis Perception of external reality intact Problem is with the self

Severity

More severe, more global disturbance

Areas of normal function remain

Relationship to normality

Clear distinction between normal and psychotic experience

Represents extremes of normal difficulties

Insight

Usually lacking

Present

8

Hallucinations A hallucination is a false perception. The patient undergoes a sensory experience in the absence of an external stimulus, for example hearing a voice when there is nobody there. For this to be diagnostic of psychosis it should be clear-cut; hearing an 'inner voice', which the person recognizes as really part of himself, is not a hallucination. An illusion should be distinguished from a hallucination in that it is an elaboration of a real percept, e.g. 'seeing faces in the fire', and is of less, if any, pathological significance. In hallucination, any of the five senses may be involved in the patient's altered perception of reality. Hearing Auditory hallucinations are probably the commonest type of hallucination, found especially in paranoid states, in schizophrenia, and occasionally in psychotic depression. Sight Contrary to popular belief, schizophrenic patients rarely 'see things': visual hallucinations occur most commonly in drug and alcohol withdrawal states, temporal lobe epilepsy and hysterical psychosis. Smell and taste Olfactory hallucinations are usually a sign of organic disease, e.g. temporal lobe seizures. Touch Tactile hallucinations are rare, but can occur in schizophrenia and drug withdrawal. Hallucinations are not always a sign of pathology. After bereavement, visual or auditory hallucinations of the lost loved one are common. Delusions A delusion is a fixed and firmly held false belief. The patient's perceptions are unaltered, but his interpretation of them is, and is at variance with common experience. The patient sees people chatting at a bus-stop and is convinced

207

they are talking about him, saying he is homosexual or a spy, for example. These beliefs may be well concealed and not evident in superficial conversation. We normally assume that our thoughts are private, that individuals are autonomous, that the non-human world is indifferent to our existence, that we are not in any special danger, and that we are not of any particular importance or interest to strangers. The deluded patient (e.g. in schizophrenia) has experiences that conflict with these everyday beliefs. The patient may feel that his thoughts can be interfered with from outside, or that his movements can be controlled by others; he may believe there is some special significance in the number plate of a passing car, that he is in great danger from hostile forces, or that television programmes or public events are caused by, or directed at, himself.

SUMMARY 3 Classification of psychosis Organic Acute (delirium) Chronic (dementia)

Functional Manic-depressive psychosis Schizophrenia

oneself and one's emotions and motives without denial or self-deception. Although insight in this second sense may be desirable, it bears no relationship to the presence or absence of illness. It is possible to be perfectly healthy and yet to lack self-knowledge, or to be profoundly insightful yet mentally ill.

THE PSYCHOSES Severity In general, psychotic disorders are more severe, more disabling and produce greater overall disturbance of mental functioning than neuroses. There are exceptions to this. An obsessional neurosis, for example one in which the individual may spend many hours a day engaged in complicated washing or dressing rituals, can be incapacitating.

Relationship to normality In psychosis there is a clear departure from normality. In contrast, minor neurotic symptoms, e.g. depression, anxiety, obsessional symptoms and hysteria, are common in normal people. The neurotic population is best seen as an extreme end of a normal distribution curve in which quantitative, but not qualitative, change has taken place.

Insight

208

It is commonly stated that the difference between neurosis and psychosis is that the psychotic 'lacks insight'. The psychotic patient attributes his symptoms to reality rather than illness, and cannot be persuaded out of these delusions or hallucinations; the neurotic, on the other hand, can usually see that his view of the world might be mistaken. However, although it is undoubtedly true that the psychotic lacks insight, there are several reasons for giving less prominence to insight than is customary. First, if the psychotic is identified on the basis of a loss of touch with reality, then he will, by definition, lack insight. If the patient knew that his delusions or hallucinations were false, he would not be deluded or hallucinating. Secondly, the patient often tacitly accepts that he is ill by coming for help, despite the apparent illogicality of complaining to a doctor about a delusion. Thirdly, the word 'insight' has two distinct meanings in psychiatry: one, given above, refers to the patient's relationship to reality and his illness; another more literal meaning refers to the ability to look into

Psychoses may be divided by their features and causes into two main groups: organic and functional (or, more accurately, 'of unknown origin'). In organic psychosis there is established biochemical, infective or structural brain disease. In functional psychosis no such disease process can be demonstrated. Organic psychoses can be divided into acute organic psychoses (also known.as acute confusional states, delirium or 'acute brain syndrome') and chronic organic psychoses (also known as dementia and 'chronic brain syndrome') (see also Ch. 24).

Functional psychosis vs. acute organic psychosis It is essential to be able to distinguish functional psychoses from acute confusional states, as the former require psychiatric, the latter medical, treatment. The key difference is the state of consciousness. In functional psychosis consciousness is clear and the patient fully orientated: it is possible to complete a complicated crossword puzzle while suffering from paranoid schizophrenia. In organic psychosis, however, there is almost always some disturbance of consciousness; this is usually variable ('fluctuating level of consciousness') and is often worse at night. Some causes of acute organic psychosis are given in Table 8.2.

Functional psychosis One hundred years ago, Kraepelin divided the functional psychoses into two main groups: manic-depressive psychosis, and schizophrenia (or, as he called it, 'dementia praecox'). Manic-depressive psychosis is primarily a disorder of mood. The patient becomes psychotically depressed or elated ('unipolar manic depression') or experiences episodes of both ('bipolar manic depression'). There are usually periods of normality between episodes of illness. Schizophrenia is primarily a disorder of thinking, and on the whole it produces, a more severe and more general disturbance of mental functioning. The prognosis is less good than for manic depression.

TABLE 8.2 Some causes of acute organic psychosis Cause

Examples

Systemic infection Metabolic disturbance Vitamin deficiency

Malaria, HIV, pneumonia (especially in the elderly) Electrolyte imbalance, renal failure, liver failure Thiamine deficiency (Wernicke-Korsakoff syndrome in alcoholics, and beriberi) Hypoglycaemia, hypothyroidism, steroid administration Anticonvulsants, hypnotics, antidepressants, LSD, amphetamines CVA, subarachnoid haemorrhage, tumour, cerebral abscess DTs, benzodiazepine withdrawal

Endocrine Drug intoxication Intracerebral Withdrawal syndrome

cal health than those who are mentally untroubled. Childhood adversity, for example, is strongly associated with poor physical health in adult life. This link was orginally conceptualized by Freud in terms of 'hysterical conversion disorders' - i.e. the 'conversion' of mental distress into bodily symptoms. The basic themes of hysteria were denial, dissociation, displacement and dependency. The patient denies a painful conflict and thus avoids anxiety or depression, but at a price: she (many hysterical patients are female) dissociates herself from her feelings and displaces the latter into her body instead. Psychic pain is thereby converted into physical symptomatology, and the patient becomes dependent on doctors and others. Although the possibly perjorative and aetiologically loaded term 'hysteria' has been replaced with the more neutral 'somatization' the underlying themes remain.

THE NEUROSES

Obsessive-compulsive neurosis

Unlike psychosis, in neurosis the patient experiences no alteration of external reality. Instead, he is troubled by, and tries to avoid, some unacceptable aspect of himself or of his internal reality. There are four main patterns of neurosis:

Here the patient deals with anxiety - often aroused by fear of his own feelings of aggression, sexuality or wish for power or independence - by an attempt at control. A ritual action or a repetitive thought is repeated over and over again in the magical hope that this will keep some imagined danger at bay. The patient knows it is absurd (thus has insight and is not psychotic), tries to resist, but still has, say, to wash in a particular way every morning in order to feel safe to start the day.

• • • •

Anxiety neurosis and phobia Depressive neurosis Hysteria and somatization Obsessive-compulsive neurosis.

Anxiety neurosis and phobia Anxiety is an almost universal feature of all psychiatric illness. It is the psychic equivalent of pain. Anxiety is normally a response to external threat or unpredictability in the environment. In anxiety neurosis, rather than being alerted and mobilized into action by anxiety, the patient is paralysed or overwhelmed by it. Where anxiety is experienced only in specific situations, the patient is said to suffer from a phobia. In phobic disorders the patient attempts to master anxiety by avoidance (e.g. an agoraphobic patient avoids public places).

Depressive neurosis If anxiety is the psychic equivalent of pain, depression is the equivalent of disability (although anxiety can also of course be disabling, and depression intensely painful). In depression, the patient has lost not a limb or a function but a relationship. It is usually a person who is lost, but the 'lost object' can also be a job or a house, or the patient's own self-esteem. Depression can be seen as an attempt to avoid painful reality by withdrawal.

Hysteria Body and mind are inseparable. There is good evidence that patients with psychiatric disorders have worse physi-

8

PERSONALITY DISORDER Individuals are recognizable not just by their appearance but by their personality which, despite ageing and change, remains relatively constant through time. Personalities can be classified dimensionally or categorically. A dimensional approach is exemplified by Hans Eysenk who, following his intellectual opponent Jung, devised a method for assessing personalities along the dimensions of extrovert (outgoing, gregarious, stimulusseeking) and introvert (inner-directed, shy, stimulusavoiding). The problem of differentiating between normal and abnormal personality is even more difficult than in neurosis. There are undoubtedly individuals whose personalities are very unusual, at least in the statistical sense, but they are not necessarily abnormal. An alternative approach to the classification of personality uses psychiatric categories to describe different types of individual. Someone suffering from schizoid personality, for example, is withdrawn and aloof, preoccupied with material objects or intellectual pursuits rather than people, lacking in close emotional contact, and is defensive and easily hurt. These features may only be a personality trait, and may even be advantageous: many creative people have schizoid tendencies. However, when taken to extremes they become a personality disorder. Here, the individual is so handicapped by

209

SUMMARY 4 Classification of personality Dimensional approaches e.g. extrovert/introvert DSM categorization A. Paranoid Schizoid

B. Antisocial (psychopathic) Borderline Histrionic

C. Avoidant Dependent Obsessive-compulsive

his personality that he seeks, or is brought for, help, e.g. someone with a schizoid personality who suffers from severe social isolation. Unlike psychosis and neurosis, a personality disorder is not an illness. Its features date back to childhood or adolescence, and so there is no premorbid period of normality. However, personality disorder often coexists with, and may predispose to, psychiatric illness (this is an example of 'comorbidity'). Personality disorders are conveniently grouped into three main types. Group A have some of the features of schizophrenia and paranoia, but without any overt signs of psychosis. They may, however, predispose to a psychotic illness, and eventually develop into one. Thus someone with a paranoid personality may, in later life, develop pamphrenia. In group B, sometimes described as 'dramatic', patients experience social difficulties that often bring them into conflict with authority. Many of the patients who present to casualty departments after episodes of deliberate self-harm (see below) suffer from borderline personality disorder, characterized by unstable relationships, substance abuse, chronic feelings of emptiness, and difficulty in regulating emotions. For histrionic personality disorder, see page 238, and for antisocial personality disorder, page 253. Group C are neurotic personalities typified by anxiety, avoidance of threat or challenge, and dependent relationships. An example is that of obsessional personality disorder. Here the individual is meticulous, controlled, orderly and rigid, and may find the sloppiness of others unbearable. Such characteristics may be advantageous in certain circumstances, e.g. the armed forces or the operating theatre, but may also bring the individual into conflict with others when flexibility is needed, e.g. under stress and in intimate relationships.

THE CAUSES OF PSYCHIATRIC ILLNESS

210

The complexity of the interactions underlying psychiatric illness can be confusing. A causal model of the type 'one pathogen-one disease' rarely applies in psychiatry. There are three main types of causal factor in psychiatric illness: genetic/biochemical, developmental, and social. All play at least some role in most psychiatric disturbances, and part of the art of psychiatry lies in assigning the appropriate weight to each factor and identifying which may be

FIG. 8.2 The origins of psychiatric disorder

the most important (Fig. 8.2). Most psychiatric illnesses are genuinely 'multifactorial' in origin. This is not surprising, considering the complexity of brain development and the numerous branching pathways that development can take. The eventual outcome in terms of mental health will depend on genetic inheritance, childhood environment, resistance or vulnerability factors, adverse and healthpromoting experiences, and accidental factors such as loss and trauma. Eliciting this personal narrative or life story is central to psychiatric history taking and provides an essential backdrop to understanding current psychiatric disorder.

Genetic/biochemical ('organic') Occasionally a single genetic defect can determine a psychiatric illness, e.g. the psychiatric manifestations of Huntington's chorea. However, although genetic inheritance may, as in schizophrenia, be a necessary organic cause of illness, it is rarely sufficient. Biochemical, viral, immunological and nutritional factors may well be important in a number of psychiatric illnesses, although few have so far been convincingly demonstrated as a unique cause.

Developmental ('infrapsychic') Each individual is a product of, and in a sense contains and continues, his past. Psychological development, especially that shaped by early childhood experience and family relationships, plays a crucial part in determining personality and ways of reacting to stress in later life.

Social ('interpersonal') Many psychiatric illnesses are triggered by recent social events, such as bereavement, divorce or unemployment. The patient's social and family situation can play a major role in both initiating his illness and helping or hindering his capacity to recover. The quality of family relationships is itself influenced by the personality, which is in turn a product of both genetic and developmental factors.

8

THE PSYCHIATRIC EXAMINATION The psychiatric examination, although similar to the medical examination insofar as it is a systematic gathering and ordering of facts and observations, is unique in a number of important ways. First, it consists of a detailed account of a person's biography and current circumstances. Second, in gathering the history the interviewer uses a much more open style than is usual in the closed-question 'checklist' approach of conventional history-taking. The 'facts' of psychiatry are often feelings, and these are often most effectively elicited by open questions such as 'How do you feel about...?' or 'What was the worst moment .. .?' Third, the interviewer's own reactions and thoughts, rather than being put aside, should be recognized as relevant data to be noted and recorded. The interviewer should maintain a non-judgemental, balanced, compassionate but neutral position. The interviewer should also be gentle but firm and should not avoid eliciting material (e.g. sexual or traumatic) that may be painful for the patient simply because of his own scruples or anxiety. The psychiatric history, which may have to be supplemented by interviewing close friends or relations, should include the following: • Reason for referral. Did the patient ask for help, was he encouraged to seek it, and if so by whom, or was he brought unwittingly or unwillingly to the doctor? Why now? • History of present complaint. A detailed account, as far as possible in the patient's own words, of what he experiences as 'wrong', unusual or bizarre, or would like to change in his life. • Previous psychiatric and medical history.

SUMMARY 5 Psychiatric history and examination History • Reason for referral • History of present complaint • Previous psychiatric history • Previous medical history • Drugs and alcohol • Family history - genogram • Personal history • Personality Examination • General appearance and behaviour • Speech - form and content • Mood • Abnormal beliefs • Abnormal experiences • Cognitive state • Insight • Interviewer's responses

FIG. 8.3 A genogram

Drug and alcohol history. Family history. A detailed description of the patient's 'family of origin' and 'family of procreation' (if they have one) can conveniently be recorded in the form of a family tree or 'genogram' (Fig. 8.3). Personal history. Detailed biography, starting at birth, recording all significant memories and events, including early separations, family atmosphere, school experiences and achievements, work history, psychosexual and marital history. Personality. A person's likes and dislikes, interests, attitudes and aptitudes should be recorded to establish a premorbid stage against which to evaluate the present condition.

MENTAL STATE EXAMINATION This is a systematic compilation of the abnormal phenomena observed and elicited by the interviewer or complained about by the patient. • General appearance and behaviour. How is the patient dressed? How does the patient relate to the interviewer - is he relaxed and friendly, or suspicious and taciturn? Does the patient sit immobile in the chair, or is he fidgety and restless?

211

• Speech. Is the speech normal in form and content, slowed down, or accelerated? Does it follow a logical progression or does it jump from idea to idea ('flight of ideas'); or is there evidence of a disorder of the logical progression of thoughts ('thought disorder')? • Mood. Is the mood normal, depressed or elated? Is the person troubled by feelings of guilt and unworthiness and suicidal thoughts, or is the patient grandiose and overexcited? Is the patient anxious or agitated? • Abnormal beliefs (delusions). Does the patient feel he is being spied on, talked about or got at? Are his thoughts or actions being interfered with? • Abnormal experiences (hallucinations). Has the patient had any bizarre or unusual sensations? Is he hearing noises or voices when no-one is around? • Cognitive assessment (for mini-mental state examination, see p. 1288). What is the general level of intelligence? Is there evidence of impairment of consciousness, immediate memory or recall? Is the patient fully oriented in time, place and person? • Insight. What is the patient's self-appraisal and theory about what is happening to him? • Response of the interviewer. How does the patient make the interviewer feel? Sympathetic? Angry? Irritated? Bored?

212

The examination should be concluded by a formulation containing the salient points of the history and mental state, with a differential diagnosis and some ideas about possible aetiological and precipitating factors. Risk assessment should be considered in each case, with a specific discussion about suicidality ('How suicidal do you feel at the moment', or 'How would you rate your suicidality on a scale of 0-10?'), and consideration of risk to others, based on past history of violence, if any. Example of a psychiatric formulation: 'Mr A is a 22-yearold unemployed man currently living in a bedsit. For the past few months he has become increasingly withdrawn and self-neglectful, and caused worry to his landlord, whom he has accused of spying on him. He has been smoking cannabis on a daily basis since leaving school at the age of 16. He has few friends, and has only held down casual jobs for short periods. In his background his maternal uncle suffered from schizophrenia; his parents split up when he was 8, he has lost contact with his father and does not get on with his stepfather, whom his mother married 2 years ago. On mental state examination he appears mildly depressed and shows evidence of paranoid delusions of being watched, and occasionally hears a voice warning him that he is in danger. There may be some suicide risk, but there is no history of any violence to others. The differential diagnosis includes schizophrenia, cannabis psychosis, depressive disorder, and insidious organic brain disease. A degree of maternal neglect (his delusion could reflect a psychotic expression of his feeling of being unnoticed), lack of a positive male role-model, low self-esteem and underachievement may all contribute to the underlying dynamic of his developing illness'.

FURTHER READING ON THE PSYCHIATRIC EXAMINATION Leff J, Isaacs A 1990 Psychiatric examination in clinical practice. Blackwell, Oxford Sims A 1995 Symptoms in the mind, 2nd edn. WB Saunders, London

TREATMENTS IN PSYCHOLOGICAL MEDICINE Psychiatric treatments are of three types: • physical - psychotropic drugs - ECT • social - housing - day centres • psychological - see Table 8.3

ANXIETY Anxiety, like pain, is an unpleasant necessity. A person unable to feel anxious would be severely incapacitated. Arousal, which may be experienced subjectively as anxiety, leads to an adaptive response to threat: fight or flight.

AROUSAL AND ANXIETY There are three main components of the arousal response: • Psychophysiological • Psychological • Interpersonal.

The psychophysiological component This is mediated by the autonomic nervous system and is accompanied by increased adrenaline (epinephrine),nora-

SUMMARY 6 Manifestations of anxiety Physical Tachycardia Tightness/pain in chest Difficulty in breathing Headache Paraesthesia Muscular pains Muscular weakness Dizziness Difficulty in swallowing Abdominal discomfort Diarrhoea Frequency of micturition

Psychological Irritability Inability to memorize Inability to concentrate Fear of impending death Fear of madness (depersonalization and derealization)

8

TABLE 8.3 Choosing a psychological therapy Principal recommendations

Evidence

• Psychological therapy should be routinely considered as an option when assessing mental health problems (B)

Psychological therapy shows benefits over no treatment for a wide range of mental health difficulties.

• Patients who are adjusting to life events, illnesses, disabilities or losses may benefit from brief therapies such as counselling. (B)

There is evidence of counselling effectiveness in mixed anxiety/depression, most effective when used with specified client groups, e.g. postnatal mothers, bereaved groups.

• Post-traumatic stress symptoms may be helped by psychological therapy, with most evidence for cognitive behavioural methods. Routine debriefing following traumatic events is not recommended. (A)

CBT has been found helpful. Some evidence of efficacy has been shown for other forms of psychological therapy. Single session debriefing appears to be unhelpful in preventing later disorders.

• Depression may be treated effectively with cognitive behavioural therapy (CBT) or interpersonal therapy (IPT). A number of other brief structured therapies for depression may be of benefit, such as psychodynamic therapy. (A)

CBT & IPT effectively reduce symptoms of depression. Benefit has also been found for other forms of psychological therapy.

• Anxiety disorders with marked symptomatic anxiety (panic disorder, agoraphobia, social phobia, obsessive-compulsive disorders, generalized anxiety disorders) are likely to benefit from cognitive behaviour therapy. (A)

CBT effectively reduces symptoms of panic and anxiety. Behaviour therapy and cognitive therapy both appear effective in treatment of obsessional problems.

• Psychological intervention should be considered for somatic complaints with a psychological component with most evidence for CBT in the treatment of chronic pain and chronic fatigue. (C)

Psychological therapies have benefit in range of somatic complaints including gastrointestional and gynaecological problems. CBT found more effective than control in improving functioning in chronic fatigue and chronic pain.

• Eating disorders can be treated with psychological therapy. Best evidence in bulimia nervosa is for CBT, IPT and family therapy for teenagers. Treatment usually includes psychoeducational methods. There is little strong evidence on the best therapy type for anorexia. (B)

Efficacy of CBT and IPT in bulimia has been established. Individual therapies have shown some benefit in anorexia, with little to distinguish treatment types. Early onset of anorexia may indicate family therapy, and later onset, broadly based individual therapy.

• Structured psychological therapies delivered by skilled practitioners can contribute to the longer-term treatment of personality disorders, (C)

Meta-analysis found a number of therapy approaches, both individual, group and milieu show some success, including dialectical behaviour therapy, psychoanalytic day hospital programme and therapeutic communities.

The recommendations are weighted as follows: A. Based on a consistent finding in a majority of studies in high quality systematic reviews or evidence from high quality studies. B. Based on at least one high quality trial, a weak or inconsistent finding in high quality reviews or a consistent finding in reviews that 'high quality'. C. Based on evidence from individual studies that do not meet all the criteria of 'high quality'. D. Based on evidence from structured expert consensus.

drenaline (norepinephrine) and corticosteroid secretion. These result in tachycardia, piloerection, sweating and pupillary dilatation; and subjective manifestations of autonomic activity, including paraesthesiae, giddiness, 'butterflies in the stomach', breathing difficulties, a sensation of painful pressure in the chest, and muscular tension, especially around the head and neck. These symptoms may be mistakenly attributed by both patients and doctors to organic illness.

The psychological component There is a continuum of arousal responses to threat, from alertness and readiness, through mild fear to severe anxiety and incapacitating panic. Effective action depends on an optimal state of arousal: little can be achieved when an individual is either barely aroused or in a state of panic. Examination nerves, if not too great, improve results. During a panic attack the patient experiences a sense of threat, or even of impending death. Chronic anxiety states

not meet all the criteria of

SUMMARY 7 Aetiology of anxiety syndromes Past Genetic factors (predisposition) Faulty learning (as a child) Early childhood relationships (Attachment theory)

Present External threat Relationships to attachment figures Internal conflict

may produce irritability, lack of concentration and a subjective sense of failing memory. Anxious patients fail to habituate to stimuli and so remain chronically overaroused; this can result in extreme sensitivity to sudden noises or movements. In extreme anxiety a state of detachment may supervene in which anxiety suddenly disappears, to be replaced by a feeling of being outside of oneself (depersonalization) and cut off from one's surroundings, which appear unreal and distant (derealizatiori). These

213

TABLE 8.4 Classification of anxiety syndromes Primary anxiety Generalized Situational Agoraphobia Social phobia Hypochondriasis Specific (e.g. spiders) Panic attacks (hyperventilation syndrome)

Secondary anxiety Psychiatric illness, e.g. depression Organic illness, e.g. thyrotoxicosis Alcohol-related withdrawal symptoms Benzodiazepine-related withdrawal symptoms

experiences may be misinterpreted by the patient as signs of 'going mad'.

The interpersonal component A frightened individual 'regresses' (see p. 217) and, childlike, clings to those he loves and on whom he depends. This is part of 'attachment behaviour' and, like anxiety, is adaptive: species survival depends on secure bonding. The dependency and excessive demands on others that so often accompany anxiety states (and which can be trying to both relatives and doctors) can be understood as part of the attachment response to threat.

anxiety syndromes because in a phobia the patient avoids the feared situation and is thus restricted by the underlying anxiety. Situational anxiety syndromes include: • Specific phobias, e.g. snake or spider phobia, or fear of heights. These have a good prognosis, are found in otherwise healthy individuals, and respond well to behavioural treatment (p. 215). • Agoraphobia is named after the Greek agora, or meeting place. In agoraphobia the patient is frightened to leave home unaccompanied, avoids public places such as supermarkets and department stores, and cannot travel on public transport, especially underground railways and aeroplanes (claustrophobia is usually part of agoraphobia). The patient is usually female and may be very dependent on her mother, daughter or husband. The prognosis is less good than with specific phobias, but behavioural therapy is usually helpful. Agoraphobic symptoms may develop as part of a depressive illness (p. 218). • Social phobia, i.e. fear of meeting people in social situations, e.g. parties, canteens. This affects mainly young people and is an extreme form of normal 'shyness'. Patients can be helped by social skills training in groups with fellow sufferers in which they 'role-play' the feared situations. • Hypochondriasis, a morbid fear of illness, leading to a frequent need for reassurance by family, friends and doctors.

ANXIETY SYNDROMES Anxiety may be a primary diagnosis, or secondary to another psychiatric or physical illness. It is a non-specific symptom which occurs in a wide range of psychiatric syndromes, including benzodiazepine withdrawal, depression, alcoholism and schizophrenia. It can also be an important manifestation of organic illness, particularly thyrotoxicosis, paroxysmal tachycardia, carcinoid syndrome, hypoglycaemic attacks, phaeochromocytoma and temporal lobe seizures.

CLASSIFICATION Primary anxiety syndromes may be classified as shown in Table 8.4.

Generalized anxiety Individuals who experience anxiety symptoms, either chronically or intermittently, without a clear stimulus or an obvious focus of fear, suffer so-called 'free-floating' anxiety.

Situational anxiety 214

In situational anxiety there is an identifiable precipitant to which the anxiety is related. Phobias are closely related to

Panic attacks (hyperventilation syndrome) These may be part of a specific syndrome, e.g. agoraphobia, or may occur on their own. They frequently present to physicians as medical emergencies and are usually selflimiting. The patient presents in a state of terror, overbreathing, complaining of chest pain, and often feeling he is dying. The diagnosis can be established by provoking these symptoms with voluntary overbreathing, which are then relieved by rebreathing from a paper bag. Tetany may be a feature. Organic causes (e.g. hyperthyroidism) must be eliminated and a positive psychiatric history should be established (i.e. a history of recent stress/conflict, or a background of chronic difficulty).

Aetiology There is no single cause of anxiety syndrome. Individuals vary in their tendency to become over-aroused: genetic factors may play a part in this. Anxiety may also arise from faulty learning. A child who senses his mother's tension in social situations may become socially phobic himself. Repeated avoidance of a feared stimulus then exaggerates the threat associated with it, as there is no opportunity for mastery through repeated exposure. Attachment theory explains how early childhood relationships may predispose to anxiety in later life. In order

to survive, a developing child must bond to an attachment figure (usually the mother). If the mother is consistently present, reliable in her handling and 'good enough' in her emotional responsiveness, then a bond of secure attachment, or 'secure base', will develop. This enables the child to engage in exploratory behaviour, to deal with new situations without becoming overanxious, and so gradually to separate, safe in the knowledge that the mother is available if necessary. The child builds up an internal image of a good parent, which provides a sense of security. If, on the other hand, the early relationship with the parent is disrupted - by separation or by inconsistent handling, for example - the child may develop a bond of anxious attachment. Rather than being able to explore and become autonomous, the child clings to the mother, at first physically and later emotionally, showing signs of distress when separation is threatened. This childhood experience then acts as a template for separations in later life. Faced with threat, the anxious individual clings to attachment figures • spouse, parent, child or doctor. The following factors should be considered in cases of anxiety: • The nature of the current external threat. This may be obvious, such as major illness in the patient or a close relative. The 'threat' may, on the other hand, appear as trivial and 'normal' as an adolescent leaving home, or a family holiday. For an anxiously attached individual these can be major hurdles. • The relationship to attachment figures, past and present. A pattern of anxious attachment in childhood now repeated in relationship to spouse or children is often found. • 'Internal threat'. Anxious patients often experience an inner conflict between their need for dependency and their feelings about those on whom they depend. The patient is threatened not so much by external circumstances as by himself. A patient who feels angry, for example with an unfaithful spouse, but is unable to express this because of fear of separation and loss, may develop anxiety symptoms. He fears to bite the hand that feeds. The patient is frightened by his own inner impulses: sexual or aggressive feelings, or the wish for independence and autonomy. Anger in particular often coexists with anxiety; the angry individual is often a frightened one, and vice versa.

SUMMARY 8 Management of anxiety syndromes • Exploratory psychotherapy • Cognitive-behavioural psychotherapy

• Psychotropic drugs

Exploration and resolution of underlying conflict. Non-directive. Aims to change patients' behaviour directly through desensitization, and/or to, challenge their underlying assumptions. Antidepressants, minor tranquillizers, B-blockers.

MANAGEMENT

8

Michael Balint, a psychoanalyst who pioneered the use of psychotherapeutic methods in medicine, wrote 'the doctor is a drug: the question is in what dose, and with what frequency it should be prescribed'. The two drugs most commonly prescribed for anxiety are tranquillizers and, via supportive psychotherapy, 'doctor'. Both can be effective in the short term, but can in the long term produce the complication of dependence, and may even exacerbate the symptoms of anxiety.

Exploratory psychotherapy Normal individuals who become anxious are usually relieved of their fears by reassurance, and the ability to reassure is thus an important part of a doctor's skill. Reassurance in neurotic individuals is often ineffective, however, because it deals only with the surface fear and not with the underlying conflict. A man who develops repeated attacks of pain, in which he is convinced that he is about to die of a heart attack, can be reassured that his ECG is normal and that he is a fit man, and be discharged. Yet a few days later he may develop another attack and present again to the doctor. The underlying fear, and the root cause of his anxiety, of which he may be unaware or unconscious, may be that his mother, on whom he is very dependent and who has recently had an illness, will die and that he will be left alone. Only when this fear is exposed and when he has learned to accept the dependent part of his himself, will his panic attacks cease. This kind of exploration is helpful in the evaluation of most cases of anxiety. In some cases it may need to be followed by longer-term psychotherapy. If the patient can learn to resolve his inner conflicts through his relationship with a doctor or therapist, and so move from a position of anxious attachment to one of secure attachment, he will be relieved of the symptoms.

Cognitive-behavioural psychotherapy (CBT) In exploratory psychotherapy the doctor aims to be understanding but is non-directive. In behavioural psychotherapy, on the other hand, the doctor aims to change the patient's behaviour directly. A phobic patient associates the feared stimulus (e.g. a spider or a department store) with unpleasant feelings. The therapy aims to desensitize the patient by gradual exposure to the feared stimulus, often in the reassuring presence of the therapist. An important feature of behavioural treatment for anxiety is relaxation training. Here the patient learns to control anxiety through muscular relaxation and deep breathing, sometimes with the aid of tape-recorded instructions. Physiological measures of arousal - galvanic skin response and pulse rate - may provide objective evidence of anxiety

215

TABLE 8.5 Side-effects of benzodiazepines Tolerance Drowsiness Incoodination, ataxia Emotional blunting Anger and irritability Diminished sexual response (women) Withdrawal syndrome: insomnia anxiety depression (may be profound) irritability muscular pains and twitching perceptual distortion fits

reduction. When self-administered, this is known as autogenie training. Behavioural therapy, as its name implies, works directly with how the patient behaves and acts in the world. The focus in cognitive therapy is on the patient's thoughts, which may themselves be as anxiety-provoking as real-life situations. Indeed, one of the basic principles of CBT is Epictetus' dictum: 'Men are troubled not so much by things as by their thoughts about things'. In cognitive therapy the patient's underlying assumptions and view of the world are challenged: for example, someone with fear of flying is asked to consider rationally the probabilities of an aircraft crashing and to compare them with accident rates for road and rail travel. Their tendencies to catastrophize (i.e. fear the worst) and dichotomize (divide the world into black and white, good and bad) are pointed out, and alternative strategies suggested.

Psychotropic drugs When anxiety is a manifestation of depressive illness, the treatment of depression with antidepressants (many of which are sedative) also relieves the anxiety. Short-term (up to 10 days) prescription of minor tranquillizers such as benzodiazepines is occasionally justified for the relief of stress-related anxiety. Longer-term prescription of tranquillizers should be avoided because of the likelihood of habituation and physiological dependence. B-Receptorblocking drugs relieve the peripheral symptoms of anxiety and may also occasionally be useful. Side-effects of benzodiazepines are common (Table 8.5) (see also p. 23).

DEFENCE MECHANISMS

216

Life is beset with threats. A variety of social and psychological defence mechanisms exist that keep anxiety within manageable limits. Many of these can be observed on

SUMMARY 9 Defence mechanisms against anxiety Denial Dissociation or conversion Splitting Projection

Idealization Regression Obsessional defence Intellectualization

hospital wards, where both patients and staff face the daily threat of illness and death and yet usually manage to remain cheerful and efficient.

Denial 'Everything's wonderful'. One way of dealing with fear or pain is simply to pretend it does not exist. Men are particularly prone to this in everyday life, but false cheerfulness (sometimes known as manic denial) may conceal unvoiced fears or sadness.

Dissociation or conversion 'I'm not worried about anything, doctor, it's just that I keep getting this pain ...'. Akin to denial is dissociation, in which a fear is displaced or converted into a bodily symptom. Patients who have inexplicable and elusive pains, in addition to established physical disease, may well be harbouring unspoken fears and conflicts.

Splitting 'You're wonderful, doctor, not like those terrible doctors at the other hospital.' As Trojans should fear Greeks bearing gifts, so doctors should beware the patient who heaps excessive praise (as opposed to expressing genuine gratitude). The patient may be dealing with anxiety by splitting his world into good and bad, and clinging to the good in the vain hope of absolute security. As with the Trojans, the tables may suddenly turn and we may be seen as bad, and they good. Patients can sometimes split ward staff in this way; physicians and nurses must learn to take both praise and condemnation with equanimity.

Projection 'It's the psychiatrists who are mad, not me'. Closely linked to splitting is projection, in which unwanted feelings are attributed instead to a convenient other.

Idealization 'The doctors here are the best in the world'. The defence of idealization is closely linked to splitting and projection. It is probably best seen as an example of

positive transference, in which past childhood feelings towards one's parents are, under the threat of illness, transferred on to current medical attendants.

Regression

SUMMARY 10 Mourning responses • • • •

Denial Numbness Searching Guilt

8 • Anger • Detachment • Acceptance

'Nurse ... nurse ...'. The sick, like children, are tucked up in bed, fed, pampered and waited on. This sometimes helps the physiology of recovery, but can also lead to helpless behaviour on the part of the patient, who may deal with the fear associated with illness by becoming more and more childlike, regressed and demanding.

Obsessional defences 'I always have my bowels open at such-and-such o'clock, doctor.'

MOURNING The mourning response is divided for convenience into a number of stages, which can be summarized as denial, despair and detachment, but emotional reality does not conform to any such neat sequence. Grief is characterized by surges and cycles of often confused feelings, rather than an orderly progression of recognizable stages.

Denial

The 'malade-d-petit-papief - the patient who presents with a written list of symptoms - is a familiar figure in medical folklore. The obsessional defence is an attempt to deal with fear by control. This may produce feelings of irritation and discomfort in the doctor that mirror the patient's feelings. This is an example of how the doctor's subjective response to the patient (technically known as countertransference} can be a useful guide to what the patient is feeling.

The initial reaction, especially to sudden bereavement, may be denial: 'Oh no .. .'. This attempt to resist the reality of death may persist long after the bereaved have consciously and rationally accepted their loss; they find themselves talking to their lost loved one, or may 'hear' or 'see' them in one of the hallucinatory experiences that not uncommonly occur after bereavement.

Intellectualization

Another early response to bereavement is a complete absence of feeling. The bereaved person may appear to be coping very well because of this emotional shutdown which, like the analgesia of a soldier injured in battle, enables him to get through the initial phase of loss. When this is prolonged, however, it produces chronic impoverishment of emotional life.

'Just a few more tests should clinch the diagnosis.' This defence applies more to doctors than to their patients. Incurable illness and death are inescapable reminders of the limitations of medicine. The feelings of inadequacy and sadness that this creates may be defended against by a preoccupation with diagnostic precision and excessive investigation and treatment.

FURTHER READING ON ANXIETY Holmes J 1993 John Bowlby and attachment theory. Routledge, London Marks I 1987 Fears and phobias. Oxford University Press, Oxford

LOSS, GRIEF AND BEREAVEMENT

Numbness

Searching Part of the 'work' of grief involves a mental combingthrough of the events that led up to the loss of the loved one. This phase is often accompanied by intense anxiety and restlessness. Searching, a form of separation anxiety, may lead to apparently aimless wandering, or visits to familiar spots associated with the lost one. The bereaved person appears to be struggling to negate the reality of the loss, or to rewrite history with a different, happy ending. Acceptance of loss only occurs when this impossible task has been attempted countless times.

Guilt Man is not an island, and human beings are inevitably vulnerable to loss. The mourning response - a patterned sequence of psychophysiological reactions to the loss of a loved one - is as deeply embedded in our psyche as are the phases of inflammation in our pathophysiology. Common to both are the experience of pain, immobility and gradual recovery.

Pangs of guilt or self-blame are a very common feature of bereavement. Survivors blame themselves, often quite irrationally, for the loss of their beloved. 'If only', they say, T had phoned the doctor earlier/not gone on holiday/been kinder...'. This response can be seen as an example of omnipotence, a relic of childish thinking which attempts to deny human vulnerability to fate and fortune.

217

Anger Anger and guilt go together in bereavement. Outbursts of anger towards medical staff or close relatives often occur after a death, and should be met by acknowledgement of the underlying pain rather than with retaliation.

Detachment As mourning proceeds, so the grieving individual begins to separate himself from the loss and develop new interests and attachments. Pangs of grief recur, particularly at times of stress and on special dates such as birthdays, Christmas, and anniversaries of loss.

one. Feelings of guilt (associated perhaps with past fantasy wishes that the person would die) are more intense in such cases. 'Symbiotic' relationships, such as those involving unmarried daughters living with their mothers, are also often difficult to recover from when they are severed. Personality of the survivor. Dependent, insecure or depression-prone individuals are more vulnerable to prolonged grief, as are those with difficulty in expressing feelings. The latter may in part account for the greater vulnerability of men to lasting problems after bereavement. Social circumstances of the survivor. A lack of close family, friends, religious beliefs or employment all make recovery from loss more difficult.

Acceptance When mourning is 'successful' and circumstances favourable, the bereaved individual may come to accept that death and loss are part of life and may sometimes even be enriched by the experience. More usually, scars remain, and many bereaved people continue to struggle with feelings of meaninglessness and diminished vitality for many years.

Physiological changes Psychological reactions to loss are accompanied by physiological changes, including disturbance of corticosteroid metabolism and depression of T-lymphocyte response. Death from 'a broken heart' is a reality: mortality rates, mainly from coronary artery disease, are higher in widowers and widows than in their non-bereaved counterparts.

DELAYED OR PATHOLOGICAL MOURNING Return to near-normality after a major loss generally takes from 1 to 3 years. However, chronic states of depression, numbness, anger, anxiety and hopelessness can continue for much longer following a bereavement. The patient becomes stuck in one of the phases of mourning and is unable to progress towards detachment and acceptance. The likelihood of pathological grief depends on four main factors: • Type of death. A sudden, untimely (of a child, for example) or horrifying death is especially hard to accept (see Recent Advances in Post-traumatic Stress Disorder). • Nature of the relationship. An ambivalent relationship (one in which love and hate coexist) is more likely to lead to pathological mourning than a straightforward

1

218

MCQ 8.1

MANAGEMENT About one-third of bereaved persons may be in need of professional support or intervention. Support groups, such as those for widows (in the UK, CRUSE) or for parents whose children have died (the Compassionate Friends), can be very helpful. Psychotherapy may be necessary in order to recapture and work through buried feelings associated with the loss. Sometimes this is systematized as guided mourning, in which the sufferer is encouraged to think about the lost one, talk about the details of his or her death, look at photographs of him or her, visit the grave, and so face and accept loss with diminished denial.

FURTHER READING ON LOSS, GRIEF AND BEREAVEMENT Garland C (ed) 1999 Understanding trauma: a psychoanalytical approach. Butterworths, London Parkes C M 1988 Bereavement: studies of grief in adult life, 2nd edn. Penguin, London

DEPRESSION The complaint of 'depression' can refer to a number of different feelings and reactions - anger, boredom, frustration,

SUMMARY 11 Features of depressive illness (as opposed to unhappiness) • • • • •

Sustained lowering of mood Severe Definite onset Low self-esteem Physiological changes: disturbance of sleep, appetite, weight, libido • Sometimes atypical features brought to the surface by depressive illness: pain, obsessional thoughts, violence

RECENT ADVANCES IN POST-TRAUMATIC STRESS DISORDER (PTSD) The word trauma comes from the Greek 'to pierce'. It is now known that trauma, whether individual or in mass disasters, pierces not just the body but also the psyche of those affected, either directly, or as witnesses or relatives. Also, the recognition that as many as 60% of psychiatric inpatients, whatever their diagnosis, may have been sexually abused in childhood suggests that many psychiatric disorders have PTSD as at least a component of their aetiology. These people, after diagnosed as 'personality disordered', are sometimes referred to as suffering from 'complex PTSN'. In the USA the lifetime prevalence of traumatic events - accidents, suicide, homicide or sexual assault is estimated to be around 70%. Of those so exposed, 20-40% will develop post-traumatic stress disorder. Diagnostic criteria include feelings of intense fear, horror and hopelessness at the time of the traumatic event; intrusive imagery ('flashbacks') in waking consciousness and dreams; avoidance behaviour, e.g. being unable to pass the spot where an accident occurred, or to read accounts of similar events; hyperarousal phenomena, such as disturbed sleep, increased reactivity to noise, muscular tension, poor concentration; and significant interference with normal coping behaviour; all of these lasting for more than one month after the event.

Risk factors include the nature of the trauma itself: the more horrific and violent, the greater the likelihood of PTSD; denial of emotional difficulty on the part of the sufferer; a previous history of trauma, especially in childhood; and dismissal of the severity of the symptoms by significant others such as family and doctors. Comorbidity (i.e. the coexistence of other diagnoses) is common, especially affective disorder. Neurobiology of PTSD. It is now known that prolonged exposure to high cortisol levels (as occurs in clinical stress) produces shrinkage of the hippocampus and that PTSD is associated with neuro-anatomical changes, especially in the non-dominant hemisphere. Management includes: (a) 'debriefing' - a detailed 'action replay' of the traumatic event and the sufferer's reactions, although there is some evidence to suggest that this may have a reinforcing effect and so should be used judiciously; (b) cognitive-behavioural treatment directed at intervening in the intrusive thought-avoidance cycle: the subject is systematically asked to recall the painful events until they can look squarely at them without undue anxiety; and (c) long-term psychotherapy if the trauma has reawakened previous childhood trauma such as sexual abuse; (d) in some cases antidepressant medication can be helpful.

anxiety, hopelessness and helplessness - as well as the specific psychophysiological state that psychiatry recognizes as 'depressive illness' and which is sometimes colloquially referred to as 'clinical depression'.

table, and takes no pleasure in anything. He feels he has no future and that it is pointless or impossible to make plans. Thoughts of suicide almost invariably present, although the patient may try to resist these by saying I would not have the courage'. • Low self-esteem. The patient feels worthless, inadequate, unwanted and useless. He feels intense guilt and plagues himself with thoughts of past failures and mistakes. He can only see himself in a negative light and emphasizes his weaknesses and faults. • Physiological changes. The most important of these is sleep disturbance. Anorexia, weight loss, loss of libido, constipation, retardation (slowing of thought and movement) or occasional agitation are also all important features, especially in severe or psychotic depression. • 'Atypical' features. Sometimes a depressed patient may present atypically, with physical symptoms such as unexplained pain, obsessional thoughts, agoraphobia, criminal behaviour (such as shoplifting) or episodes of anger and violence. These may have been brought to the surface by a depressive illness. In a so-called 'smiling' or 'masked' depression, the patient conceals severe depressive feelings behind a cheerful facade.

DEPRESSIVE ILLNESS Clinical features Clinical depression can be distinguished from both 'ordinary' depression or sadness (transient depressive moods) and depressive personality by three main features: • Depressive illness is a sustained depression of mood lasting for weeks or months. • The seventy of the depression is great enough to interfere with normal functioning. • The depression has a definite onset, with a distinct change from normal to depressive thinking. Depressive illness involves an alteration of many different aspects of mind, body and behaviour. It is nearly twice as common in women as in men. The features of depression may be summarized under four main headings: • Depressed mood. The patient feels weepy and miserable, 'low in spirits', anxious, lacking in concentration and irri-

Although the presentation of depressive illness can vary greatly, there are two common clinical pictures (Table 8.6). 1

8

219

TABLE 8.6 Clinical features of neurotic and psychotic depression Clinical features

Psychotic

Neurotic

Severity of depression Delusions/hallucinations Sleep disturbance

More severe Yes Early morning wakening Yes, prominent

Less severe No Difficulty in getting off to sleep Yes/No

Worse on wakening and a.m. Yes Yes

No, or worse p.m.

Yes No

Sometimes

No

Yes

Yes

Yes

No

'Vegetative features', e.g. weight loss, constipation Diurnal mood variation 'Reactive', i.e. precipitated by loss Genetic factors, e.g. family history of manic-depressive psychosis Dexamethasone suppression test abnormal Response to tricyclic antidepressants Response to ECT

Neurotic depression Here the patient, despite depressive feelings, remains in contact with reality: 'I know it is stupid, but I feel as though I have let everybody down'. The patient might be a woman in her 30s with several small children, living in poor housing and with a husband who drinks and who is having an affair which she has just discovered. She is tearful, has lost weight and cannot fall asleep at night. She feels she is not coping with her job or housework, and no longer enjoys company or television.

Psychotic depression

220

Depressive psychosis is a part of manic-depressive psychosis, an affective psychosis. The depth of the depression is here much greater than in neurotic depression. The ratio of women to men affected is more equal. The patient becomes immobilized in both thought and action. He wakes very early in the morning and may have diurnal mood variation, feeling at his worst in the morning and improving slightly towards the evening. Depressive delusions develop, the patient becoming convinced that he has, for example, brought ruin on his family, that he is evil or ugly or radioactive. Depressive hallucinations may occur: the patient may hear a voice telling him that he is evil, or he may feel that his insides are rotting and complain that he smells. Endocrine disturbance may occur in psychotic depression, with disruption of the pituitary-adrenal axis. In some cases the normal diurnal variation of cortisol output is abolished, providing the basis of the dexamethasone suppression test, in which dexamethasone is administered but fails to reduce urinary steroid levels.

SUMMARY 12 Factors contributing to depression • Loss • Genetic • Biochemical, e.g. reduced catecholamine levels at postsynaptic nerve endings • Social, e.g. lack of close relationships, unemployment, lack of good childhood experiences to build self-esteem, learned helplessness

DIFFERENTIAL DIAGNOSIS OF DEPRESSION The differential diagnosis of depression includes: • Organic illness, especially hypothyroidism • Drug administration, especially steroids and some antihypertensive drugs • Alcohol-induced depression, where depression is secondary to alcoholism (or drug addiction), and withdrawal of the intoxicant results in lifting of depressive symptoms • Anxiety states, as anxiety is a common feature of depression. When they coexist, treatment of the depression usually (but not always) removes the anxiety • Other psychoses, especially schizophrenia.

THE CAUSES OF DEPRESSION Depression is often divided into: • Reactive depression, whose symptoms correspond with neurotic depression and which is said to be a response to external events; • Endogenous depression, where the symptoms are psychotic in character and where there is no clear environmental precipitant. The distinction is misleading. Psychotic depression is usually also 'reactive' in the sense that it is triggered by a painful event in the patient's life; and neurotic depression can be 'endogenous' in the sense that genetic and personality factors influence the likelihood of an individual developing depression in response to misfortune. Loss Loss plays a central role in the origins of depression and is the link connecting grief with depression. The grieving person is in pain not just because of the loss of an external relationship, but because he has lost a part of himself. His internal world as well as his external world has been impoverished. The similarity between the clinical features of grief and those of depression are striking. In grief, an external loss dominates the picture; in depression, it is an internal feeling of loss. In 70% of depressions there has been a preceding external loss or stress: a bereavement (delayed grief is best seen as a special case of depression),

loss of job, break-up of a relationship, a child leaving home or a theft, for example. It is when this then leads to a feeling of internal loss - of optimism, vitality and (especially) of self-esteem - that a depression occurs. Whether or not a particular individual reacts to loss by developing depression depends on a number of interrelated factors, genetic, social, developmental and psychological.

Genetic/biochemical make-up Twin studies have shown that genetic factors play an important part in manic-depressive psychosis. Changes in neurotransmitter biochemistry in depression are important: current views emphasize a reduction of catecholamine levels at postsynaptic nerve endings. Antidepressant drugs may work by preventing the breakdown of cerebral amines (monoamine oxidase inhibitors), or reducing their reuptake (tricylic drugs) at central nerve endings. Similarly selective serotonin reuptake inhibitors (SSRIs) increase the concentration of serotonergic neurotransmitters in the synaptic cleft.

Current social situation The presence of a close confiding relationship, usually with a spouse, protects individuals from developing depression after the experience of loss. Employment is another protective factor; the unemployed are thus doubly vulnerable to depression, the loss of a job being a precipitant and lack of a job being a vulnerability factor. Women with three or more children under 15 and long-term social and financial difficulties are more vulnerable to depression following loss.

Early childhood experience Women who have lost their mothers before the age of 11 - whether through death, separation (for example, due to illness) or divorce - are especially vulnerable to depression in later life. A reservoir of good experiences in childhood is needed for an individual to develop a lasting sense of self-esteem. This provides a child with a good and secure internal world which can withstand the inevitable separations and losses of later life, and enables them to respond to such situations with appropriate sadness rather than by developing depression.

Psychological factors Recent studies of depression have focused on the vicious circle of depressive thinking and its accompanying negative assumptions about the self and the world. The depressed individual has developed learned helplessness, feeling that nothing he can do will make any difference to the situation. Actions and experiences that might make the individual feel better are avoided, and he continues to view himself in a negative light. His depression thus becomes a self-fulfilling prophecy. Depression produces and may be produced by patterns of dysfunctional thought, such

SUMMARY 13 Management of depression Psychological treatment Exploratory therapy Cognitive-behavioural therapy Family/marital therapy

8

Physical treatment Anti-depressants ECT

as dichotomous thinking (seeing everything as black or white), catastrophization (making a drama out of every crisis - inability to buffer and cope with minor setbacks), and a cascade of negative thoughts about the self that are triggered by everyday events, minor losses or difficulties. These serve to reinforce a basic view of the world characterized by hopelessness about the future, helplessness in the present, and guilt and remorse about the past.

Management Mild depression This is best dealt with by psychotherapy, which may take one of three main forms. Exploratory and psychodynamic psychotherapy Here, the patient is encouraged to express depressive feelings, to vent anger and despair and - through the relationship with the therapist (who may be a GP, social worker, psychotherapist or physician) - to experience the responsive empathy and containment that was lacking in childhood and so to recognize and recapture some of the good experiences the depression has blotted out. As self-esteem improves, so the patient can begin to tackle some of his current difficulties, which in a depressed state seem overwhelming. Cognitive-behavioural therapy (CBT) and interpersonal therapy (IPT) Controlled trials have shown that CBT is as effective as antidepressants in the treatment of mild-to-moderate depression. CBT counteracts learned helplessness and negative depressive thinking by challenging the erroneous assumptions on which they are based (e.g. if I phone my boyfriend and he is out, it does not mean that he does not love me, merely that he is working late), and by considering positive features of the self. The patient is encouraged to set himself small, easily accomplished tasks and to feel pleased if he succeeds. IPT is similarly focused and time limited. It explicitly adopts a medical model, seeing depression as an illness over which the patient has little control. IPT homes in on one of four underlying factors that contribute to depression: 'role transitions', e.g. depression associated with retirement, or becoming a mother, or divorce and so becoming single; grief and bereavement linking depression with unworked-through losses; interpersonal conflict, usually with partners; and social skills deficits: here depression is linked with lack of personal effectiveness and competence, and therapy is specifically designed to improve communication and social flexibility.

221

Family and marital therapy Here the aim is to improve the marital relationship, which can then provide a bulwark against depression in response to loss. Partners often unwittingly contribute to the maintenance of a depressed state, perhaps by being overconcerned and 'helpful' and so reinforcing the sufferer's sense of their own helplessness and loss of role. The safe expression of anger can strengthen a relationship and reverse the inward turning of anger that is so common in depression. Moderate depression A combination of antidepressant drugs and psychotherapy can be more effective than either alone in treating moderate depression. Both depressed mood and biological symptoms improve with drugs; low self-esteem and social difficulties improve with psychotherapy. Antidepressants are always indicated when sleeplessness and guilt are prominent features. Tricyclic antidepressants (TCA) such as amitriptyline (75-200 mg/day) take 10-14 days to produce their full effect. Side-effects include parasympathomimetic symptoms of dry mouth and urinary retention, constipation, blurred vision and 'muzziness'; and cardiac arrhythmias can occur. Dosulepin (dothiepin) and lofepramine have fewer side-effects. Imipramine and clomipramine are less sedative. Tricyclics should be avoided in patients with prostatism or a history of heart disease. Mianserin, a tetracyclic, has fewer side-effects but can produce haematological reactions and is possibly also less effective in its antidepressant action. If mianserin is used, a blood count should be carried out every 4 weeks. Selective serotonin re uptake inhibitors (SSRIs) such asfluoxetine,fluvoxamine, sertraline and citralopine have fewer antimuscarinic side-effects and low cardiotoxicity, and are therefore preferable, especially when the suicide risk is strong. They may, however, cause nausea and vomiting. Venlafaxine has mixed SSRI and TCA properties and is another popular antidepressant. Monoamine oxidase inhibitors (MAOIs) such as phenelzine (15-90 mg daily) are still occasionally useful as second-line drugs, especially when phobic symptoms are present, but dietary restrictions (e.g. no cheese or yeast-extract) are needed because of harmful (hypertensive) interactions with tyramine-containing substances. Moclobemide is a newer MAOI that can be taken without any major dietary restrictions. Severe depression Depressed patients who are severely incapacitated, psychotic or actively suicidal require hospital admission. Antidepressants, which may have to be given in large doses (up to 250mg of amitriptyline or equivalent daily), are the first line of treatment.

222

ECT Despite debate about its mode of action and efficacy, and despite popular prejudice, controlled trials of electroconvulsive therapy (ECT) have shown it to be an effective and rapid treatment for depressive delusions and hallucinations (Fig. 8.4); it is also free from side-effects, apart

FIG. 8.4 Effect of ECT on severity of depression

from the dangers of the anaesthetic and some short-term memory impairment. ECT is a first-line treatment in depressive stupor or where the patient is refusing food and drink. When the patient refuses consent, in the UK it can only be given after Section 3 of the 1983 Mental Health Act has been applied and a second, independent, psychiatric opinion obtained, although in a life-threatening situation one application of ECT may be given pending the second opinion. ECT produces a rapid improvement in severe depression and is especially effective in elderly patients, in whom tricyclics may produce confusion and cardiac arrhythmias. The basis of ECT is the delivery of a convulsion or fit, and its effectiveness depends on an adequate fit of around 25 seconds. A 'course' of ECT, given twice or thrice weekly, should deliver a total of around 250 seconds of fit if relapse is to be avoided. ECT can be given to both hemispheres (bilateral) or to the non-dominant one (unilateral), depending on the position of the electrodes. Unilateral ECT produces less memory impairment but may also be less effective. ECT is very effective in the short run but, unless supported by psychological and pharmacological therapies, relapse is also common. Chronic or recurrent depression Most depressive illness has a tendency to remit, but relapse is common. Prolonged treatment with tricyclics has a prophylactic effect, as does 'maintenance' psychotherapy. Relapse of depressive psychosis may in some cases be prevented by prophylactic lithium carbonate. Chronic depressive states are rare but disabling, and prolonged attendance at a supportive day hospital may be required.

FURTHER READING ON DEPRESSION Brown G, Harris T 1978 The social origins of depression. Tavistock, London

TABLE 8.7 Clinical profiles of suicide and parasiticide Suicide

Parasuicide

Central theme

Despair, depression

Anger, frustration

Epidemiology

Increase in men aged 15-25, 1985-1999

Increase in all groups 1985-1999

8

Commoner in women aged Age, sex, class Commoner in age range 25-34; commoner in men 18-35 and in social than women; no class classes IV and V preference; recent increase in death by suicide in young men Social circumstances

Isolated, e.g. unemployed, retired, divorced, children left home, few friends or religious/social groups

Disturbed but not isolated, e.g. teenager unhappy with parents or boyfriends

Method and build-up

Planned. Usually leaves a note; method often violent, e.g. hanging, shooting, drowning

Impulsive. Commonest method is swallowing pills. Cutting (usually wrists) a more disturbed subgroup

Alcohol

May play a part - 'Dutch courage'

Attempt very commonly associated with alcohol/ drugs, i.e. reduced impulse control

Illness

90% of suicide case are mentally ill, usually with depressive illness. Physical illness also common

Only about 30% are mentally ill. Some have personality disorder. Majority normal or with 'minor affective illness'

Outcome

Not all 'succeed'. Survivors need hospital admission for treatment of mental illness

Not all survive, but most do. About 5% repeat. May benefit from counselling or psychotherapy

FIG. 8.5 Death rates for suicide and undetermined injury in England 1969-91

Lee A 1999 (ed) Affective and non-pychotic disorders. Gaskell, London Scott J 1995 Psychological treatments for depression. Br J Psychiatry 167: 289-292

SUICIDE AND PARASUICIDE Suicide appears to be an integral part of human psychology. There are few individuals lucky and happy enough never to have had even a fleeting thought of suicide at some point in their lives. There has been a worrying increase in suicides in young men (Fig. 8.5). 'Parasuicide', a syndrome of deliberate self-harm, appears to be a relatively recent phenomenon and has reached epidemic proportions over the past 25 years. It is the commonest cause of hospital admission for women under 65, and the second commonest for men. Many cases of deliberate self-harm go unreported; probably only about one-third reach hospital.

Clinical profiles The clinical profiles of suicide and parasuicide are, on the whole, distinct (Table 8.7). A person who commits suicide could be a male with a previous history of depression and alcoholism, recently divorced or made redundant, living alone and with few close friends or relations. He deliberately tells the neighbours and milkman that he is going away for the weekend, and then takes a bottle of aspirin or shoots himself. A parasuicidal person might be a woman in her late teens living with her mother and stepfather, and getting on badly with her boyfriend. After a row with him, in which both have been drinking, she goes home, swallows some of her mother's sleeping tablets, and is then noticed to be drowsy. When challenged about what has happened, she immediately 'confesses' and is rushed to hospital.

The central theme of suicide is despair and depression; in parasuicide it is anger and frustration. Although there are important differences there are also psychological similarities. Both involve a physical attack on self, on the body. In both cases the patient feels a combination of personal inadequacy and anger towards friends and family. This anger is directed inward, but there is often a wish to punish or harm those to whom the patient feels close. This is done indirectly via the suicide or suicide attempt. It is often said that parasuicide cases 'do not really mean it'. This is only partially true, and can sometimes be used by staff as justification for adopting a punitive attitude. It is best to assume that both suicide cases and parasuicide cases want both to die, and not to die. The relative proportion of the two feelings varies greatly, but both feel both. A third point of overlap follows from this. Not all suicide cases succeed and, conversely, it is possible in parasuicide to die by mistake. Patients who have made major suicide attempts may recover and, with the help of drugs, psy-

223

FIG. 8.7 Primary diagnosis of suicides (England and Wales Suicide Inquiry cases), 1996-2000 FIG. 8.6 Age and sex profile of general population suicides (England and Wales), 1996-2000

chotherapy and time, regain the will to live. In the majority of parasuicide cases the attempt resolves the difficult social situation that prompted it, and so is not repeated. In a minority - about 5% - the attempts are repeated, and a proportion of these will eventually 'succeed'. Parasuicide repeaters are more likely to have a personality disorder, to be addicted to drugs or alcohol, or to have been in trouble with the law. (See Figs 8.6 and 8.7.) 1

Management of 'failed suicide' The causes of suicide are essentially the causes of depressive illness, and the management of attempted suicide is that of prevention, and particularly the early detection, of depression. About half of successful suicide cases have contacted a helping agency - a counselling organization, a GP, social worker or priest - in the month before they kill themselves. It is important to ask troubled, unhappy or depressed patients about suicidal feelings. Fleeting suicidal thoughts - 'I would like to sleep' or 'I would like to be dead but would never have the courage to do anything' - are less worrying than a definite suicide plan. Talking about suicidal thoughts may in itself be a relief. Suicidal patients with symptoms of depressive illness should be admitted to a psychiatric unit. Patients who are managed as outpatients need to be seen frequently while they are in a suicidal state.

1

224

MCQ 8.2

Antidepressants can be prescribed, but only in small amounts, as hoarding of tablets can be a danger. Although the prevention of suicide remains the ideal, this is difficult to achieve in practice. Counselling organizations such as the Samaritans seem not to have had a major impact on overall suicide rates, but the incidence of suicide in the UK has fallen slightly over the past 30 years compared with the rest of Europe. Suicide is a reminder of the limitations of psychiatric and social care. When a patient who is receiving treatment commits suicide, it is especially stressful for the doctors and nurses involved, as it always is for the patient's family. Feelings of guilt and anger are inevitable. Suicide, like any death, is a reminder that doctors are not omnipotent. In dealing with suicide or the threat of it, it should be remembered that, ultimately, a person determined or ill enough can always find a way to end his life.

DELIBERATE SELF-HARM Causes The causes of attempted suicide cannot be found solely within the individual. Social factors that need to be considered in understanding the increase in parasuicide rates include: • Complex family structures without clear rules or roles, such as unsupported one-parent families; adolescents who want freedom but lack the emotional or financial resources to leave home; families with step-parents; • Drugs being a socially acceptable means of dealing with distress: the commonest tablets taken in parasuicide are mild analgesics and minor tranquillizers;

8

CASE STUDY 8.1 SUICIDE OF A 20-YEAR-OLD ART STUDENT A 20-year old art student was brought by her friends to the psychiatric ward in a highly aroused state. She had not slept for several days, was dishevelled, could not sit still, and was talking 'nineteen to the dozen'. She was convinced that she was a direct descendant of van Gogh and that her paintings were worth several million pounds. She came from a loving family but one in which there was a very high incidence of mental illness. Her mother suffered from alcoholism, her half-sister had committed suicide, and her father, a very successful businessman, had marked mood swings. A diagnosis of hypomania was made. She was treated with haloperidol and started on lithium carbonate. After 2 weeks in hospital she was well enough to return home, but not to resume her studies. She was followed in outpatients and offered fortnightly supportive psychotherapy in addition to her continuing drug therapy. She found the side-effects of the neuroleptic drugs troublesome, and often 'forgot' to take her haloperidol. Her mood continued to be somewhat elated, and she worried her parents by mixing with what they saw as a 'wrong' crowd, many of whom had also suffered psychiatric illnesses or abused drugs. She suffered an episode of depression when an exboyfriend killed himself while drunk, and had once more to be admitted to hospital for a brief period. Two years after her original admission she felt well enough to resume her studies at art college. Things seemed to be progressing well, and she reduced her contact with psychiatric services to a

3-4-monthly outpatient attendance. She wanted to come off all her drugs, and was only persuaded to continue with lithium with some difficulty. Then one weekend, apparently without warning, she was found dead in her flat having taken an overdose of aspirin. On reconstructing the preceding history it appeared that she had taken no lithium for several weeks prior to her tragic death. Lithium levels were supposedly checked by the student health service every 3 months, but she had failed her most recent appointment.

Questions 1. What was the diagnosis? 2. Could her suicide have been prevented?

Discussion She showed many of the features of manic-depressive psychosis, bipolar type. There was a strong family history of mood disorder. She showed features of hypomania on first admission, which responded to neuroleptics and a mood stabilizer. Later she developed depressive symptoms in response to a painful life event. The mortality rate from suicide in manic-depressive illness is around 10%, and there is some evidence that taking a mood stabilizer such as lithium reduces mortality. There was an administrative and ethical conflict between the patient's reasonable desire to lead as normal a life as possible and the fact that she was

An epidemic factor. Deliberate self-harm is often behaviour learned from friends, fellow patients, relatives or television.

Management The medical management (p. 23) of the parasuicide patient takes initial precedence over the social and psychiatric

suffering from a potentially lethal psychiatric illness. Insisting on regular contact with medical/psychiatric services - her GP, community psychiatric nurse or consultant psychiatrist - might well have meant that her deteriorating mental state and poor drug compliance would have been picked up. Even better would have been attendance at a 'lithium clinic' run as part of a mental health service. Ongoing psychotherapy could also have identified the recurrence of her depressive symptoms and her irrational view of herself as useless and a burden to everybody, which was revealed in her suicide note. But all of this is to be wise after the event. She appeared to be doing well, and there is a limit to the extent to which treatment can be imposed on someone who feels and seems to be functioning well. Her case illustrates one of the most poignant dilemmas of psychiatry. Severe mental illness affects young people and has a mortality comparable to that of diseases such as leukaemia. She survived for nearly 5 years after diagnosis, a result which would be considered quite good in physical medicine but can lead to a very pessimistic appraisal of psychiatric treatments. All this had to be discussed in the case review, which occurs whenever there is a suicide among patients known to psychiatric units - open meetings in which staff can express their feelings without fear of censure, and work through the inevitable emotions of guilt, failure, anger and despair that suicide can evoke.

assessment. Medical staff often feel angry with overdose patients compared to those whose illness is not selfinflicted and, like alcoholics and drug addicts, such patients may become scapegoats. Their irresponsibility affronts the overworked nurses and doctors, who are often similar in age and maturity to the patient. In addition, medical staff work in a stressful and hierarchical situation in which they can at times feel angry or frustrated, and these emotions

225

can be displaced into a punitive attitude towards parasuicide patients. This can be avoided if ward procedures give time for discussion of these patients and the feelings they arouse.

History A careful history must be taken from the overdose patient, who should be kept in hospital until alert and cooperative enough for this to be possible. The presence of depressive illness and suicidal rather than parasuicidal features must be identified. The risk of a repeat attempt has to be assessed (Table 8.8). Social isolation, marked sleep disturbance preceding the attempt, strong feelings of guilt or unworthiness, evidence of preplanning, inability to look to the future and previous psychiatric history - especially if there are drug or alcohol problems - all point to serious disturbance and should lead to referral to a psychiatrist. A detailed account of the attempt itself and of events surrounding it should be taken. It is often unhelpful to ask the patient 'why' he harmed himself. If the patient knew, he might not have had to do so. In attempted suicide one often finds an acting-out of feelings, rather than the capacity to contain or verbalize them. The patient will find it much easier to answer questions about 'what happened,

TABLE 8.8 Deliberate self-harm: assessment of severity of disturbance. Ideally, all eases should be referred for psychiatric assessment, Psychiatric referral is essential if any of the following are present Circumstance of the attempt Evidence of preplanning (e.g. suicide note, use of car exhaust) Violent method used (e.g. attempted hanging) Social circumstances of patient Alone Away from home Personal history of patient Recent loss (e.g. bereavement, divorce/separation, unemployment) Physical illness Previous attempts Alcohol or drug addiction History of depressive illness Family history of patient Suicide Depressive illness Mental state examination Sleep disturbance, low mood before attempt Low self-esteem, hopelessness Psychotic features (e.g. 1 deserve to die')

226

Future No relatives or other support Unresolved crisis Continued determined suicidal intent

and how'. This provides an 'action replay' of the attempt, which will contain the clues to the question of 'why'. Most commonly the patient feels angry but is unable to express this directly, usually for fear of rejection by a partner, spouse or parent on whom he feels dependent. As in depression, he dare not bite the hand that feeds, and so attacks himself instead. It can be useful to ask the patient who he thinks would have been most distressed if he had died. Later, it is often helpful to discuss alternative strategies with the patient, to examine what he could have done in the same situation other than attempt suicide. Followup counselling after discharge, either as an individual or in a family, by a social worker, psychiatrist or psychotherapist, may be offered, although less than half of patients accept such help. However, even without follow-up the prognosis of parasuicide, in the absence of depressive illness or personality disorder, is fairly good.

FURTHER READING ON SUICIDE AND PARASUICIDE Hawton K 1987 Assessment of suicidal risk. Br J Psychiatry 150: 145-153. Hawton K, Catalan J 1987 Attempted suicide: a practical guide to its nature and management. Oxford University Press, Oxford Owens et al 1999 Assessment of deliberate self-harm in adults. In: Lee A., ed. Affective and non-psychotic disorders. Gaskell, London.

MANIA AND HYPOMANIA Clinical features In depression, mood is abnormally lowered. In mania, or its milder form hypomania, it is abnormally elevated. In unipolar manic-depressive psychosis the patient may have recurrent depressive or recurrent manic episodes. If both manic and depressive episodes occur, the patient has bipolar manic-depressive psychosis. There are three main groups of symptoms in mania and hypomania: • Psychological symptoms. The patient feels elated, cheerful and expansive, and his thoughts are speeded up. He may develop grandiose delusions of wealth, potency and fame. He is full of optimistic (and unrealistic) plans for the future. • Physical symptoms. The patient is restless, sleeps little, eats voraciously or not at all, often loses weight, has poor concentration, and tries to do several things at once. • Behavioural symptoms. The patient often spends large amounts of money, dresses flamboyantly and may behave in an outrageous, disinhibited and promiscuous manner. Outbursts of irritability and aggression are common. Speech is rapid, may be incoherent, and flits from subject to subject (known as flight of ideas). The patient is often humorous in an irreverent and infectious way.

CASE STUDY 8.2 A 32-YEAR-OLD WOMAN WHO HAS TAKEN AN OVERDOSE A 30-year-old married woman was sent to A&E by the emergency GP service having been found by her husband in a comatose state with a bottle of Prozac beside her and smelling of alcohol. She was admitted to the medical ward, given a fluid diuresis, and made a full physical recovery within 12 hours. The duty psychiatric team was then called to evaluate her. She said that she had been feeling low for several weeks, with poor sleep and reduced appetite. She had had a difficult year, since her mother had been killed in a car accident, and her 9-year-old son had been diagnosed as suffering from hyperactivity attention deficit disorder and excluded from his school. Her husband, a builder, had been unemployed because of a bad back, and on the day of the overdose they had received a final demand for a gas bill which she could not pay. This felt like the final straw - she felt she could not cope, her husband was out at the pub with his friends, and her son would be better off without her. She and her husband had been getting on badly, with lots of rows, she had no libido and they had not had sex for several months. She had a history of puerperal depression after the birth of her son, and had taken a minor overdose in her teens following a row with her boyfriend. She was an only child. Her parents had separated when she was 13. She had stayed with her mother who soon remarried, but the patient had not got on with her stepfather. She saw her father infrequently and felt awkward with him. On examination she was a bedraggled, miserable-looking person, appearing older than her 30 years. She was overweight. History-taking was difficult and she made little eye

contact. Her speech was normal in form but slightly slowed. She was subjectively and objectively depressed. She felt guilty towards her family, feeling she had let them down and that she was in some way responsible for her son's problems. She had no delusions or hallucinations. Cognition was normal and she was of average intelligence. The interviewer felt a mixture of pity and irritation towards the patient. She said that she no longer felt suicidal, but was unable to visualize how things would be different if she returned home. On being offered admission to a psychiatric ward, she refused.

Questions 1. What is the diagnosis? 2. How serious is the risk of further self-harm or suicide? 3. Should she be compulsorily admitted to hospital?

Discussion The diagnosis of depressive illness is fairly clear in this case. She gives a history of low mood, poor appetite, loss of libido and difficulty in sleeping over several weeks, culminating in an overdose of tablets with suicidal intent. She felt guilty, unattractive and useless. This was not a purely situational episode of deliberate selfharm, although there was an impulsive element to it, and there was no evidence of preplanning and she had not written a suicide note. The depression was of moderate severity, but there were no psychotic features. She claimed that her suicidal feelings had resolved, but was unable to give

8

a convincing account of how things could change in her life. Her home support was problematic. There was a definite, but not extreme, continuing suicide risk. Compulsory admission is a grave step, especially in the absence of psychotic features, but to let her go home might be to court disaster. During these deliberations her husband appeared on the ward. He was interviewed and revealed himself to be a reasonable man, desperately worried about his wife and shocked by her suicide attempt. He admitted that he too had felt depressed and had tended to resort to alcohol and avoid his home problems by going to the pub with his friends. He was strongly opposed to compulsory admission and said that he would guarantee to keep a close eye on his wife. Compulsory admission requires the consent of the nearest relative, unless that power is taken away by applying to a local magistrate - an extreme measure. A further attempt to persuade the patient to stay in hospital was made, and she reluctantly agree to stay voluntarily. While on the ward she was started on a more sedative antidepressant, sertraline, joined a self-esteem and assertiveness group, and she and her husband attended marital therapy sessions, in which her anger about the way in which she felt he had abandoned her was clearly expressed. He accepted the legitimacy of this and apologized. She could link this with her feelings about her father, who had abandoned her after her parents' divorce, and began to see that, despite his faults, her husband was a loyal and loving man. Her suicidal feelings lifted, she began to sleep better and was discharged from hospital, although therapy continued.

227

Hypomania frequently coexists with depressive symptoms: the patient laughs and cries almost simultaneously. This is known as a mixed affective state. Hypomanic episodes may, like depression, be triggered by a painful rejection or loss. This has led to the concept of manic defence, whereby the patient escapes into mania as a way of avoiding unhappy or sad feelings.

Management A hypomanic patient can cause havoc to both himself and his family. Hospital admission is often indicated but may be hard to achieve voluntarily, as the patient denies illness, often claiming to have never felt better, or to have discovered the secret of life. Immediate management is with tranquillizers such as chlorpromazine (100-600 mg daily) or haloperidol (15-60mg daily). Subsequent relapses may be prevented by prophylactic lithium carbonate (800-1600 mg daily). Lithium also has a mild immediate antimanic effect. Blood levels should be monitored regularly and be in the range 0.8-1.5mmol/L. Lithium may produce hypothyroidism and renal damage, leading to nephrogenic diabetes insipidus, and thyroid and renal function should be monitored. Other side-effects include abdominal pain, fine tremor, subjective memory impairment, thirst and lowered zest for life. Patients should be encouraged to keep up fluid intake in hot weather. Anticonvulsants such as carbamazepine (400-1600 mg daily) can be used as second-line antimanic prophylactics if lithium is ineffective or cannot be tolerated.

SCHIZOPHRENIA Just under 1 % of the population develop schizophrenia at some point in their lives, a statistic that appears to hold true for all cultures and countries. Chronic major mental illness in people under 65 is due mainly to schizophrenia. Until the move towards community care began, about half of all hospital beds in the UK were in mental hospitals and the majority of these (for the under-65s) were occupied by patients with a schizophrenic illness. As the number of hospital beds has fallen, a significant proportion of the vagrants living in large inner-city hostels and prison recidivists are found to be suffering from schizophrenia.

Aetiology Schizophrenia is the core disorder in adult psychiatry. Much debate has raged about the nature of schizophrenia, or even whether it is a valid clinical entity. It is probable

1

228

MCQ 8.3

Chances of developing schizophrenia (%) Relationship Parent: one Parents: both Sibs (neither parent schizophrenic) Sibs (one parent schizophrenic) Children Half sibs Uncles and aunts Nephews and nieces Grandchildren First cousins Fraternal twins Identical twins FIG. 8.8 Lifetime expectation of schizophrenia in relatives of schizophrenics

that schizophrenia is not a single disease but a group of related conditions. Both genetic and environmental factors are important in the aetiology of schizophrenia. It is best seen as a developmental disorder of the brain to which genetic and perinatal factors (birth trauma or maternal viral infection) contribute, but manifesting itself in late adolescence when brain maturation is finally completed. Pathogenic environments contribute to relapse in schizophrenia, but their role in aetiology is uncertain. It is possible that abnormal brain development may result from unstimulating or traumatic environments. • Twin and adoption studies have demonstrated the role of heredity in schizophrenia (Fig. 8.8). The concordance rate for identical twins is about 50%, that for nonidentical twins around 15%. • The efficacy of antipsychotic drugs such as chlorpromazine, with its inevitable parkinsonian side-effects, suggests a disturbance of midbrain dopaminergic pathways in schizophrenia (the 'dopamine' hypothesis). There is convincing evidence of increased D2 activity in patients with positive symptoms of schizophrenia. Dopamine abnormality may be the end result of other metabolic aberrations. It seems likely that serotonin (5HT) is also involved, perhaps by increasing D2 levels. Also glutamate, an excitatory neurotransmitter, may behave abnormally, leading to underactivity of Nmethyl-D-aspartate, which in turn may stimulate abnormal dopamine and 5HT activity. • A subgroup of patients show ventricular abnormalities on CT scan. There is slight ventricular enlargement, a reduction in the normal asymmetry between the cerebral hemispheres and, particularly in males, a reduction in cortical grey matter, especially in the temporal lobes. It has been suggested that structural abnormalities are more common in patients who lack a family history of the illness, suggesting that non-genetic organic factors, perhaps associated with perinatal brain injury, may be important in some cases.

SUMMARY 14 Schneider's first-rank symptoms

TABLE 8.9 Relapse rate for schizophrenia

• Delusional perception • Auditory hallucinations Audible thoughts (thought echo) Voices arguing or discussing ('third person') Voices commenting on the patient's actions • Thought disorder Thought withdrawal Thought insertion Thought broadcasting • Passivity experiences: delusions of control 'Made' feelings (includes somatic hallucinations) 'Made' actions Somatic passivity (body invaded from outside)

A. At 9 months

• Family and social factors have also been shown to be important, although not pathognomonic. In established schizophrenia the relapse rate is high when the patient is living in an over-involved family where there is a negatively charged emotional atmosphere (high 'expressed emotion', or 'EE') (Table 8.9). • Schizophrenic patients have a high state of psychophysiological arousal, possibly related to family stress. This may reflect limbic system dysfunction and lead to a difficulty in processing sensory stimuli.

8

Relapse rate 50% 12%

High EE homes Low EE homes B. In high EE homes Face-to-face contact

Maintenance therapy

Relapse rate

High High Low Low

No Yes No Yes

92% 53% 42% 15%

Delusions In schizophrenia these include the feeling that one's thoughts originate from outside ('thought insertion'), that they are being interfered with, removed ('thought block' and 'thought withdrawal') or transmitted to outsiders as though on a loudspeaker ('thought broadcasting'). 'Passivity feelings' are the feeling that one's thoughts or actions are 'made' from outside.

Delusional perception Clinical features Schizophrenia usually starts in young adulthood (the exception being paraphrenia, a form of paranoid schizophrenia affecting the elderly). There is a slight preponderance of men. The features of the illness can be divided into two parts: the effects on mental processes, and the effects on social functioning.

Mental disturbance There is a general disturbance of mental functioning in schizophrenia. The normal progression of logical thought is disrupted ('thought disorder'). The privacy of the self is breached. Certain patterns of disorganization of thinking are characteristic of schizophrenia, and are known as Schneider's first-rank symptoms.

Hallucinations These are usually auditory. Characteristic of schizophrenia are 'third-person' hallucinations, in which more than one person is heard discussing the patient and referring to him as 'he', 'she' or 'it'. Voices commenting on a person's actions like a 'running commentary' are also characteristic, as are voices that echo the patient's thoughts. Tactile or 'kinaesthetic' hallucinations, for example electric-shock feelings in the limbs, are rare but pathognomonic.

In delusional perception, a neutral stimulus (e.g. a car number plate or a traffic light) suddenly acquires special and often frightening significance for the patient. 1

SUMMARY 15 Management of schizophrenia • First attack: admit to hospital or day hospital for assessment, confirm diagnosis, establish contact and initiate treatment. • Treat symptoms with neuroleptics. Chlorpromazine 100mgt.d.s. by mouth Trifluoperazine 5mg t.d.s. by mouth Risperidone 2-4 mg b.d. by mouth Examples of Olanzapine 5-10 mg b.d. by mouth moderate moderatedoses doses Flupentixol ('Depixol') 40 mg 2-weekly Fluphenazine ('Modecate') 25 mg 2-weekly Treat side-effects as necessary with antiparkinsonian agents, e.g. Kemadrin 5 mg t.d.s. • Assign 'community key worker' as part of community care programme who will offer supportive psychotherapy and coordinate care package. • Assess family situation: offer psychoeducational programme to lower 'EE' if indicated. • Rehabilitation: attend day hospital, day centre, sheltered workshop, live-in hostel, sheltered housing. • Where resistant to medication consider newer neuroleptics: clozapine and risperidone.

229

Social deterioration The second main feature of schizophrenia is social deterioration. This is known as a negative feature of the illness, as opposed to the positive features of thought disorder, delusions and hallucinations. The patient may withdraw from social contact, give up his job, shun his friends and family, and spend many hours in isolation.

Differential diagnosis

Antipsychotic drugs These are of two main types: the so-called 'typical' antipsychotics, and the newer 'atypicals', orginally developed in the search for drugs with fewer extrapyrimidal side-effects (EPSs), which are rapidly becoming standard treatments. 'Typicals' can be divided into three main groups. All are potent D2 receptor antagonists.

An important subtype of schizophrenia is paranoid schizophrenia, characterized by paranoid delusions. Here the onset is later (around 30-40 years of age), social deterioration is much less marked, and the personality is relatively well preserved. Schizoaffective disorder has features of both schizophrenia and an affective illness, e.g. first-rank symptoms plus considerable depression. Its diagnosis is intermediate between that of the two 'parent' disorders. Not all madness is schizophrenia. The differential diagnosis includes:

• A group characterized by profound sedative effects and moderate antimuscarinic and extrapyramidal effects, e.g. chlorpromazine (50-1000 mg daily); • Moderate sedative effects, marked antimuscarinic effects, but fewer extrapyramidal effects, e.g. thioridazine (25250 mg daily), often used as a sedative in the elderly; • Fewer sedative and antimuscarinic effects but more pronounced extrapyramidal effects, e.g. trifluoperazine (2-30 mg daily); haloperidol (2-80 mg daily); flupentixol (2-100 mg daily).

• Organic psychosis, such as acute confusional states and drug- and alcohol-related psychosis; • Manic or depressive psychosis (in which auditory hallucinations occur but are more often 'second-person', in which voices speak directly to the patient); • Hysteria; • Stress-induced or 'psychogenic' psychosis; • Severe personality disorder.

Atypical' antipsychotics tend to have strong 5HT2a receptor antagonist activity. They include:

Prognosis The course of schizophrenia is very variable. About 30% of patients have only one episode. A good prognosis is more likely if the psychosis has an acute onset, a clear precipitant, florid symptoms, marked mood change in addition to disturbance of thinking, and good previous social adjustment and personality. At the other end of the scale, about 15% remain severely disabled and will still be in institutional care after a year. In these patients the negative features of schizophrenia (withdrawal and inertia) may predominate. In the middle group, representing the majority of patients, the illness has a fluctuating course but relapse and some residual disability are likely to occur.

Management Acute schizophrenic symptoms respond well to antipsychotic medication. There is some evidence that prompt treatment is associated with a good prognosis. The patient will normally be admitted to a psychiatric unit and treated initially with 100-1000 mg daily of chlorpromazine or its

1 230

equivalent. The longer-term management of schizophrenia is difficult. It involves both prevention of relapse and management of chronic disability.

MCQ 8.4

• Clozapine. This has serotonergic and dopaminergic properties and is effective against positive and negative symptoms of schizophrenia which have failed to respond to typicals. Its main drawback is that it causes agranulocytosis in 3% of subjects, and weekly blood count monitoring is necessary. Excessive salivation and sedation can also be problematic. • Risperidone. Also effective against positive symptoms, but can cause agitation and insomnia. • Olanzapine and quetiapine (an anagram of quiet-apine!, subtly drawing attention to its sedative properties) are also effective drugs and generally well tolerated, although they can cause alarming weight gain. A wide range of side-effects and idiosyncratic reactions may be produced with the antipsychotic drugs (Table 8.10). Perhaps the most serious of these is the rare but potentially fatal neuroleptic malignant syndrome, characterized by hyperthermia, confusion, muscular rigidity and autonomic disequilibrium (tachycardia, labile blood pressure, sweating). The patient should be admitted immediately to a medical unit; all antipsychotics should be discontinued and bromocriptine and dantroline given. Psychosocial interventions Analytical psychotherapy is generally not favoured in schizophrenia, and indeed the emotional arousal it may evoke can cause deterioration. There are, however, a number of 'psychosocial interventions' that significantly improve the wellbeing of patients with this disease. These comprise: • Family intervention. Because high 'EE' is associated with relapse, measures to reduce EE, either in families or in the hostels where some patients with schizophrenia live,

TABLE 8.10 Side-effects and toxic effects of antipsychotic drugs (e.g. chlorpromazine, flupentixol) Type of effect

Examples

Autonomic (antimuscarinic)

Blurred vision Dry mouth Constipation Precipitation of glaucoma Postural hypotension Inhibition of ejaculation Urinary hesitancy Akinesia (Parkinson-like) Acute dystonia Tardive dyskinesia (involuntary repetitive movements) Akathisia (motor restlessness) Weight gain Galactorrhoea and amenorrhoea Gynaecomastia Agranulocytosis Cholestatic jaundice Skin rashes

Extrapyramidal

Endocrine effects

Hypersensitivity reactions

Neuroleptic malignant syndrome Provocation of epileptic fits Hypothermia (especially in the elderly) Skin photosensitivity

will be valuable. A programme of 'psychoeducation' for families teaches them about the nature of the disorder and, via family group therapy or regular therapy sessions with families, works to reduce the emotional tension and foster a more benign atmosphere. Long-term supportive psychotherapy. Therapy in schizophrenia has to be long-term: perceptible change may take 4 years or more. A stable consistent therapeutic relationship with a psychiatrist or community key worker is needed and often bears fruit. Cognitive-behavioural therapy (see Recent Advances box). 1

Management of chronic disability Chronic mentally ill patients are severely handicapped. Their prospects of employment are poor, they may have little or no family support, and their capacity to care for and occupy themselves may be very limited. They are sometimes also shunned or made scapegoats by society, which tends to fear and despise madness. For at least 150 years these problems were dealt with by removing patients from society and locking them away in mental hospitals, thereby reinforcing the helplessness and isolation produced by the illness. Because of this, and also because mental hospitals are expensive to run and often sited on valuable land, government policy in the UK has now shifted away from the mental hospitals and towards community care. Patients are supposed to be cared for within the local community, in a network of hostels, sheltered houses, day hospitals and day centres, and community mental health centres. These are run by a locally based team of professionals, the 'multidisciplinary team' including community psychiatric nurses (CPNs), social workers and general practitioners, as well as psychiatrists. However, although this pattern of care is desirable for many patients, and is already a reality for some, it is not necessarily suitable for all patients, some of whom continue to need highly staffed or secure facilities on an intermittent or continuous basis. The care programme approach aims - through ensuring that every patient has an identified key worker, a definite 'care programme' and regular reviews with all concerned - to prevent patients from drifting away from services and thus to prevent the relapse that is likely to ensue when continuity of care is lost.

8

FURTHER READING ON SCHIZOPHRENIA Drug and Therapeutics Bulletin 1996 Drug treatment of patients with schizophrenia. Drug Ther Bull 33:81-86 Kane J, McGlashan T 1995 Treatment of schizophrenia. Lancet 346:820-825 Lewis S, Buchanan R 1998 Schizophrenia. Health Press, Oxford

PARANOID STATES Long-term maintenance on medication (which in the case of 'typicals' can sometimes be given in the form of a weekly or fortnightly injection because of problems of compliance) prevents relapse in schizophrenia. Phenothiazines can, however, have unpleasant and serious side-effects, especially a feeling of tiredness and extrapyramidal syndromes of stiffness and involuntary movements (Table 8.9). Antiparkinsonian drugs such as orphenadrine (100 mg t.d.s.) or Kemadrin (5 mg t.d.s.) should be given to counteract these. Reducing family tension has also been shown to prevent relapse. This can be achieved by family counselling sessions and by physical separation of the patient from an over-involved family through the provision of hostel accommodation and/or attendance at a day centre.

Paranoia means literally 'to be beside oneself, or 'out of one's mind'. In psychiatry, it refers to a psychosis whose

SUMMARY 16 Conditions resulting in paranoid states • • • • • •

Organic psychosis Manic-depressive psychosis Paranoid schizophrenia Paranoid psychosis Paranoid personality Paranoia in the elderly

231

RECENT ADVANCES IN THE PHYSHOSOCIAL MANAGEMENT OF SCHIZOPHRENIA Schizophrenia remains psychiatry's greatest challenge. Although the essential nature of the schizophrenias remains as elusive as ever, the past few years have seen, for the first time for two decades, significant psychotherapeutic advances. There is evidence that the longer the time between the onset of symptoms and beginning treatment, the worse the prognosis. This has led to much current interest in the notion of 'early intervention' in schizophrenia - trying to pick up the prodromal stages of the illness and initiating treatment, both psychosocial and pharmacological, as early as possible.

Psychosocial treatment of schizophrenia Exciting results have recently been claimed for cognitive-behavioural treatment in schizophrenia. Contrary to earlier views, it seems that patients can develop some control and mastery over their symptoms, especially abnormal beliefs and auditory hallucinations. Patients can, for example, distract themselves from voices by concentrating on external tasks, by playing music on 'Walkman' headphones, or even by 'bargaining' with the voices so that they confine 'their' attentions to particular times of the day. With the help of a therapist, or even a self-help 'voices' group, patients can also be encouraged to test the validity of delusions, e.g. 'if I go into that shop I will be killed', and on the basis of the 'results' of these 'tests', modify their beliefs. Early studies of these approaches are encouraging in producing, for some patients at least, reduced symptoms, improved

central theme is a feeling of persecution and fear of being attacked. Paranoid states are common in general medical and hospital settings. A postoperative patient may suddenly rip out an intravenous cannula and attack one of the nurses. A patient may refuse all drugs, convinced that the staff are torturers dressed up as doctors and nurses. The conditions resulting in paranoid states are as follows: • Organic psychosis. This is by far the commonest cause in hospital settings. Paranoid delusions are frequently found in patients with acute confusional states. Any of the causes of organic psychosis (p. 208) may be responsible, with steroid-induced psychosis, alcohol withdrawal syndrome and amphetamine psychosis being particu-

1

232

Case 8.1

social functioning and compliance, and reduced drug dosages.

Compliance therapy A major problem in schizophrenia is drug noncompliance. Borrowing on the techniques of 'motivational interviewing' developed for addiction, researchers have shown that compliance can be dramatically improved if the clinician provides much greater information about the drugs, and attends closely to patients' feelings about side-effects and the stigma associated with taking long-term antipsychotic medication. A simple useful intervention appears to be to compare the schizophrenia sufferer's need for medication with the diabetic's need for insulin. From this perspective schizophrenia becomes a disability to be lived with, rather than a disease which is either denied or devastating. Assertive outreach programmes aim to tackle non-compliance by treating the patient in his or her home environment and by offering intensive therapeutic nursing, family and practical help at a time of maximum stress.

Early intervention The prognosis in schizophrenia is worse the longer the gap between the onset of symptoms and starting therapy. Early intervention programmes aim to identify and treat young schizophrenia sufferers using a combination of destigmatization, drug therapy, combating concomitant substance misuse ('dual diagnosis') and psychosocial intervention.

larly common. Alcoholism may also produce a chronic paranoid state in which auditory hallucinations predominate (p. 251). Manic-depressive psychosis. Feelings of being disliked and shunned are common in depression, and may assume psychotic proportions if the patient feels he is so wicked that he is deservedly in danger of attack. Paranoid ideas can occur in mania, but elation and overactivity are usually more prominent features. Paranoid schizophrenia (see p. 229). Paranoid psychosis. Sometimes the patient develops paranoid delusions and hallucinations without the full range of features of schizophrenia. There is often a sexual content to the psychosis. The patient may, for example, become convinced that his spouse is having an affair (morbid jealousy, or Othello syndrome). The feeling that workmates or strangers are accusing the patient of homosexuality is often a prominent feature of paranoid psychosis. Paranoid personality. People with paranoid personalities have a tendency to mistrust and blame others, and so

frequently become involved in complaints and battles with authority. The patient is often sensitive, isolated, insecure and quick to take offence, and misinterprets the behaviour of others as hostile. When his fragile self-esteem is threatened by illness or rejection, he may develop a paranoid psychosis. • Paranoia in the elderly. Elderly patients may develop a paranoid psychosis (paraphrenid) with schizophrenic features. Deafness, social isolation and previous paranoid personality are important predisposing factors. Common to all the different forms of paranoia are isolation, insecurity, mistrust and misperception. In some cases the defence of projection seems to be operating. Here an inner state of persecution is transferred on to the external world, presumably as a defensive manoeuvre designed to reduce inner mental tension. Thus a cancer patient may develop a paranoid state in which he denies there is anything physically wrong, but refuses to leave hospital on the grounds that he is in danger of being attacked. Here the 'hostility' felt from the cancer within is attributed to the outside world instead. Paranoid states can be treated symptomatically with phenothiazines (see p. 230), but the underlying cause itself should be treated whenever possible.

SOMATOFORM DISORDERS Somatoform disorders are conditions in which (a) physical symptoms are present for which no organic or physiological basis can be found, and (b) where there are demonstrable underlying psychological factors or conflicts, i.e. the physical symptoms are manifestations of underlying psychological distress or disease.

Terminology and definitions There are several terms commonly used to describe physical disorder in which psychological mechanisms may be important. The outmoded but still widely used term psychosomatic illness refers to conditions in which there is definite structural physical illness (e.g. bowel inflammation in ulcerative colitis) but whose precipitant or maintaining factors are thought to be in part psychological. The term functional is sometimes used to describe symptoms produced by changes in organ function in the absence of structural change. Functional illness may or may not be psychological in origin. Functional bowel disease (irritable bowel syndrome, see p. 817), for example, may result from faulty diet or disturbance of the autonomic system (potentially stress related) or both. Controlled trials suggest that chronic intractable irritable bowel syndrome responds well to psychotherapy. The term psycho genie usually implies that the symptoms are entirely psychological in origin. Many of the ill-defined

conditions seen in general practice and medical outpatient clinics could be labelled psychogenic, although this term may obscure the fact that physical factors (e.g. muscular tension) are also important. The concept of hysteria (p. 235) has a long history and has become differentiated from the more non-specific psychogenic symptoms. Hysteria is divided into two types:

8

• Conversion disorders. These are psychogenic illnesses often simulating neurological disorders - based on unconscious mechanisms, in which mental conflict is avoided both by 'converting' anxiety into physical symptoms (hence conversion hysteria) and by dissociation (i.e. a split between the mental self and the physical self). • Somatization disorders. These are polysymptomatic disorders in which the patient experiences multiple unexplained symptoms without physical disease and in the context of a disturbed personality. 1 In psychosomatic illness psychological factors aggravate, prolong or cause relapses in an illness whose basis is physical (just as a poor ward atmosphere has been shown to delay postoperative wound healing). Psychogenic symptoms are often based on circular patterns in which, for example, emotional conflict produces muscular tension, which results in pain, which leads to greater emotional stress, or anxiety (e.g. fear that the pain may be due to cancer), which results in more muscular tension and so on. Chronic fatigue syndrome (CFS, see p. 236) can be thought of as a somatoform disorder; some however would classify it as a psychosomatic illness, in which psychological factors exacerbate an organic illness, albeit one whose pathology is currently unknown.

MIND-BODY MODELS There are two extreme but widely held views on the nature of psychosomatic illness. The first, corresponding to the philosophical position of materialism, holds that all phenomena are physical in origin and denies the existence of the mind. For the materialist there are no psychogenic conditions, only illness whose physical cause has yet to be determined. The second, idealist, view holds that psychosomatic illness is mental in origin. There are problems with both these models. The materialist might point to the numbers of patients labelled as hysterical who have subsequently died of carcinoma or multiple sclerosis; but the existence of mind is undeniable. The idealist, on the other hand, tends to confuse the symbolic meaning of an illness with its cause, claiming, for example, that a man who developed oesophageal cancer did so because he could not swallow his wife's infidelity. A classic error of this type was the pre-1956 theory of Down's syndrome. This held that Down's syndrome was caused by physical or mental trauma in pregnancy, as mothers of Down's syndrome babies regularly reported such traumatic incidents. Only when the chromosomal abnormality

233

234

responsible was found was it realized that the trauma theory was a retrospective rationalization rather than a causal explanation. In reality, the relationship between mind and body is far more complicated than either the pure materialist or the idealist model allows. The interesting questions concern the interactions between the mental and the physical, just as many important problems in genetics involve interactions between heredity and environment. For example, there is good evidence that psychological adversity in childhood - neglect or abuse - is associated with poor physical health in adult life. The mechanisms by which this is mediated remain to be elucidated. To some extent they may result from excessive alcohol and drug abuse, and smoking, which are associated with both psychological stress and poor physical health. The effects of stress via the hormonal system on immune mechanisms may also play a part. Thus the 'Descartian' split between mind and body is beginning to be seriously challenged. It has been cogently argued that we 'think with our bodies': that the body is more than just a metaphor for much emotional experience (we speak of 'gut feelings' and 'emotional pain'), but that we use bodily sensations to help us choose between different equally rational courses of action. Body and mind, reason and emotion are ultimately inseparable.

more likely to develop coronary artery disease than those with more placid temperaments. This is particularly true where aggression plays a large part in the 'type A' makeup. Similarly, the prognosis of breast cancer may depend in part on personality: women who are emotionally inhibited appear to have, on average, a less good prognosis than those more able to express anger and sadness. The intervening variables that link the psychological to the physical are largely unknown. Autonomic and corticosteroid responses to stress play their part, as do changes in the immune system (especially T lymphocytes), but the precise pathways and links remain to be elucidated. Current evidence does not support the view that psychosomatic diseases are primarily psychological in origin, nor that any specific relationship exists between personality type and a particular disease or target organ.

Experimental evidence

Development of the body image

There is evidence that painful events in a person's life can precipitate physical illness. The link between stress and myocardial infarction, for example, is well documented (Fig. 8.9). Bowel disease, such as ulcerative colitis or even appendicitis, may also be triggered by a preceding painful or stressful life event. Personality plays an important part in determining the impact of such events. Individuals with 'type A personality' (competitive, driving) are, other factors being equal,

In adult life mind and body are separate realms. The body is experienced as an object among other objects in the world, but is also, for its 'owner', the source of knowledge and sensation and the means by which contact and communication with others is possible. The mind, though experienced as a separate domain, is indissolubly connected to the body. As Freud pointed out, it only takes a toothache to feel that 'the molar hole contains my soul'. The adult state in which the self comfortably inhabits both mind and body, experiencing each as separate and yet connected, is a complex product of health, maturation and culture. The infant is born in a state of psychosomatic unity: pain for a young child is both mental and physical. In adult life physical illness may produce regression to a childlike state. Childhood experiences may play a crucial part in determining later relationships between mind and body. In early childhood, the parents' love and care for the child and his body sow the seed of healthy self-love in later life. A girl who was 'meant' to be a boy may have to struggle with the effects of her parents' disappointment for years to come. Later, the adolescent has to learn to accept his or her body and sexuality. Self-acceptance is an essential precondition of loving others. When this process is disturbed, some individuals may, when anxious, ask their doctor for an examination, complaining that there is something wrong with their body or even asking for it to be changed by plastic surgery. The patient is looking for the acceptance and care he feels he missed from his parents. If external relationships fail, the patient may turn for interest and gratification to his

FIG. 8.9 Experience of potentially stressful life events by myocardial infarction patients in the 12 weeks before Ml occurred

PSYCHOLOGICAL APPROACHES TO SOMATOFORM DISORDERS Psychiatric understanding of psychosomatic and psychogenie illness can be considered under two headings: the development of the body image, and repression and the unconscious.

own body, and this too can lead to development of, and preoccupation with, psychogenic physical symptoms.

Repression and the unconscious Repression is the active removal from consciousness of painful or disturbing thoughts, feelings or memories. The mind's eye is blinded to problems and conflicts. The difficulties and the feelings they evoke remain, however, even though the patient is unconscious of them. They may then manifest themselves in physical symptoms, by the process of displacement. Thus a 20-year-old student presented with a hysterical hemiplegia and unexplained fits. She described her family life as happy and uneventful. In the course of therapy it emerged that her father had had a disabling stroke when she was 10, and that a fellow student had attempted to rape her a few weeks before her hemiplegia presented. She had 'forgotten' both of these painful events, but the neurological aspect of her presenting symptoms were reminiscent of her father's disability, suggesting that at some level they were still active in her mind, albeit at an unconscious level. The trauma of the rape re-evoked the shock, terror and vulnerability she had experienced when her father became ill.

'MINOR' PSYCHIATRIC ILLNESS Minor psychiatric illnesses are hard to classify and often go undetected, hiding behind and contributing to physical symptoms (Tables 8.11 and 8.12). In these illnesses a somatic 'facade' conceals difficulty in adjustment to social stress ('adjustment disorders'), symptoms of anxiety and/ or depression, or personality difficulties. Once these are explored and the patient reassured, the physical symptoms may be relieved. In considering patients with somatization disorders, a number of psychosocial factors should be explored: • Stress factors, especially recent losses, such as bereavement, unemployment, children leaving home, financial difficulties, divorce and termination of pregnancy; • Social support, relationship with partner (including sexual relationship), parents, children and friends, housing conditions; • Personality: how the person has reacted to stress in the past, previous stressful or painful experiences, especially in childhood, and whether the patient is able to respond with appropriate anger or sadness; • Alcoholism, drug abuse and eating disorders, which frequently present indirectly via physical symptoms.

FURTHER READING ON SOMATOFORM DISORDERS Goldberg D, Huxley P 1980 Mental illness in the community - the pathway to psychiatric care. Tavistock, London Lloyd G 1991 Textbook of general hospital psychiatry. Churchill Livingstone, Edinburgh

TABLE 8.11 Characteristics typical of doctors who frequently detect psychological problems • • • • • • •

8

Show empathy Sensitive to emotional cues Use appropriate psychiatric questions and probes Tend to ask for clarification of the patient's complaints Make early eye-contact with the patient Do not bury themselves in notes Good at dealing with interruptions and garrulous patients

TABLE 8.12 Psychiatric illness in general practice: how much goes undetected? Category

All patients

Probability of detection by GP

Physical illness without psychiatric disorder Physical illness and unrelated psychiatric disorder Physical illness with secondary psychiatric disorder

67%

not available

8%

19%

1%

74%

19% 13.5%

47% 33%

5.5%

85%

5%

95%

Somatization Physical illness with symptoms exacerbated or caused by psychiatric disorder Somatized psychiatric illness (no diagnosable physical illness) Psychiatric disorders presenting with psychological symptoms

Smith G 1999 Managing somatoform disorders. In: Lee A (ed) Affective and non-psychotic disorders. Gaskell, London Wilson A, Hickie I, Lloyd P 1994 A treatment of chronic fatigue syndrome: science and speculation. Am J Med 96:544-550

HYSTERIA The essential features of hysteria are: (a) physical symptoms, (b) absence of organic pathology, and (c) psychological conflict.

Classification and clinical features There are three main patterns of hysteria: monosymptomatic, polysymptomatic and epidemic. Monosymptomatic hysteria (conversion disorder) Typically there is gross loss of neurological function. This may involve: • the limbs, e.g. paralysis, anaesthesia or dystonia

235

RECENT ADVANCES IN CHRONIC FATIGUE SYNDROME (CFS) Patients suffering from feelings of chronic physical exhaustion frequently present to physicians and sometimes to psychiatrists. Although there may be nothing new in this - such patients were formerly seen as suffering from 'neurasthenia' - much recent effort has been devoted to the clarification of diagnostic criteria, the search for possible causes, and developing structured treatment approaches. The criteria for CFS include disabling physical fatigue which is unexplained by primary physical or psychiatric diagnoses, myalgia, depression and poor concentration, all lasting for more than 6 months. CFS subsumes other diagnostic labels such as Royal Free disease, postviral syndrome and, where there is or has been definite evidence of infection, postinfectious fatigue syndrome. There is no generally accepted theory about causation. Enteroviruses have been implicated, and increased litres of anti-Coxsackie B IgM are reported; muscle fibre damage is found in some cases; some have claimed deficiency of intracellular magnesium. In contrast to the metabolic explanations, psychiatric intervention shows much promise. Fifty per cent of CFS sufferers meet the criteria for major depression. Whether this is cause or

consequence is unclear: one study found that 100% of patients who lacked evidence of infection were depressed, suggesting that 'CFS' represents a number of disorders of differing aetiology. Management includes:

• the higher functions, e.g. hysterical fits, hysterical blindness, fugue states (in which the patient has complete memory loss), or even hysterical psychosis (where the patient develops a short-lived and dramatic psychosis in response to external stress).

with monosymptomatic hysteria, their prognosis is poor and is related to general personality difficulties rather than a specific conflict.

The pattern of dysfunction follows the patient's 'idea' of how the body works, rather than anatomy. Similarly, in hysterical psychosis the patient behaves in the way he thinks a madman ought to behave, rather than demonstrating the specific symptoms of schizophrenia. A neurologist can usually determine the integrity of the nervous system, either by subtle physical examination (extensor activity in apparently paralysed limbs) or by investigation (intact visual evoked potentials in hysterical blindness). The patient may also exhibit the belle indifference described by Janet, i.e. an apparent lack of anxiety in the face of a crippling disorder. Psychophysiological measures of arousal show that hysterical patients are in fact often highly aroused, despite the absence of manifest anxiety.

236

Polysymptomatic hysteria (Briquet's syndrome, somatizotion disorder) Patients suffering from polysymptomatic hysteria (usually women in their 30s) develop a wide variety of physical complaints with no organic basis over a number of years. These often involve pain in the abdomen, back and pelvis. Such patients have often been extensively investigated by physicians and frequently admitted to hospital. Their medical notes are thick but inconclusive. Unlike patients

• A sensitive and tactful approach in which the patient is engaged and supported via the adoption of a psychosomatic approach (see main text) which does not attempt to deny the reality of the symptoms or to undermine directly the patient's belief system about his illness; • Psychotherapy directed to helping the patient to adjust to the losses (e.g. of job, mobility, social life) that result from CFS; • Cognitive-behavioural therapy aimed at reattribution of symptoms (e.g. 'I am tired because I am unused to exercise, rather than because my muscles are irremediably damaged'); • Anxiety management to reduce tension and panic; a programme of graded exercise • if depression is a prominent feature; prescription of antidepressants.

Epidemic hysteria This rare and self-limiting condition occurs in schools and other situations where young people are in close proximity. Fainting, nausea and paraesthesia brought on by overbreathing are common presentations. The condition is sometimes wrongly assumed to have an infective cause.

Diagnosis Hysteria is not an easy diagnosis to make. A proportion of patients diagnosed as hysterical turn out to have genuine organic disease: multiple sclerosis or occult cancer, for example. Others may develop another psychiatric condi-

SUMMARY 17 Psychological mechanisms in hysteria • • • • • • •

Dissociation Denial Primary gain - elimination of mental pain Secondary gain - attention, care Identification Illness behaviour Unconscious avoidance of conflict

tion, such as a depressive illness or even schizophrenia. However, it is important to recognize and diagnose hysteria when it does occur, so as to spare the patient unnecessary investigation. Hysteria and organic illness can also, and often do, coexist; establishing one diagnosis should not prevent the search for another.

SUMMARY 18 Management in hysteria • • • • • •

8

'Psychosomatic' attitude on part of doctor Exploration of conflict Behavioural approaches (Antidepressants) Abreaction Long-term psychotherapy

Meaning and aetiology The manifestations of hysteria are often dramatic, and therein lies a clue to its nature. Most people, especially as children, have the capacity to act a role. This involves an internal splitting, or dissociation, between the actor and his part. The actor is not really Hamlet, yet becomes Hamlet as he acts the role. The hysteric is not paralysed and yet, psychologically, can become so. According to Freud, the symptoms in hysteria have a meaning which can be understood in terms of the patient's life. Psychological theories of hysteria also highlight several other important aspects of the illness: • Defence. Hysteria is a defence against the anxiety created by conflict. A young man from an academically minded family was facing exams which he feared he would fail; he became paralysed. • Unconscious behaviour. Although a conflict is often obvious to others, the patient is unconscious of it. • Personal gain. There are two kinds of gain for patients in being ill. The primary gain is the fact that they are unaware of their conflict and thus do not experience anxiety. The secondary gain is the care and attention they get from family, friends and relatives. This is only possible if the family to some extent go along with it, or allow the patient to manipulate them. Here they may be meeting their own needs, e.g. the husband who only feels potent when he has a sick wife to look after. 'Manipulation' is a normal process in childhood, and hysteria always has childlike aspects. The child who does not want to go to school develops a tummy-ache; his mother lets him stay at home with her. A similar process is at work in hysteria. The child is not consciously making himself ill or malingering (as, for example, is the soldier who deliberately shoots himself in the foot to avoid going to battle): there really is a pain, even if it clears up quickly once the danger of having to go to school has passed. • Identification. A common finding in the histories of hysterical patients is that there has been a close relative who has suffered illness, often serious. The young man who became paralysed because he feared his exams had had a younger brother who had died with cerebral palsy, and on whom his parents had lavished all their care and attention. The patient often has such a model of illness and its accompanying behaviour which influences him. Sociologists speak of 'illness behaviour' to describe the socially sanctioned behaviour that accompanies illness and which is not strictly related to the pathophysiology of the disease. Sick people are allowed to retire tem-

porarily from work and spend time in bed; they receive flowers, cards and gifts from relatives and are treated with special consideration; they are allowed to be childish and demanding. Their medical attendants also adopt a particular role, usually one of reassurance, benevolence and quasi-omniscience. One of the rules of illness behaviour, however, seems to be that the patient should recover, or at least attempt to recover. When this does not happen there may be a sudden switch: the previously privileged patient may be seen as manipulative, attention-seeking and deviant, and be shunned or covertly punished by irate relatives and medical attendants. In hysteria this often results in a worsening of the symptoms.

Management It is tempting to try to argue a hysterical patient out of his illness. Occasionally this can be helpful, but more often it entrenches the patient more deeply in his symptoms and an escalating battle between doctor and patient may then ensue. There is, moreover, a sense in which the patient is right: he is ill, but psychologically, not physically. There is something wrong, namely a psychological conflict, albeit one of which the patient is unconscious. A more helpful approach is therefore to adopt a psychosomatic attitude. This involves explaining to the patient that stress can cause illness, that he may have a psychological conflict, and that there may be a connection between this and the physical problem, both of which are real. The patient who has committed himself to a physical illness should be offered appropriate physical treatments (e.g. physiotherapy, massage) and should not be deprived of it out of spite. When recovery does take place the temptation to say T told you so' should be resisted. Instead, the patient needs a face-saving formula: the paralysed person must be allowed to put his crutch down gradually, inadvertently even, not have it knocked from under him. The psychological principles of treatment are: • Exploration of conflict. With allowance for face-saving, this is perhaps the most helpful strategy. If the conflict is removed (e.g. the time for exams past) the symptoms will usually remit. • Cognitive-behavioural approaches. These are often useful. They (a) aim to avoid reinforcement of the symp-

237

toms, e.g. by suggesting to ward staff or relatives that they respond positively to any signs of movement in the paralysed patient, but as far as possible ignore (nonpunitively) helpless behaviour; (b) are directed towards helping the patient overcome his cognitive responses (e.g. of panic and catastrophe) towards the imagined illness. The latter is sometimes known as 'reattribution therapy' - the patient learns how he wrongly attributes normal bodily sensations to illness, and that this may be influenced by emotion, especially anxiety. Drugs. Where hysteria is secondary to a depressive illness antidepressants should be prescribed. In general, however, drug treatment should be avoided in hysteria, as the patient is often very suggestible and tends to develop dramatic reactions and side-effects. Hysterical amnesia can sometimes be approached by abreaction, in which painful experiences are 'relived' with the help of an intravenous injection of a sedative drug such as diazepam. Psychotherapy. Some patients will benefit from longterm psychotherapy.

HISTRIONIC PERSONALITY AND RELATED CONDITIONS There are a number of conditions that are related to hysteria but which take the form of a lifelong personality disorder rather than an illness. These conditions share with hysteria the capacity for dissociation and the use of illness for manipulation of those around them.

HISTRIONIC PERSONALITY A patient with histrionic personality (usually female) finds it difficult to form close attachments, and when she does they are highly dependent. Her emotions are often superficial and rarely satisfying. She may be very alluring and attractive to men, but finds it difficult to enjoy sex herself. She carries with her an atmosphere of drama: her life is often a series of emotional or practical crises. Her feelings

SUMMARY 19 Conditions related to histrionic personality

238

Multiple personality

Aspects of the self entirely dissociated from one another, with lives of their own

Factitious illness

Self-inflicted illness

Fictitious illness Patient travels from hospital to hospital (Munchausen's syndrome) with fictitious illnesses. Pseudologia phantastica.

are dramatized and exaggerated; her headaches are always 'migraines', her diarrhoea 'dysentery' and her pains invariably 'excruciating'. She uses her crises to coerce those around her (including doctors and nurses), and sometimes has a husband whom she abuses and yet clings to, and who appears to submit willingly to all her demands. As a child she may have felt emotionally neglected and insignificant, helpless and dependent. The anger and manipulation that manifest themselves in adult life can be seen as a continuation of this, an attempt, through the coercive power of illness, to keep attached to her those whom she feels may otherwise ignore or abandon her. Patients with histrionic personality sometimes develop the physical symptoms of hysteria.

MULTIPLE PERSONALITY The adult personality is made up of a number of elements: • Some come from identifications based on parents ('he has his father's temper and his mother's charm'). • Some represent different aspects of the self: the caring self, the aggressive self, the childish part of the self, and so on. • Some can be defined in terms of relationships: in a marriage a partner may be expected to be a spouse, lover, parent, colleague and friend. Eric Berne has summarized the different aspects of the self as 'parent', 'adult' and 'child', all of whom coexist within one person. In the mature personality there is a blending together of the different aspects into a coherent whole, so that a person is all of a piece, although minor discontinuities and splits in the personality are common. The fascinating but rare clinical condition of multiple personality, in which several quite separate personalities appear to inhabit the same individual (fictionalized in such works as DrJekyll and Mr Hyde and The Three Faces of Eve), is an extreme example of this everyday splitting. In multiple personality the different personalities have lives of their own which are entirely dissociated from one another; the good self is totally unaware of the bad self and would be horrified by its activities. This enables, say, a prudish aspect of the self to live alongside the sexual side without being aware of it, thereby avoiding anxiety or guilt. Patients with multiple personality can occasionally be helped towards integration with psychotherapy.

FACTITIOUS ILLNESS Patients with unexplained symptoms are sometimes found - often after lengthy investigation - to have caused their own illness. A patient with unexplained and extensively investigated 'haemoptysis' was found to be venepuncturing herself, swallowing blood and then spitting into a sputum pot; a person with unexplained intestinal bleeding

may be inserting foreign bodies into the rectum; hypoglycaemic attack can be produced by insulin injections in non-diabetics; dermatitis artefacta (p. 436) is yet another example of such factitious illness. Occasionally such behaviour is due to psychosis, but more often it is a manifestation of a personality disorder. These patients are often very 'nice'; they may be nurses or members of the other caring professions. They are often on good, even intimate, terms with medical staff. They may induce incomprehension, abhorrence and anger once they are discovered. Factitious illness is best seen as an extreme example of histrionic personality disorder. The patient has split off an angry, destructive, defiant part of herself which acts semiautonomously and of which the 'good' self may be partially unaware. The principles of management of hysteria may be applied in these cases. A full history will reveal underlying conflicts, and the patient can then be told, in as gentle and unconfrontational a way as possible, that these conflicts may be responsible for her symptoms and that she may thus be playing a part in causing her own illness. Although the patient may become angry, the results of such an approach are rarely disastrous. The patient will often be relieved and even begin to accept psychological help.

FICTITIOUS ILLNESS: MUNCHAUSEN'S SYNDROME Baron von Munchausen was an 18th-century fictional character who travelled extensively and told fantastic stories. His name is now used to describe a group of patients with characteristic features: • Patients often experience recurrent unexplained symptoms, especially abdominal or chest pain, which are often thought to be due to renal colic, myocardial infarction or intestinal obstruction. The patient may be given opiates and have had several inconclusive laparotomies, which reveal themselves as multiple abdominal scars. • These patients travel from hospital to hospital, usually discharging themselves after a brief admission, and using a variety of assumed names. • They give dramatic accounts of their lives, for example stories of heroic wartime feats or tragic losses, none of which can be corroborated and none of which quite add up. These stories are sometimes labelled pseudologia phantastica, Munchausen patients may be seen as the male counterparts of factitious illness patients. They are hard if not impossible to manage or help, but awareness of the diagnosis will avoid unnecessary or dangerous investigations. Most doctors have had at least one experience of being hoodwinked in this way, and this may be a necessary learning experience. The doctor must be friendly but firm in refusing to treat these patients as though they are ill. The patient usually disappears, but may occasionally accept social help and often become highly dependent on medical services.

SUMMARY 20 Factors contributing to anorexia nervosa • • • •

8

Contradictory social attitudes towards eating and thinness Enmeshed, i.e. mutually overdependent, families Fear of loss of control of eating and of sexuality Fear of separation

EATING DISORDERS Eating is not just about physiology: food has great psychological and sociological significance. Morality starts with the child's discovery of what is good and bad to eat, and a mother's self-evaluation may continue to depend on how well she feels she feeds her family. Birth, the reaching of adulthood, marriage and death are all marked by a meal. In addition, there are two factors, specific to many highly industrialized western societies, that are of especial relevance to eating disorders: the role of the food industry, with its large marketing system; and the emphasis on a slim body image as a mark of vitality and attractiveness. Eating disorders are increasingly common, especially among women (Fig. 8.10). Although the overall incidence of anorexia nervosa is only 1 per 100000 population, 1% of 16-18-year-olds are affected; 2% of women attending a family planning clinic suffer from bulimia nervosa; and the prevalence among dance students is 20%.

ANOREXIA NERVOSA Aetiology and pathogenesis No single cause of anorexia nervosa has been found. A number of factors appear to be involved: • Social pressures towards both eating and thinness, though contradictory, are important.

FIG. 8.10 Epidemiology of anorexia nervosa Social class distribution of anorexic patients (father's occupation) compared with that for general population.

239

CASE STUDY 8.3 A 30-YEAR-OLD WOMAN WITH AN EATING DISORDER AND A WEIGHT OF 30 KG A 30-year-old woman was admitted to a psychiatric unit weighing 30 kg. Her body mass index (weight in kilogrammes divided by height in metres, squared) was, at 13, near the lower limit considered to be compatible with life. She was brought in by her husband, who had been increasingly worried about her but up to this point had been unable to persuade her to accept help. She had had eating disorder symptoms since at least the age of 15, when she had left home to live with her boyfriend, who eventually become her husband. She was terrified of excess weight and weighed herself several times a day. She ate little more than a crispbread and a slice of cucumber daily. She had never had regular periods, although, unusually for someone with anorexia nervosa, had managed to get pregnant with the help of ovulationassisted fertility treatment. She was closely involved with her mother, to whom she presented an 'everything is normal' front, while seeing her almost every day. She was very worried about her parents' marriage, which she saw as in great jeopardy. She was treated with a refeeding schedule, with which she was surprisingly compliant, until she weighted 45 kg, and then was discharged from hospital. Her outpatient care consisted of a combination of individual psychotherapy and family therapy. Her individual therapy concentrated on looking at all the ways in which she pretended to herself and others that she was eating when in fact she was avoiding food. It used an 'addiction' model and suggested that she was addicted to not eating in the way that an alcoholic is addicted to alcohol - using it as a refuge whenever she felt stressed or confused. Therapy also looked at her

240

image of herself as a woman, and at how frightened she was about becoming 'like' her mother, whom she saw as being trapped in a loveless marriage, constantly at her father's beck and call. The family therapy sessions tried to help extricate her from the parent-child relationship she had established with her husband and to foster a more equal partnership, in which she took charge of her own eating, rather than either deceiving him or relying on him to prompt her to eat. There were no further admissions, but her weight hovered just under 7 stone for many years, and she said that the idea of being 7 stone terrified her. Over the years she made many life changes. She left her husband and formed a new partnership, which was more emotionally open and caring, although not without its problems. Her new partner was more able to help her to eat at times when she was in the grip of her anorexic symptoms. Her parents' marriage also broke up, and she become much less involved with her mother.

Questions

1. Could intervention have started earlier? 2. What is the long-term prognosis of anorexia nervosa?

Discussion It is notoriously difficult to engage people suffering from anorexia nervosa in treatment. Teenage sufferers are often taken unwillingly to the doctor by worried parents. It is always important for the doctor usually a family doctor in the first instance - to interview the patient on her own and to make determined

attempts to establish rapport, a process which may take several interviews. There is a danger that the doctor will be pulled into identification either with the worried and exasperated parents, in which case he may become heavy handed and insensitive, or with the patient, when he may minimize the seriousness of the situation and fail to be appropriately firm and challenging. Anorexia is 'delusional' in the sense that such patients are convinced that they are fat, despite every rational argument to the contrary. There has been much debate about the ethics of admitting people with anorexia compulsorily to hospital. The current consensus is that the thinner the patient the less rational they are, and that in lifethreatening situations the use of the Mental Health Act is justified. On the other hand, it is essential to help the patient move towards taking responsibility for her own illness, and a Section may militate against that. Many patients put on weight while in hospital but immediately revert to dangerous weight loss once they are back home. Anorexia nervosa is a very varied illness, with many different trajectories, making it difficult to generalize about prognosis. In general, the shorter the interval between onset of illness and treatment, the better the prognosis. Most studies suggest a 60-70% recovery rate within 5 years. The longer the period of illness the greater the mortality. Different studies have shown lifetime mortality of between 5 and 10%. Partial recovery is probably the best that can be hoped for in a chronic form of the illness such as that suffered by the patient described, whose symptoms had lasted for more than 15 years.

• The fear of losing control, both of eating and of sexuality, and a feeling of inadequacy and of not being a separate person are characteristic of the anorexic's inner world. Anorexia is an attempt by the individual to remain in complete control by regulating her food intake and her weight, and thereby putting her sexuality into reverse. • Family function in anorexia is often abnormal. The families (where food is frequently a central theme) tend to be 'enmeshed', i.e. mutually overdependent. This causes little difficulty in early childhood, but with the advent of adolescence the family may be threatened by the possibility of autonomy and separation. • Endocrine abnormalities occur in anorexia nervosa. Hypothalamic function is disturbed, with reduced luteinizing hormone (LH) and gonadotrophic hormone output, but this is more probably a consequence rather than a cause of weight loss. Low levels of plasma thyroxine and plasma TSH are also common. There may be alkalosis if there is surreptitious vomiting.

Clinical features

overweight, which may appal the girl, who then begins to pursue thinness. Her thoughts are dominated by food and how to avoid it. She absents herself from family meals, although she may pride herself on providing food for others, while secretly despising and envying those who eat normally or, as she sees it, gluttonously. As long as she is thin she will be happy. To be thin is to be good; to be fat, bad. She may augment her efforts at dieting by taking laxatives, amphetamines or diuretics, and by self-induced vomiting. Battles about food may develop with her parents. Denial of weight loss (p. 216) is an important feature: she may conceal her emaciation within voluminous and floppy clothes. She is physically overactive as part of her secret plan for weight loss. Sometimes, deliberately excessive exercise may be a major factor in the weight loss. She complains of a feeling of 'fullness' in her stomach which prevents her from eating. She will be emotionally and socially, but not intellectually, less mature than her peer-group. Physical examination will reveal thinness, scanty secondary sex characteristics, possibly peripheral cyanosis, a soft downy covering of the face and body (lanugo hair), hypotension and bradycardia.

8

Four features are characteristic of anorexia nervosa.

Differential diagnosis • The most important is phobic avoidance of normal weight. The patient is terrified by the idea of her normal weight. • With this goes a distortion of body image, known as overperception. By asking women to choose a silhouette that they feel represents their shape, it can be shown that it is normal for women aged 15-30 to perceive themselves as up to 50% fatter than they really are. Patients with anorexia nervosa see themselves as at least 100% bigger than they are: an emaciated girl of 35kg may be disgusted by what she perceives as her gross obesity. • The consequence of phobic avoidance and overperception is relentless dieting, especially the renunciation of carbohydrates. • Amenorrhoea is an essential diagnostic feature. In males (10% of all anorexics) there is disturbance of sexual function, usually with loss of libido. Anorexia nervosa usually starts in early or midadolescence. There is frequently an initial period of mild

SUMMARY 21 Features of anorexia nervosa • • • • • • • • • •

Phobic avoidance of normal weight Lanugo hair Hypotension Relentless dieting Denial, concealment Self-induced vomiting, laxatives Overperception of body image Excessive exercise Enmeshed families Amenorrhoea

The clinical picture may be one of two types: • In typical anorexia the only obvious psychological abnormality is the difficulty in eating. Here, the main differential diagnoses are organic causes of weight loss in a young woman, for example malabsorption syndrome, thyrotoxicosis or occult neoplasm. • An atypical picture also occurs. Here the woman is often older (in her 20s) and her anorexic symptoms coexist with other psychiatric features, such as depressive symptoms of misery and withdrawal; obsessional symptoms centring, for example, on cleanliness; and personality difficulties.

Management A patient weighing less than 35 kg should be admitted to hospital. A behavioural regimen in which the patient agrees on a target weight and is rewarded with privileges (watching television, being allowed visitors) in return for weight gain is usually effective in the short term. Occasionally a small dose of a phenothiazine (e.g. chlorpromazine 50 mg b.d.) is helpful. However, the patient will often lose weight again once she is discharged, unless psychotherapy is also started. Psychotherapy is the treatment of choice for outpatients. Mild anorexia can respond to simple counselling. In more severe cases family therapy is the best treatment for those under 17, whereas for older patients (especially if they have left home) individual psychotherapy is the preferred treatment. In each case, the aims of therapy are: • to maintain normal weight (body mass index 18-25) and a regular eating pattern. Some patients use therapy as a diversion from this central task;

241

to help both patient and family to move away from the area of food as a focus for all conflicts and difficulties; to help the patient accept herself as she is, in both her body and her feelings, i.e. to cope with the problems in living from which the eating disorder may have been a diversion as well as a cause.

SUMMARY 22 Treatment of bulimia nervosa • Keep a diary of everything eaten and all episodes of vomiting • Re-establish dietary control; eat three meals per day even if bingeing • Identify triggers to bingeing (anger, loneliness) • Challenge erroneous assumptions, e.g. 'If I eat a chocolate bar will become grossly obese and no-one will love me'

Prognosis Twenty-five per cent of anorexic patients recover spontaneously; 25% make a full recovery with medical intervention; about 25% continue to experience some difficulties with eating but can lead a normal life; and 25% remain quite severely handicapped. About 5-10% die of the condition. Poor prognosis is associated with later onset, 'atypical features' (see above), marked body image disturbance and self-induced vomiting. Recovery is accompanied by a reversal of the hormonal abnormalities and the onset of normal menstruation. 1

BULIMIA NERVOSA Bulimia nervosa is a syndrome of 'binge eating' (bulimia: to stuff oneself with food) and self-induced vomiting in the presence of normal weight. There is an overlap here with anorexia nervosa, where binge-vomit cycles can also occur, but in anorexia the patient is always abnormally thin and also amenorrhoeic. The usual age of presentation in bulimia also tends to be older, with patients (usually women) in the 20-30 age range, rather than the teens. The bulimic is frightened of becoming fat, but can accept her normal weight. She is terrified by her greed and tries to avoid its consequences by making herself sick. She normally feels deeply ashamed about this. She lives in a state of dietary chaos, in which she is determined to fast, succeeds for a while, then succumbs to hunger and starts eating. She binges, then feels guilty, makes herself sick and then starves; and so the cycle starts again. The physical consequences of bingeing and vomiting include erosion of dental enamel from exposure to gastric acid, Mallory-Weiss oesophageal tears, and even oesophageal rupture. Repeated vomiting causes hypokalaemia, which may in turn produce tetany, paraesthesiae, ileus, cardiac arrhythmias and epileptic fits. Bulimia is more common than anorexia: the prevalence rate in Europe and North America is 1-2%, compared to 0.5-1% for anorexia; 20% of women attending a family planning clinic admitted to bingeing, and 2% had the full syndrome. Like anorexia, bulimia may exist either on

1

242

MCQ 8.5

2

MCQ 8.6

its own, or as part of a general disturbance of personality, and be associated with alcoholism, deliberate self-harm or shoplifting, all of which make the prognosis less good. Treatment of bulimia nervosa is usually successful and involves helping the woman to re-establish a regular eating pattern, and to feel less guilty. The bulimic often feels a great relief when dietary control is re-established, unlike the anorexic, who feels threatened at the idea of giving up her own control. In addition to a cognitive-behavioural approach, exploration of the conflicts underlying the bulimia (such as marital difficulties or feelings of anger which the patient is unable to express) is also necessary. A proportion of bulimics have experienced sexual abuse in childhood, and this too needs psychotherapeutic exploration.

OBESITY Obesity in itself is not a psychiatric problem, but morbid obesity can be experienced as a severe handicap, comparable to amputation or disfigurement. Respiratory, endocrine and cardiovascular complications are common and serious, and gross obesity should, like anorexia nervosa, be considered a life-threatening condition. Obese people feel ostracized and experience self-disgust. They long to be 'normal' and may become depressed by unsuccessful efforts at dieting. There are often marital and family difficulties which are both a cause and a consequence of their obesity. The prognosis for weight loss, either by dieting or surgery (e.g. intestinal bypass) is not good. Psychotherapy can be a precursor to weight reduction if it helps the patient to accept himself as he is. The patient may then find it easier to diet, being less inclined to resort to the comfort of food whenever an emotional upset occurs.

FURTHER READING ON EATING DISORDERS Fairburn G C, Wilson G (eds) 1992 Binge eating. Guilford, New York. Schmidt U, Treasure J 1992 Getting better bit(e) by bit(e). Laurence Erlbaum, Hove Tiller J, Schmidt U, Treasure J 1993 Treatment of bulimia nervosa. Int Rev Psychiatry 5:75-86.

by doctors) and 'bad' ones (self-administered by addicts). This distinction does not take account of the fact that almost all drugs, even when medically prescribed, have harmful effects and can be misused: analgesics and tranquillizers are the best examples of this. Conversely, opiates can be of great benefit in the treatment of severe pain. Current definitions of drug dependence are based on the interaction between the individual and the drug, leading to a compulsion to take it:

8

• In order to experience its psychic effects, and/or • To avoid the unpleasant effects of its withdrawal (abstinence syndrome). There is no single cause of drug addiction. The prevalence of addiction is determined by:

FIG. 8.11 New and former addicts notified to the Home Office

• The availability of the drug. The greater the availability, the greater prevalence of addiction. In Hong Kong, where heroin is widely available, 8-10% of the adult male population is addicted. • The accessibility of the drug. This explains the relative commonness of drug addiction among doctors and nurses, and in inner-city and coastal areas. • The susceptibility of individuals to addiction or misuse. Management of drug addiction is equally multifaceted and involves a combination of physical, psychological and social rehabilitation.

DRUG DEPENDENCE Addiction to illicit drugs, especially opiates, is a source of widespread medical and social concern (Table 8.13) 2. There are several reasons for this. First, there has been an epidemic increase in opiate abuse over the past 50 years. In 1936 there were about 500 known addicts in the UK (of whom one-quarter were doctors); by the mid-1970s this figure had risen to 6000; and in 1988, following the introduction to the UK of cheap heroin, the number of known addicts was estimated to be at least 50000 (Fig. 8.11), of whom 9000 were registered with the Home Office. By the year 2000 this figure had reached 20000. Secondly, the impact of opiate addiction has profound medical and social consequences. Within 5 years of diagnosis 10% of addicts will have died from drug-related causes, including AIDS, hepatitis B and overdose. Thirdly, 'drug addiction' is a microcosm of a much wider social problem. Finally, there is increasing concern about the way that substance abuse complicates and worsens the prognosis of many psychiatric illnesses. For example, schizophrenia may be accompanied by, and sometimes be difficult to distinguish from, cannabis, heroin and/or amfetamine abuse, leading to a so-called 'dual diagnosis'.

Nature and causes of drug dependence In the past, definitions of drug abuse have concentrated on a contrast between the harmful and the beneficial effects of drugs, thus dividing drugs into 'good' ones (prescribed

OPIATE ADDICTION Causes The most important cause of opiate addiction is availability. Susceptible individuals expose themselves to the drug, either because they inhabit a subculture in which drug taking is prevalent, or because they are psychologically vulnerable owing to family difficulties, depression or boredom. The pharmacological features of opiates - tolerance and withdrawal syndrome - ensure the establishment of a habit, and social as well as physiological factors then maintain the addiction and make rehabilitation difficult.

Biochemical factors The discovery in 1977 of the encephalin neurotransmitters was a major advance in the understanding of opiate addiction. Endorphins are naturally occurring, opiate-like substances which mediate pain and other sensations in the midbrain. It is possible that individuals susceptible to addiction may genetically be deficient in endorphins and so 'opiate hungry'. Both opiates and endorphins reduce the amount and effects of other cerebral neurotransmitters, such as acetylcholine. This has led to the 'supersensitivity' theory of withdrawal syndrome, which postulates that addicted individuals have reduced amounts of transmitter reaching postsynaptic receptors. Stopping the drug then

243

TABLE 8.13 Addictions: Prevalence data Substance use is one of the major social and health problems facing society Illicit drugs • 23% of those aged 14-15 have taken an illegal drug in last year, 13% in last month, 29% of those aged 16-24 in last year, 14% in last month. • Known heroin addicts have increased from about 5000 in 1980 to about 50000 now, with growth continuing at about 20% per year. • The number of seriously problematic drug misusers in UK is estimated to be between 100000 and 200 000. • Amfetamine seizures have increased from about 3000 in 1987 to about 18000 in 1997. Cocaine use is also increasing rapidly. Alcohol • 41% of men aged 18-24 drink above recommended levels, but recent increase highest among young women (currently 25%). • Alcohol use increasing rapidly among schoolchildren - currently 5-10% 14-15-year-olds drinking more than recommended adult levels. • 51% of rough sleepers are alcohol dependent. • More elderly people are drinking above recommended limits - from 1984 to 1996 increase from 12% to 18% of men, and 3%-7% of women. Smoking • By 13 years, 19% smoking cigarettes daily. 29% of 16-19-year-olds are current smokers, 39% of 20-24-year-olds and 28% of all adults. Mortality • Between 1000 and 3000 drug misusers died as a result of overdose in 1998 in England and Wales (uncertainty relates to problems with current databases). Compared to the general population, mortality rates are higher among addicts (SMR = 7) • Injecting drug users have a mortality 12-22 times greater than their peers. Compared to the general population, suicide rates are high among addicts (SMR for males = 4.4; SMR for females = 11.3). • 25% of road accident fatalities have blood alcohol above 80mg%, 30% of pedestrians sustaining injuries. Illicit drugs are found in blood of 15-20% of drivers involved in fatal road accidents. • 40000 alcohol-related deaths per annum. • 120000 smoking related deaths per year in England and Wales. This accounts for 20% of deaths at all ages, and 25% of deaths between 25 and 35. Psychological and physical morbidity • 62% long-term injectors are positive for hepatitis C, and 52% for hepatitis B. About 80% of those positive for HCV will develop long-term infection, and 5% of those positive for HBV. About 20% of those with chronic infection will develop cirrhosis or liver cancer. Between 152000 and 228000 drug users in the UK are currently infected with HCV. • 10000 individuals seek help for their own or someone else's drinking problems each day. • 30-40% of hospital in-patients drink alcohol excessively. Alcohol is associated with heart disease, liver cirrhosis, cancer, neurological and gastro-intestinal problems. • 30-40% of substance misusers also have a psychiatric diagnosis. 60% of parasuicides are alcohol related. In a health district of 250000, 3000 problem drinkers will also have severe mental health problems. • Alcohol is strongly related to high-risk sexual practice, unwanted pregnancy and sexually transmitted disease. • Domestic violence, accidents and social disharmony: 30% of divorces, 40% of domestic violence, 20% of child abuse cases and 30% of domestic accidents are associated with alcohol misuse. Crime • 32% of proceeds of acquisitive crime is spent on heroin or cocaine. • The costs of drug-related crimes to the criminal justice system is at least £1 billion per year. Approximately 30% of prisoners are problem drinkers. 30-60% prisoners claim to have offended under influence of alcohol. 60% of those arrested test positive for drugs, 20% for opiates. Costs • The costs to industry of excessive alcohol consumption are estimated at £2330 million per year, as a result of unemployment, premature death and sickness absence. • Smoking costs the NHS approximately £1500 million per year. For a typical health authority this means £14 million per year, broken down as GP visits £2.5 m, prescriptions £1.5m, in-patient stays £3.2m, day-care £1.9m and outpatient visits £4.9m. Cost-effective treatments require proper training and specialist support • Many studies have shown that the treatment of opiate users saves much more money than it costs. • In North American studies it is estimated that for every $10000 invested in alcohol treatment, $30000 is saved by the managed care provider. • Smoking cessation interventions cost between $212 and $873 per life year gained, compared with £17000 as the median cost per life year gained of 310 common medical interventions.

244

results in a sudden increase in transmitter and stimulation of the already supersensitive receptor.

Psychological factors Drugs can be administered ('dropped') by injection ('fixing'), either subcutaneously ('skin-pop') or intravenously ('mainline'); or by smoking ('chasing the dragon'). All these methods are powerful reinforcers in the pavlovian sense:

TABLE 8.14 History and examination of suspected opiate abuser History Previous contact with a prescribing doctor or clinic History of recent use (doses, frequency, legal, illegal) Assessment of demands or requests Assessment of withdrawal symptoms Current legal problems

• Primary reinforcement from both the pleasure ('buzz') and the removal of withdrawal symptoms; • Secondary reinforcement from the associated rituals of administration, e.g. filling the needle; • Social reinforcement from fellow addicts.

Examination Injection sites: needle marks, tracks, bruises, thrombosed veins, abscesses, skin ulcers Arterial spasm or gangrene (caused by, e.g. dipipanone, cyclizine) Jaundice from hepatitis Emaciation

Psychiatric factors

Investigation Urine for drug assays Liver function tests Hepatitis serology HIV test (with consent)

Opiate addiction is more likely to be a cause than a consequence of psychiatric illness. 'Loners' may well be suffering from anxiety and depressive symptoms which they are trying vainly to treat with opiate drugs. Hospitalized addicts (as opposed to the generality of addicts) tend to be suffering from personality disorder and to have difficulty in forming close relationships. Addicts are caught in a vicious circle in which their primary relationship comes to be with the drug, which is felt to be more accessible, reliable and instantly gratifying than a person. There is some evidence that family discord and alcoholism are more common in the childhood histories of addicts than in controls. The anxiety associated with the early stages of opiate withdrawal may mirror early childhood experiences of separation from desired, but inaccessible, parents.

Social factors There is no particular social class preponderance for opiate addiction.

Presentation Addiction to opium and its derivatives, morphine and heroin (diacetyl morphine), can be considered as a model of drug addiction in general. The physical signs that should arouse suspicion of opiate abuse are listed in Table 8.14. Addicts present to doctors in one of four ways: • With infective complications of self-administered injections, e.g. septic abscess at injection site or septicaemia, infectious hepatitis or manifestations of HIV; • With confusion or unconsciousness due to drug overdose, either deliberate or accidental; • Wanting, or apparently wanting, help with their habit; • With a withdrawal syndrome. This consists of a craving for the drug together with physiological features, which include rhinorrhoea, lacrimation, yawning, perspiration, irritability, gooseflesh, nausea, diarrhoea, cramps,

8

muscle pains, tachycardia, flushing and involuntary movements. It is not life-threatening and the symptoms begin to recede about 72 hours after last taking the drug.

Management Legal requirements In the UK, any doctor attending a patient who is known or suspected to be addicted to opiates is required to inform the Home Office Drug Branch. Only specially licensed doctors are permitted to prescribe morphine and diamorphine (heroin) for the maintenance of addiction. Other opiates, such as methadone, may be prescribed by any doctor for the management of withdrawal syndrome, but this should be done only with great constraint and in an emergency. General management Drug abusers may not be open about their habit, which should be suspected in unexplained cases of psychosis, sepsis, serum hepatitis or HIV infection. Scapegoating should be avoided, and drug users should be treated with the same concern and courtesy as any other patient. At the same time, caution should be exercised: addicts may steal prescription pads and syringes, and exaggerate their drug needs in order to obtain drugs, which are then sold. In an initial interview a urine screen should be carried out to confirm that the patient is a user, and blood should also be tested for hepatitis serology and HIV (the latter normally with the patient's consent). Whenever possible, the patient should then be referred for specialist help. Occasionally

245

TABLE 8.15 A typical detoxification schedule for opiate addiction 10-day detoxification (volume kept at 25 ml)

28-day detoxification (volume kept at 25 ml)

Methadone 25 mg/day - 2 days 15mg/day-2days 8 mg/day - 2 days 4 mg/day - 2 days 2 mg/day - 2 days

Methadone 30 mg/day - 4 days 25 mg/day - 3 days 20 mg/day - 4 days 13 mg/day-3 days 10 mg/day-4 days 6 mg/day - 3 days 3 mg/day - 7 days

TABLE 8.16 Characteristics of methadone linctus which justify its use as a treatment • • • • •

Non-injectable Not attractive to users looking for buzz Low street retail value Controlled drug - able to be prescribed on daily collection Long half-life for established users, hence reliable urine testing and stabilizes drug-taking experience • If it does 'leak' from the drug service to the streets it is likely only to be used by established users for whom it is a safer preparation. Some of these will use purchased methadone for detoxification • As a linctus', consumption can be supervised if necessary

(e.g. in a casualty department at a weekend) withdrawal symptoms have to be treated and a dose of methadone given equivalent to the average daily intake which the doctor calculates is being used. The usual dose is between 20 and 70 mg of methadone. Typical detoxification schedules are given in Table 8.15. Where heroin intake is low, diphenoxylate and clomethiazole (chlormethiazole) can be used. Prescriptions should be on a daily basis for not more than 2 days. A clear management policy should be instituted and explained to the patient. Treatment of medical complications of addiction The main medical complications of addiction are septicaemia (p. 273), hepatitis (p. 846), endocarditis (p. 574) and HIV infection (Ch. 11).

Rehabilitation In the UK, all regions have drug clinics to which the addict should be referred. Here, the patient may be offered an oral methadone maintenance programme (Table 8.16) in the hope - not always realized - of reducing the use of illicit drugs. An attempt is made to form a rapport with the user. After establishing a baseline habit, some addicts will be gradually withdrawn from the drug over a 10-week period, either as an outpatient or, more rapidly, as an inpatient. Others will be stabilized on a maintenance methadone programme as a way of facilitating a period of stability in their life which reduces the need for crime and enables them to form a relationship with the drug clinic team. Withdrawal is an essential, but by no means sufficient, aspect of treatment. Addicts exposed to enforced abstinence, e.g. in prison, almost always relapse when released. Psychotherapeutic and social help is also needed if the addict is to remain drug-free.

FIG. 8.12 Changing patterns of consumption of non-opiate drugs in the UK A Prescriptions for sedatives/tranquillizers and hypnotics, 1974-1985. B Current benzodiazepine use and dependence in the UK. In one year, (A) 9-10 million people use benzodiazepines. (B) Of these, 3.2-3.5 million use them for more than 4 months and are at risk of dependence. (C) Of these, 0.64-1 million are likely to suffer withdrawal problems.

Benzodiazepines ADDICTION TO NON-OPIATES

246

Dependence on non-opiate drugs has become a major health problem in recent years. The changing pattern of dependence is shown in Figure 8.12.

Benzodiazepines have now replaced barbiturates both in popularity and as a cause for widespread addiction. Although the three million annual prescriptions cause less mortality than barbiturates (as they do not produce respiratory depression), they are an important source of mor-

TABLE 8.1? Symptoms of benzodiazepine withdrawal • • • • • • • •

Insomnia Anxiety Depression Outbursts of anger Muscular pains Tremor Headache Nausea, weight loss, anorexia

• Sweating • Tiredness • Perceptual abnormalities Visual disturbance Depersonalization Vertigo • Fits • Psychosis

bidity (see p. 216). There is an increased incidence of road traffic accidents in benzodiazepine takers. A withdrawal syndrome occurs in patients who have taken more than 20 mg daily over 6 weeks (Table 8.17). The symptoms of minor affective disorder (often self-limited), for which benzodiazepines are initially prescribed, may also then become part of a withdrawal syndrome, which can be mistaken for the original illness and lead to further prescriptions. Like barbiturates, benzodiazepines are best withdrawn in a slowly reducing dosage over several weeks. Antidepressants or even ECT (see earlier) may be needed to treat the depression associated with withdrawal.

Barbiturates Barbiturate addicts now comprise a small number of older patients (mostly women) who are relics of the barbiturate era, and a number of young multiple-drug users for whom barbiturates are one of a number of drugs in their repertoire. The effects of barbiturates are similar to those of alcohol, i.e. intoxication, ataxia and depression of consciousness. Respiratory depression due to overdose is a frequent cause of death. A withdrawal syndrome occurs in which convulsions are common. This should be controlled by pentobarbital (short-acting) 4-6-hourly. Withdrawal should be carried out slowly over several weeks.

Stimulants Amfetamines do not appear to cause physiological dependence, but produce a psychosis which is clinically very similar to paranoid schizophrenia. Symptoms abate within 2 weeks of withdrawal. Diagnosis is by urine screening. Treatment is by neuroleptics such as phenothiazines. Withdrawal from amphetamines often leads to a severe state of depression. Ecstasy, a synthetic amfetamine derivative, is widely used, especially at 'rave parties'. Fits, hyperpyrexia, renal failure and death may follow. Tactile hallucinations ('formication') are a feature of cocaine intoxication, which otherwise produces effects similar to amfetamines.

Removing the hydrochloride molecule from cocaine produces 'crack', which can be smoked rather than snorted. Crack is said to be the most powerful stimulant of all and produces profound depression and anxiety, and concomitant craving, following withdrawal.

8

Hallucinogens Hallucinogenic drugs such as LSD (lysergic acid diethylamide) and 'magic mushrooms' produce a toxic psychosis characterized by vivid hallucinations and delusions. Prolonged psychosis may be precipitated by LSD in susceptible individuals. 'Flashbacks' may occur for several months after an LSD 'trip'.

Cannabis Cannabis, like alcohol, appears to have few harmful effects if taken in moderation. Ten per cent of all young people have tried cannabis at some time, and 25% of students use it at least once a week. In some cultures and subcultures (e.g. Rastafarians) regular cannabis use, especially among males, is the rule. The active principle is tetrahydrocannabinol (THC), which is sometimes used as an antiemetic in terminal care. As with LSD, excessive use (or occasional use in susceptible subjects) produces a psychosis, and a withdrawal syndrome may occur. It has been suggested that regular use of cannabis may produce a state of apathy or amotivational syndrome.

Solvent abuse Inhalation of solvents such as toluene and acetone (used as vehicles for glue and other agents) produces a temporary state of euphoria. Glue-sniffing is common in certain youth subcultures. It is dangerous: an appreciable mortality has been recorded from direct cerebral damage, as well as asphyxia from the use of polythene bags to concentrate the vapour. Morbidity includes hepatitis, renal damage, cardiac arrhythmias, polyneuropathy and aplastic anaemia.

Drug screening Urine can be analysed for benzodiazepines, cannabis, amfetamines and opiates (but not LSD), and this is a useful test in cases of psychosis where drug abuse may be a possible cause.

ALCOHOLISM Alcohol is a major cause of medical and psychiatric morbidity in the UK (Table 8.18; see Table 8.13) and worldwide. In the UK, a 200% increase in alcohol consumption since 1950 (Fig. 8.13) has been accompanied by a 25-fold increase in admissions for alcoholism to psychi-

247

TABLE 8.18 Alcohol consumption and alcoholism in the United Kingdom (population 55 million) Alcohol consumption

Estimated numbers

Regular drinkers Heavy drinkers Problem drinkers Addicted drinkers

36m 2m 700000 200000

TABLE 8.19 Criteria of alcohol dependence • • • • •

More than 80 g of alcohol consumed daily Tolerance: blood alcohol levels greater than 150mg/100mL Withdrawal symptoms Continued drinking in spite of psychological, social, and physical problems Abnormal blood tests, e.g. macrocytic anaemia and raised y-GT

Definitions Some criteria for alcohol dependence are given in Table 8.19. Two important patterns of alcoholism require definition: problem drinking and alcohol dependence. Problem drinking The WHO defines alcoholism as drinking that causes emotional, social or physical damage to the individual. Problem drinkers may suffer from all three, e.g. anxiety or depression related to alcohol; divorce, unemployment or financial difficulties; and gastritis or physical trauma associated with drinking. Alcohol dependence The essential feature of alcohol dependence is the development of a withdrawal syndrome when the drug is removed. The dependent alcoholic will drink, usually early in the day, in order to stave off withdrawal symptoms such as nausea, tremor, sweating or agitation. Other features of alcohol dependence include: • Narrowing of the drinking repertoire. The ordinary drinker's consumption varies from day to day. The dependent alcoholic has a daily drinking routine that must be followed. • Predominance of drinking over other activities. For the alcoholic, drink, even if it is clearly damaging their life, comes first. • Tolerance. The capacity of the dependent individual to drink large quantities of alcohol is an expression of physiological tolerance.

FIG. 8.13 Trends in alcohol consumption and expenditure in the United Kingdom A Alcohol consumption 1910-1990. 00 Consumer expenditure on alcohol, 1965-1990.

248

atric hospitals. About 20% of admissions to medical wards are due to alcohol or alcohol-related diseases. Forty per cent of male attenders at hospital casualty departments have drink-related problems (p. 861). In 2000 40% of UK students drank above safe limits and 25% of the general population drank above safe limits.

Presentation: the hidden alcoholic Alcoholics have to be very ill or very brave to complain of their condition directly. Only about 10% of alcoholics in the community are in contact with psychiatric services, and about half are known to other agencies such as GPs, social services or parish priests. The remainder are unknown and untreated. Doctors need a high index of suspicion if they are to diagnose these hidden alcoholics correctly. The following factors should alert the examining doctor: • Unexplained gastrointestinal complaints • Complaints of anxiety and depression or panic attacks without obvious external precipitants

SUMMARY 23 Some presentations of undeclared alcohol abuse (should raise Index of suspicion') Psychological Memory lapses Depression Anxiety Outbursts of unprovoked anger Suicide attempts Morbid jealousy Paranoia Acute confusional state (DTs) Unpredictability

Physical Gastrointestinal symptoms Dyspepsia Pain Nausea Haematemesis Neurological symptoms Tremor Paraesthesia Muscular weakness Unexplained fits

Social Marital difficulties Problems at work Accidents at home or on the roads Delinquency, crime Violence

TABLE 8.20 The CAGE questions* • • • •

Have you ever felt you ought to Cut down on your drinking? Have people Annoyed you by criticizing your drinking? Have you ever felt bad or Guilty about your drinking? Have you ever had a drink first thing in the morning to steady your nerves or get rid of a hangover (Eye-opener)?

'Positive answers to two or more of these suggests a drinking problem.

Table 8.21 Factors associated with risk of alcoholism Factor

Examples of occupations affected

Availability of alcohol

Workers in the drink trade, caterers, business executives Coal miners, seamen, medical students, journalists Company directors, lawyers, GPs Commercial travellers, seamen, oil-rig workers, service personnel Coal miners, doctors, military personnel, actors Doctors, the unemployed Doctors, workers in the drink trade

Social pressure to drink

• • • • • •

Social difficulties, e.g. marital disruption, absenteeism Unexplained accidents Smelling of alcohol at unexpected times Single males over 40, socially isolated females Certain occupations, e.g. publicans, journalists, doctors Unexplained macrocytosis, or elevated liver enzyme values.

Where alcoholism is suspected, a detailed alcohol history should be taken. Vague questions such as 'How much do you drink?' are unlikely to be informative, but positive answers to two or more of the CAGE questions (Table 8.20) suggest a drinking problem. A detailed account of a typical 'drinking day' is required. Alcohol intake is measured in units. One unit equals half-a-pint of beer, a quarter-of-a-pint of strong lager, one 'short' or one glass of wine. Regular intake of six units or more per day puts men at risk of alcohol dependence. Women need half as much (i.e. three units per day) for half as long (e.g. as little as 2 years) for dependency to occur. The patient should be asked when he last had a day without a drink; about the longest period without a drink in the preceding 5 years; about withdrawal symptoms; and about blackouts. A 'drink diary', in which the patient records his drinking behaviour for a week, can be a useful aid to history taking. Raised serum y-glutamyl transpeptidase (y-GT) and increased mean corpuscular volume (MCV) are useful supportive evidence of abnormal drinking and are also helpful in monitoring progress. A blood alcohol concentration may confirm suspicions of recent excessive intake.

Aetiology Mutually interacting genetic, psychological and cultural factors underlie alcoholism. Adoption studies show that adopted children of alcoholics are four times as likely to develop alcoholism as are adopted children of nonalcoholics. The biochemical mediator of an inherited

8

Freedom from supervision Separation from normal social or sexual relationships Strains, stresses, and hazards High or low income Collusion by colleagues

tendency to alcoholism is unknown; however, active alcohol dehydrogenase leading to higher (and unpleasant) acetaldehyde levels has been postulated as protective against alcoholism. Psychological factors operate in two ways in alcoholism: • Imitation learning. An individual's tendency to drink follows that of his subculture, parents and peer group (Table 8.21). • Alleviation of unpleasant feelings. Alcohol reduces unpleasant feelings such as social anxiety and guilt. Each time an individual drinks, the experience of reduction of unpleasant mood ('drowning your sorrows') is reinforced. Family disruption is a common experience in the history of alcoholics. The true aetiology should be distinguished from the reasons for drinking offered by the patient, as the alcoholic's explanation of why he drinks is often a rationalization. The alcohol addict drinks because of his physiological state of addiction.

Physical complications The physical complications of alcoholism (see Chs 12,16, 24) include liver disease (Ch. 16), neurological syndromes (Ch. 24) (including peripheral neuropathy, cerebellar

249

CASE STUDY 8.4 A 50-YEAR-OLD FARMER WITH CONFUSION A 50-year-old separated farmer was brought to A&E by his neighbour because he had been acting strangely for 2 days and was saying things that did not make sense. He was unable to give a clear history, but it seemed that he had separated from his wife 6 months previously and had been living on the farm on his own, although she continued to visit by day to help with the farm work. He had a new girlfriend but she lived in a different part of the county and he felt he could not afford the petrol to visit her. He had no previous psychiatric history and denied illicit drug taking, but was a smoker and admitted to heavy drinking over the past few months. He had been eating poorly. His farm had been in the family since the 14th century, but times had been hard recently and he had had to diversify into holiday letting, and had sold his milk quota. On examination he was a well-built man with obvious signs of weight loss. He was dishevelled and confused. His speech was normal in form but rambling. He did not appear depressed. He believed that he was being spied on by some unknown power, and that his thoughts could be extracted from his brain. His level of attention and orientation seemed to fluctuate, but on systematic examination he was generally orientated although he got the date wrong. Physical examination was essentially normal, apart from the evidence of weight loss and slight generalized tremor and sweatiness. He was admitted to the psychiatric ward for further observation but almost at once appeared to fall and injured the back of his head, with extensive bleeding. He remained unconscious for several minutes. The emergency crash team was called, but his vital signs were normal and consciousness returned. He was very confused, complained of headache,

250

and on examination, although his pupils reacted equally to light there was weakness of elevation of the eyes, which was hard to evaluate as his cooperation with examination was poor. He became more lucid as the day went on, and was able to explain that he had been drinking a bottle or more of vodka a day for several months, but that he had decided to stop drinking some 5 days before admission, and confirmed that he had eaten little for several weeks. Later he had a grand mat fit.

Questions 1. What is differential diagnosis in this man's case? 2. What further investigations are needed? 3. What treatment, if any, should be instituted?

Discussion There are several diagnostic possibilities here. 1. The combination of delusional ideas with fluctuating consciousness makes this likely to be a confusional state ('acute brain syndrome') rather than a functional psychosis such as late-onset schizophrenia ('paraphrenia'), although at this stage no possibilities should be ruled out. 2. A first presentation of a psychotic illness in a 50-yearold man, accompanied by weight loss with a history of smoking, who then develops fits, should arouse the suspicion of organic pathology, such as a cerebral secondary tumour from a bronchial primary, or possibly cerebral non-metastatic manifestations of lung cancer.

3. The head injury was confusing: was his confusional state the result of the head injury and a possible subdural haematoma, or was some intracerebral pathology causing fits? 4. The history of heavy drinking, poor diet, sudden cessation of alcohol intake and possible eye muscle dysfunction makes alcohol withdrawal syndrome ('DTs') or even Wernicke's encephalopathy a distinct possibility - alcohol problems are common and 'common things commonly occur'. An emergency CT scan could help decide whether there was any intracerebral pathology - a haematoma or a tumour - although this investigation is difficult in such a restless and confused subject. Chest X-ray would eliminate the possibility of a lung primary. Liver function tests and haematology would help clarify whether there was any evidence of alcohol-related damage. Interviewing informants such as his ex-wife or girlfriend would help confirm and clarify the history - a good history is worth a thousand investigations in psychiatry. In the absence of any positive evidence of intracerebral pathology a presumptive diagnosis of DTs with possible Wernicke's encephalopathy was made. His confusional state coming on 5 days after stopping drinking, 'rum fits', and eye movement disorder were consistent with this diagnosis. He was given injections of parenterovite and started on a 'detox regimen' of reducing doses of carbamazepine, starting with 400 mg q.d.s. and reducing to 200 mg daily over a 5-day period. He made a good recovery and returned to his farm, determined to decide between his wife and girlfriend, to employ a neighbour to provide him with regular meals, and to abstain from alcohol.

degeneration, myopathy and convulsions), cardiomyopathy, pulmonary tuberculosis, anaemia, hypoglycaemia and pancreatitis. Children of alcoholic mothers may be born with a syndrome of low birthweight and developmental retardation known as the fetal alcohol syndrome.

Psychiatric complications There are three levels of severity at which psychiatric complications of alcoholism occur. In the initial stages the patient may present with minor psychiatric symptoms of panic attacks, anxiety, depressions or sexual difficulties. If withdrawn suddenly from the drug, the alcoholic addict is at risk for delirium tremens and sometimes alcoholic hallucinosis. Prolonged and severe alcohol addiction leads to alcohol-related dementia, including Korsakoff's psychosis (Ch. 24). Delirium tremens (DTs) is discussed on page 1438. Alcoholic hallucinosis Alcoholics may experience auditory hallucinations, usually derogatory or persecutory, in the presence of clear consciousness (consciousness is always clouded in DTs). These usually stop when drinking ceases but occasionally persist, suggesting coexisting schizophrenia. Morbid jealousy Alcohol is a frequent cause of a syndrome of morbid jealousy, in which an individual (usually male) becomes convinced that his spouse is unfaithful. Wernicke-Korsakoff syndrome (q.v.) and alcoholic brain damage are discussed in Chapter 24.

Management The outcome of treatment for alcoholism (Table 8.22) depends less on the nature of the treatment than on the social situation and personality of the patient. Addicts with a stable family and employment and a good personality do best. The impulsive or psychopathic drinker lacking home, family or job has a very poor prognosis for sobriety. The aim of treatment should almost always be abstinence. Any doctor who attempts to help alcoholics must be prepared for disappointment, as relapses ('slips') are common. A combination of limit-setting and acceptance is needed in all of the following treatment approaches.

TABLE 8.22 Outcome of treatment for alcoholism Outcome

1 year (%)

2 year (%)

Improved Still drinking Dead Lost to follow-up

35 45 1

25

19

21

51 3

Psychotherapy A supportive relationship, with either a spouse, a peer (e.g. through Alcoholics Anonymous) or a professional helper, is probably essential. Formal psychotherapy or social help should be conditional on sobriety, on the principle that whatever an alcoholic's problems (and they will be many), drink makes them worse. Wherever possible, the spouse and family should be included in any psychotherapy programme.

8

Inpatient treatment Brief admission for 'drying-out', or longer admission to a therapeutic community with group therapy, can help maintenance of sobriety. Patients who need admission are by definition more disturbed, and the long-term prognosis is therefore often poor. Drug treatment Disulfiram (Antabuse) reacts with alcohol to produce acetaldehyde, producing unpleasant feelings of flushing and nausea. Some patients find it helpful to have this internal 'chemical fence' to discourage drinking. Alcoholics anonymous Alcoholics Anonymous (AA), started in the 1930s by an alcoholic doctor ('Bob'), is a worldwide self-help organization for alcoholics run by ex-alcoholics. It offers a structured personal and group psychotherapy programme aimed at sobriety. An important feature of AA is that it provides a social life that is not based on pubs or drinking. A parallel organization, Al-Anon, offers support to the families of alcoholics. 'Shop-front'/'dry house' programmes The 'skid-row' alcoholic has usually lost everything - job, home, family, possessions, friends - through drinking, and special facilities are needed for this group of patients (about 10% of alcoholics). These include non-threatening 'drop-in' day centres, after-care hostels and long-term medical and social support.

Prevention There is a direct relationship between per capita consumption and the price and availability of alcohol. As the real cost of alcohol has fallen steadily in the UK over the past decade, so per capita consumption and hence alcoholism rates have risen accordingly (Fig. 8.11). Internationally, there is a direct correlation between the price of alcohol and the incidence of cirrhosis and other indices of severe alcoholism. Governments, through pricing policies and licensing laws, can have a direct effect on alcoholism rates. Well-funded education programmes are also needed to alert the public to the dangers of alcohol.

251

FURTHER READING ON ALCOHOLISM AND DRUG ABUSE Chick J 2000 Treatment of alcohol dependence. In: Gelder M, Lopez-Ibor J, Andreason N (eds) New Oxford textbook of psychiatry. Oxford University Press, Oxford Ghodse A 1997 Drugs and addictive behaviour. Blackwell, Oxford Heather N, Robertson I 1997 Problem drinking. Oxford University Press, Oxford Royal College of Psychiatrists and Royal College of Physicians 2000 Drugs: dilemmas and choices. Gaskell, London

MENTAL HANDICAP OR LEARNING DIFFICULTIES Mental handicap can be defined as the arrested or incomplete development of the cognitive (i.e. intellectual) part of the mind. Such a broad definition covers a wide range of individuals: • Severely mentally handicapped, who lack speech, are unable to learn, cannot dress or feed themselves, may be incontinent, and may have stereotyped movements such as rocking or thumb-sucking. The majority are looked after in institutions. • Moderately mentally handicapped, who have a considerable handicap (e.g. very restricted speech) but are able to live in sheltered accommodation in the community. • Mildly mentally handicapped, who have the capacity to learn and to live normal, if somewhat restricted, lives. Their primary problems are social and educational rather than medical. Most can live independently in the community.

MENTAL ILLNESS AND MENTAL HANDICAP It is only recently that mental illness and mental handicap have been properly distinguished. It is now understood that the majority of mentally handicapped people are not psychiatrically ill, and that they should be cared for by the social, rather than medical, services. Nevertheless, the mentally handicapped are particularly vulnerable to mental illness because of both their physical handicap and their emotional and social immaturity. They can develop the full range of mental illnesses, including schizophrenia, manic depression, anxiety states and hysteria. Distress is likely to be expressed behaviourally, for example through withdrawal, temper tantrums, compulsive masturbation or lawbreaking (e.g. shoplifting or self-exposure), rather than verbally.

• A single gene disorder, e.g. untreated phenylketonuria, tuberous sclerosis, mucopolysaccharidoses, untreated galactosaemia and Tay-Sachs disease; • Chromosomal, e.g. Down's syndrome, fragile-X mental retardation, deletion and duplication syndromes; • Traumatic, e.g. birth injury or anoxia. Here, the mental handicap is usually associated with spasticity; • Infective, e.g. maternal toxoplasmosis, cytomegalovirus and rubella; bacterial and viral meningitis. All may cause severe mental handicap; • Other causes include rhesus incompatibility, lead poisoning, and maternal drugs such as anticonvulsants or alcohol. Mild mental handicap usually lacks a specific inherited or environmental cause. It is thought to be a result of polygenie inheritance, often coupled with an impoverished environment.

AUTISM AND ASPERGER'S SYNDROME Autism is one of the most serious of the development disorders. It is characterized by 'autistic aloneness', in which the child is unable to make warm relationships; speech and language disorder, in which conversation as opposed to monologue may be deficient; an obsessive desire for sameness; and bizarre mannerisms. IQ is usually impaired, although the child often has artistic or musical abilities. The aetiology is unknown but is thought to be genetic, and possibly also due to perinatal minor brain injury. There is no known treatment that alters the condition, although individual psychotherapy has been claimed to have achieved remarkable results in some cases. The management usually consists of support for the family and special educational arrangements for the child. As in autism, in Asperger's syndrome there are marked impairments in social interaction, but the development of language and curiosity about the environment is normal. Sufferers usually show intense circumscribed interests or restricted repetitive stereotyped patterns of behaviour. The abnormalities persist into adult life, and most Asperger sufferers are 'loners', who are able to work but rarely able to sustain intimate relationships.

Management The principles of management of adult mental handicap are shown in Table 8.23.

Prevention CAUSES OF MENTAL HANDICAP 252

Moderate and severe mental handicap usually has a distinct cause. This may be:

Primary prevention is exemplified by rubella inoculation and the provision of lead-free paint. Secondary prevention is made possible by genetic counselling and by prenatal diagnosis and screening (see Ch. 3, p. 74).

TABLE 8.23 Principles of management of adult mental handicap • Accurate diagnosis • Treatment of coexisting mental and physical illness • Provision of training facilities, day centres, sheltered workshops and social clubs • Residential provision, e.g. hostels for those whose relatives cannot cope or are not available • Provision of emotional support for carers who, for example, worry about what will happen when they are no longer able to care for their mentally handicapped dependant

FURTHER READING ON MENTAL HANDICAP OR LEARNING DIFFICULTIES Bouras N 1994 Mental health in mental retardation. Cambridge University Press, Cambridge Matson J, Mulick J (eds) 1991 Handbook of retardation. Pergamon, Oxford

PSYCHIATRY AND THE LAW Mental illness may be associated with crime in one of three ways: • Psychosis. The most famous example of a crime being the result of a psychosis is the 19th century McNaughten case, in which a man who had tried to shoot the American President was acquitted 'by reason of insanity'. • Neurotic disorder. Depression, for example, may make a criminal act such as shoplifting more likely ('diminished responsibility'). • Psychopathic personality.

ANTISOCIAL PERSONALITY DISORDER People with an antisocial personality disorder may be described as those with persistent social difficulties (often, but not always, involving law-breaking) in the absence of a psychosis or neurosis. Their crimes, if they do occur, may be as trivial as repeatedly breaking shop windows, or they may be serious, such as murder. In the 19th century these individuals were classified as 'morally insane'; they were not deluded or hallucinated, but they appeared not to know right from wrong, or to be aware of the consequences of their actions. The term antisocial personality disorder is now used in order to avoid the derogatory connotations of 'psychopath'. The concept implies: • A persistent abnormality of interpersonal functioning, extending backwards to a childhood which was often disrupted and poor ('depraved because deprived');

• A major problem of social adjustment; • The absence of another psychiatric disorder.

8

An antisocial individual has trouble not just in relation to himself or to immediate family (although this usually does happen), but also to institutions such as the law, school, the armed services, employment and marriage. He is both dependent on institutions and unable to coexist peaceably with them. There is often associated alcohol and/or drug abuse. The dividing line between normal and abnormal is harder to draw in this than any other psychiatric category and it is not a diagnosis that should be applied lightly. It is also important not to make such patients scapegoats. Antisocial individuals sometimes benefit from psychotherapy (providing firm limits are set), and from help with the social consequences of their disorder, such as homelessness and unemployment.

Dangerousness and risk assessment Psychiatrists are often called upon to assess the dangerousness of an individual so that decisions can be made about whether he should be allowed to live freely in the community. Assessment of dangerousness, which cannot be an exact science and will always to some extent be dependent on clinical judgement, rests on two main criteria: • Past behaviour. Violent or law-breaking behaviour in the past is the most reliable predictor of future trouble. A very careful history, taken in an atmosphere of trust, is essential if this is to be reliably assessed. • Psychological understanding of the causes of specific episodes of violence. Unless caused by organic factors such as epilepsy, violence rarely, if ever, emerges out of the blue. There are always antecedents and triggers that can be identified and which constitute a psychological explanation for the violent episode. These can best be elucidated by reconstructing a detailed 'action replay' of the events leading up to the episode. Thus many people resort to violence when they feel cornered, threatened, or are made to feel inadequate. Substance abuse often plays a major part in lowering thresholds to violence. Risk can be reduced if these factors are known, and steps can be taken to avoid them; conversely, situations likely to increase the likelihood of violence can be identified.

MENTAL HEALTH LEGISLATION Psychotic patients lack insight and may occasionally be dangerous, either to themselves or to others. In the UK, until the 1930s nearly all psychiatric inpatients were legally detained in hospital. Today, less than 10% of patients are detained on a 'Section'. Although compulsory admission is sometimes necessary for the protection of the patient and the public, this power is open to misuse or even abuse. The

253

CASE STUDY 8.5 APPLICATION OF THE MENTAL HEALTH ACT TO A 45-YEAR-OLD MAN WITH HYPOMANIA A 45-year-old man was brought by his son to the accident and emergency department at 11.30 at night. The history was that since losing his job recently he had not slept for several days, had run up huge debts on his credit cards, and had become irritable and aggressive, so much so that his wife had temporarily moved out of the family home. He was complaining of a headache and his son, who was at his wits' end, had persuaded him to come to the hospital to have it investigated. On assessment there was no previous history of formal psychiatric disorder, although he had been prone to mild episodes of depression and withdrawal. Physical examination revealed no abnormality. He was restless, spoke rapidly without sticking to the subject, and was hard to interrupt. He appeared to have no insight into his situation and repeatedly demanded to leave. The provisional diagnosis was hypomania. The junior psychiatrist was worried about him, and felt that he needed further psychiatric assessment. Not surprisingly, he refused the offer of informal admission to hospital. Arrangements were therefore made for him to have a formal mental health assessment. This involved assembling a consultant psychiatrist (who was approved under Section 12 of the 1983 Mental Health Act), a mental health-approved social worker and his GP. During this time the patient was insisting on leaving, but with the help of psychiatric nurses he was distracted enough to stay. Eventually the assessment team arrived, he was duly put on a Section

254

for assessment and admitted, with much protest, to the psychiatric ward, with a diagnosis of hypomanic disorder.

Questions 1. Under what powers can patients be held in hospital against their will? 2. What safeguards are there for both patient and doctor in this situation? The law requires that a strict procedure be followed if someone is to be deprived of their liberty under the Mental Health Act. There is often a period prior to bringing that procedure into play in which the patient may need to be restrained. The legal safeguard here is 'common law', i.e. the right (and duty) of a citizen to protect an individual from causing harm either to others or to himself. In this instance the patient was clearly in danger of harming others or himself. His aggression had resulted in his wife having been physically hurt; equally, he was likely to provoke others to hurt him; his financial profligacy (a typical symptom in hypomania) was also self-injurious. His headache was unexplained, and he needed investigation and overall assessment as well as restraint. Following the mental health assessment he could be given appropriate treatment - in this case tranquillizing drugs such as haloperidol - and if he refused to take them forcible administration was

United Kingdom Mental Health Act of 1983 (currently under review) aimed to provide safeguards against such abuse, and the Sections of the Act of particular relevance to general physicians are described below. In all cases it is a duty of the hospital administration to inform the patient of his Section, and a statutory right of the patient to appeal against compulsory detention.

permissible (although this is rarely necessary). The assessment procedure could begin in a safe environment. The 'Sectioning' procedure involves a clear reading to the patient of his 'rights' including the right of appeal against the Section, to be heard by independent experts in as short a time as possible. In this case the full mental health assessment team had to be brought together. However, had the patient already been formally admitted to hospital the junior psychiatrist would have had the right to impose a 'Section 5.2' on the patient, giving the staff the powers to hold him against his will for up to 72 hours, until a formal assessment could be mounted. Thus the combination of the 'common law' provision, with rights of appeal, gives both doctor and patient some safeguards. The Section 2 imposed on the patient was in fact lifted some 2 weeks later as his clinical condition improved, his insight returned, and he could begin to see that he needed help. Had this not happened, the Section 2 would have expired by the end of 28 days, and were he to have needed further restraint a 'Section 3' treatment order would have been needed. The rules of appeal apply equally to this procedure. The Mental Health Act is currently under review in the UK. It is likely that any decisions about detention for longer than 28 days will be taken out of the immediate hands of clinicians altogether, and referred to a mental health panel comprising an independent psychiatrist, a lawyer and a lay member.

Admission for assessment (Section 2) The patient may be detained for up to 28 days if (a) he is suffering from mental disorder which warrants assessment; or (b) he is a danger to himself or others. The application for such assessment must be made by a social worker or the patient's nearest relative. There has also to be a medical

recommendation from two medical practitioners. One of these must be a doctor officially approved as having special psychiatric experience ('Section 12' doctor, often a consultant psychiatrist); the other should, if possible, have known the patient previously (and is thus usually the patient's GP). In an emergency the patient may be detained for up to 72 hours (Section 4) on the basis of only one medical recommendation, which need not necessarily be by a psychiatrist.

detained patients can still be treated without their consent (e.g. given ECT or phenothiazine drugs), but a second opinion must first be obtained from a doctor appointed by the Mental Health Commission. Patients may not undergo psychosurgery (a brain operation occasionally undertaken in cases of depression or phobic anxiety) if they do not consent to it. Even if they do, a doctor and two nonmedical appointees of the Mental Health Commission must scrutinize the case to ensure that consent has been validly given.

8

Admission for treatment (Section 3) The patient may be detained for up to 6 months. The conditions of the application and recommendation are therefore more stringent, and the rights of appeal greater than under Section 2. The patient must have a defined mental disorder, be thought likely to improve, and be unable to be cared for within the community. Section 3 can be extended for a further 6 months, with the approval of an independent tribunal.

Emergency detention of patients already in hospital (Section 5) An informal patient may be held in hospital for up to 72 hours on the recommendation of the doctor in charge of his treatment without a Section 2 or 3 being instituted. Under Section 5(4), a psychiatric nurse can detain an informal patient already being treated for mental illness for up to 6 hours. Under Section 5(2) a patient in a general hospital who is being treated by a non-psychiatrist can be detained, under the report of the doctor in charge of the case (the 'responsible medical officer' RMO), for up to 72 hours. A psychiatrist should see the patient as soon as possible to decide whether further detention is needed. Section 5(2) cannot be renewed.

Powers of the courts (Sections 36, 37, 38) Under these Sections, mentally ill offenders can be compulsorily admitted to psychiatric units for assessment and treatment, rather than going to prison.

Place of Safety Order (Section 136) Police officers have the power to remove to a place of safety any person whom they find in a public place who appears to be suffering from mental disorder and who is in immediate need of care and control for his own sake or that of others. Once removed, the patient must be medically examined and assessed. The place of safety is often a police station, but may also be a hospital A&E department.

Consent to treatment (Sections 57, 58) An important feature of the Act is a tightening of the procedure surrounding consent to treatment. Compulsorily

FURTHER READING ON MENTAL HEALTH LEGISLATION Faulk M 1994 Basic forensic psychiatry, 2nd edn. Blackwell, Oxford Gunn J, Taylor P 1993 Forensic psychiatry: clinical, ethical and legal issues. Butterworth-Heinemann, Oxford

MEDICINE AND THE EMOTIONS There are two basic types of mental activity. One is pictorial, non-rational and emotional, and concerns the biological fundamentals of birth, attachment, sex, separation and death, as well as art and creativity; the second is rational, verbal and logical and concerns the achievement of specific goals and the solving of particular problems. Freud called these the primary and secondary processes. Non-rational thinking is primary because it precedes rational thinking developmentally: infants and young children think in feelings and pictures. The secondary processes are only gradually acquired. In healthy adult life both coexist. Much of waking life is dominated by secondary processes, but in dreams, daydreams and instinctual activity the primary processes take over. In mental illness the balance of primary and secondary processes may be disturbed: the psychotic, for example, is dominated by primary processes, whereas the patient with obsessional neurosis is trying to apply the false logic of secondary processes to his emotions. Modern medicine is based on scientific rationality. This means that at times doctors may find it difficult to comprehend the emotional reactions of the patient to illness, death or sexuality. For the doctor, the patient who needs a routine hernia operation presents a technical problem to be solved; the patient, however, may at some level be wondering whether he will survive the operation and, if so, whether his sex life will continue. Medical training is predominantly a training in rationality, in learning to set feelings aside. In order to understand the 'irrational' anxieties that trouble patients the doctor has to remain in touch with and respect the primary processes. The following sections on sex, and on death and dying, highlight some of these issues.

255

SEX AND MEDICINE Sex matters for three main reasons: identity, attachment and continuity. Sexual abuse in childhood threatens all three and is an important underlying cause of psychiatric disturbance.

Identity Gender is fundamental. When a baby is born the mother asks 'Is it alright?', then 'Is it a girl or a boy?'. Gender is integral to a sense of identity, and this can be threatened by illness at any of the stages of development. Hypospadism, hirsutism, hernia, delayed puberty, amenorrhoea, infertility, hysterectomy, prostatectomy, diabetes-related impotence are just a few of the many medical conditions that may threaten the sense of identity that is rooted in gender and sexual functioning.

Attachment Human attachment occurs through the body. Even when communication is at a distance, hands are, needed to write a letter, a voice to speak on the telephone. Sexual bonding is a powerful and rewarding form of attachment. The adolescent worries whether his or her body is 'good', 'big' or attractive enough to find a partner. Where sexuality is threatened by illness and disability, the patient fears that the basis of his or her attachment may be undermined: 'Who will want me i f . . . I have a colostomy ... a mastectomy ... am impotent... ?' Psychogenic symptoms such as lower abdominal discomfort, pruritus or back pain may both arise from and lead to sexual difficulties. Such symptoms may draw attention to an underlying sexual anxiety, and at the same time provide the patient with the care and attention that he fears may be lost through sexual failure. Occasionally, guilt about masturbation may be based on the misconception that it is 'wrong' or harmful, and this too may lead to psychogenic symptoms.

Continuity Psychologically, sex and reproduction may serve as an insurance against the fear of extinction. The idea of personal death is more acceptable if there is family continuity, or if an individual feels he has contributed something: that children, memories, ideas or objects will live on in the next generation. Threat to this sense of continuity can lead to feelings of anxiety or depression.

1

256

MCQ 8.7

TABLE 8.24 Common anxieties about sex Adolescence Masturbation Homosexuality Genital inadequacy Young adults Impotence Anorgasmia Infertility Lack of libido

Middle-aged Extramarital affairs Sex and illness Any age Incest Fetishism Non-consummation of marriage

Taking a sexual history Sex, at least in theory, is no longer a taboo subject, but in practice both doctor and patient are likely to feel anxious about some aspects of their sexuality. It is important that the doctor feels reasonably comfortable when asking about a patient's sexual life. All human beings have a sexual life - in the sense of thoughts, fantasies and feelings about sex - regardless of whether or not they are celibate. The doctor needs to be sensitive to the common anxieties about sex and to be aware of those that may apply to a particular patient's age and situation (Table 8.24).

SEXUAL AND MARITAL DIFFICULTIES Divorce and family disruption are becoming increasingly common. More than one in three marriages in the UK now ends in divorce; in the USA the figure is one in two. There is good evidence that in general marital disharmony and divorce are psychologically damaging for children, who are more likely to have emotional disorders in adult life and to underachieve academically than their counterparts from stable families. Nevertheless, single mothers and even same-sex parents are capable of providing a loving and nurturing environment for their children. Partly in response to this, there is now great interest in marital and family therapy. The commonly encountered sexual difficulties are impotence, premature ejaculation and delayed ejaculation for men, and vaginismus and lack of sexual enjoyment (including anorgasmia) for women. Both sexes may suffer diminished libido. Transient difficulties are common; it is only when they are persistent and cause distress that they constitute a problem. Sexual problems may be primary, i.e. present from the onset of sexual life, or secondary, i.e. developing after a period of satisfactory sexual functioning.

Erectile dysfunction This can be defined as the inability to achieve an erection sufficient for satisfactory sexual intercourse.

Aetiology Contrary to popular belief, as many as 50% of cases of impotence are organic in origin. Where impotence is psychologically based, it arises out of anxiety. This may be in the context of an episode of anxiety/depression, or as a manifestation of personality disorder. Organic causes include: • Endocrine causes, such as hypopituitarism and hypogonadism • Diabetes. At least 30% of diabetic men experience impotence by the age of 50 • Vascular disease, such as Buerger's disease • Alcoholism • Drug induced, e.g. diuretics, p-blockers and neuroleptics.

•

•

•

Treatment • If psychogenic, the Masters and Johnson approach outlined below is normally effective. • The underlying cause should be treated whenever possible. • Symptomatic treatment: - Intracavernosal injections of papaverine, papaverine plus an a-blocker such as phentolamine, or alprostadil (prostaglandin E2) can be self-administered and are effective in up to 70% of cases. The main danger is priapism, and erections lasting more than 4 hours should be treated as an emergency, with venous aspiration together with a vasoconstrictor. The drug Viagra is a significant advance over these cumbersome measures. In the presence of sexual desire and stimulation it produces good erections that have a more 'natural' quality. - Mechanical aids such as a vacuum condom are also useful in some cases. The Masters and Johnson approach Current thinking about sexual dysfunction has been greatly influenced by the work of Masters and Johnson, who studied the physiology of sexual response and introduced a new classification and treatment for sexual difficulties. General principles of the Masters and Johnson approach include the following: • Treatment is in couples. Sexual difficulties are often complementary (e.g. premature ejaculation and an-

SUMMARY 24 The Masters and Johnson approach • • • • • •

Treatment in couples Physical examination/history-taking (to reassure they are 'normal') Instruction in sexual arousal Instruction initially not to make love (to dispel anxiety) Graded exercises based on 'sensate focus' Psychotherapy if necessary

•

•

orgasmia), even though one member may present as the person with the problem. The non-presenting partner is given the role of co-therapist. Physical examination and careful history-taking of both partners are required both to rule out organic causes of sexual dysfunction and to reassure the patients that they are 'normal'. There is more to sex than penetration. Sexual arousal, especially in women, depends on a series of steps, arid a man may need to learn that foreplay cannot usually be omitted if his partner is to be fully aroused. Anxiety is seen as a central element in sexual dysfunction. Anxiety is antagonistic to sexual arousal, and a vicious circle can arise in which the patient fears they will fail (for example to maintain erection, not to ejaculate prematurely, or to have an orgasm), which makes them 'try' harder, which makes them less able to relax and respond to sexual stimuli and thus more, rather than less, likely to 'fail'. The aim of treatment is to break this circle. Paradoxically, the couple are initially instructed not to make love, in order to prevent performance anxiety. They are given a series of graded exercises based on sensate focus or stroking, in which they learn to give and receive pleasure. Initially this is non-genital, but later moves on to 'genital sensate focus', i.e. mutual masturbation. It is ironic that whereas masturbation was viewed (wrongly) by 19th century psychiatry as a major cause of insanity, it is now considered important that both sexes are able to masturbate without guilt. Psychological factors in sexual dysfunction may also need treatment, through either individual or marital therapy. Unexpressed anger is a common cause of sexual disharmony: a couple who cannot row may also not be able to make love. Fear of loss of control, ignorance or guilt may also affect sexual pleasure.

8

The results of the Masters and Johnson approach are generally good, with 60-70% improvement rates reported in well-motivated couples. 1

FURTHER READING ON SEX AND MEDICINE Hawton K 1985 Sex therapy: a practical guide. Oxford University Press, Oxford Hawton K 1995 Treatment of sexual dysfunction by sex therapy and other approaches. Br J Psychiatry 167:307-314

DEATH AND DYING Death is an integral part of medicine, but doctors and other medical staff often feel confused by the psychological and ethical questions it poses. Should a terminally ill patient be kept alive as long as possible, or allowed to die in peace? Should every patient with a fatal illness be told of his diagnosis, and if so how should they be told? Can a dying

257

patient be helped to come to terms with death, and how can this happen? What are the psychological reactions in the relatives of dying patients, and what support do they need? These questions are problematic not only because of a lack of training and technique, but also because death is an intrinsically difficult subject which all but the exceptionally compassionate find hard to face. Patients and doctors alike deny the reality of death in order to carry on with life.

REACTIONS IN THE PATIENT The response of the patient to the knowledge that he has a terminal illness can be compared to a bereavement reaction. The patient enters a state of grief for the loss of his own life. The initial reaction is usually one of numbness and shock, a struggle between denial and acceptance. The patient attempts to fight off what is happening and cannot believe it is true, that it is happening to him. This is often followed by a period of severe anxiety, in which the patient becomes dependent on visitors, family and medical staff, and finds it very hard to be alone. A period of sadness and weeping is inevitable. Anger is also a normal feature of this phase, and the patient may become difficult, complaining, ungrateful and demanding. With time, however, the patient gradually becomes calmer, and during this phase the opportunity to talk may be very helpful. This series of reactions may be compressed into a few hours or spread over months, and, as with bereavement, does not follow a neat, orderly course.

TALKING TO DYING PATIENTS Breaking bad news and talking to dying patients is an art that can be learned through watching others, and through discussion of the feelings and difficulties it arouses. The patient can often be helped if staff and family recognize the subsequent reactions for what they are and then let the patient talk about his anger, panic and sadness. When this has happened the patient may feel better.

258

SUMMARY 25 Emotional reactions in terminally

ill patients • • • • • •

Numbness and shock Struggle between denial and acceptance Severe anxiety Sadness and weeping Anger Eventual calmness

The issue of 'to tell or not' is a false alternative. Some patients want to know their diagnosis in great detail; others would rather not know. It is unnecessary and inappropriate to confront all patients with the stark reality of their illness. On the other hand, far more patients than are officially 'told' want to know their diagnosis or know it already. The patient needs space and time in which to discuss his feelings and ask questions. Gentle probing may be needed to help the patient make use of this opportunity.

REACTIONS IN THE FAMILY Like the patient, families also go through an anticipatory grief reaction when a loved one is dying. The death of a child or adolescent is especially painful and unsupportable. Spouses of dying patients are particularly vulnerable. They may, for example, feel angry with their husband or wife for being ill, and feel very guilty about having such unvoiced thoughts. The stress of death may lead to tense and angry outbursts, either within the family or outside it at medical staff. Doctors and nurses should be prepared for this. Family counselling can help with these grief reactions and help to make death and bereavement, when they come, more bearable.

FURTHER READING ON DEATH AND DYING Holland J, Rowland J (eds) 1990 Handbook of psycho-oncology. Oxford University Press, Oxford

9 Infectious, Tropical

and

Parasitic Disease Mark H Wansbrough-Jones and Stephen G Wright

The relationship between humans and microorganisms 259 Principles of diagnosis and management 261 Antimicrobial agents 262 Infection in the immunocompromised patient 269

Rickettsia 305 BACTERIAL INFECTIONS 307 Gram-positive cocci 308 Gram-negative cocci 310 Gram-positive bacilli 311 Gram-negative bacilli 315

Control of infection 271

Mycobacteria 325

Syndromes of infection 273

Spirochaetes 329

VIRUS INFECTIONS 283

Actinomycoses 330

Enveloped double-stranded DNA viruses 283

Madura foot 331

Enveloped single-stranded DNA viruses 289 Non-enveloped double-stranded DNA viruses 290 Enveloped single-stranded RNA viruses 290 Non-enveloped single-stranded RNA viruses 302 Non-enveloped double-stranded RNA viruses 303 CHLAMYDIAL, MYCOPLASMA AND RICKETTSIAL INFECTIONS 303

Nocardioses 331 SYSTEMIC MYCOSES 332 PROTOZOAL INFECTIONS 335 Protozoal infection of the gut 335 Genital tract protozoa 342 Free-living protozoa causing human disease 342 Non-pathogenic gut protozoa 343

the upper hand, particularly over children. Worldwide, infectious gastroenteritis is the most common cause of death in children under the age of 1 year. The causative organisms are similar in underdeveloped and western countries, but in the former there is more frequent exposure from water supplies, methods of food preparation and poor standards of hygiene, including sewage disposal. The outcome is worse because of the severity of dehydration. Table 9.1 shows how the frequency of some infections varies in different areas of the world. The figures illustrate differences attributable to climate (malaria), standards of hygiene (typhoid) and immunization (poliomyelitis, measles and diphtheria). The mortality associated with these diseases is also higher in underdeveloped countries because of a poor nutritional state at the onset of illness and limited access to medical facilities. The pattern of infections is constantly changing. Infectious diseases become less important as a cause of death as the standard of living improves. England in 1860 was much the same as parts of Africa and southeast Asia today with respect to some infectious diseases (Table 9.2). Improvements in standard of living and immunization account for the change in many of these diseases, but sometimes the pathogen itself changes in virulence. Scarlet fever (scarlatina) is no longer the feared disease it was 50 years ago because erythrogenic toxin-producing streptococci have become less virulent. Changes in microorganisms may produce little change in the clinical disease but are of no less significance. The type of meningococcus most commonly incriminated in bacterial meningitis in the UK has for many years been group B, but recently there has been a resurgence of group C infections. Meningococcal vaccine will protect against group A or C infection, but group B meningococci are poor immunogens. Similar antigenic variation occurs repeatedly in influenza virus. Most of the changes in the pattern of infections are gradual. New infections are extremely rare but can have an enormous impact; an example is human immunodeficiency virus (HIV), the cause of the acquired immune deficiency syndrome (AIDS). All of the infections encountered in temperate and developed areas of the world will be found in tropical regions. Certain infections are exclusive to the tropics because of climate, and the presence of animal reservoirs and insect vectors.

Tissue and blood protozoa 343 HELMINTH INFECTIONS 358 Nematodes 359

Chlamydia 303

Trematodes 370

Mycoplasmas 305

Cestodes 375

The relationship between humans and microorganisms is usually one of balanced conflict. In the first months of life encounters with microorganisms are both a threat and a stimulus to the development of complex immune responses. In many parts of the world microorganisms appear to have

THE RELATIONSHIP BETWEEN HUMANS AND MICROORGANISMS Microorganisms that can cause disease in humans are widely distributed in nature, and humans are colonized by vast numbers of organisms. Whether a particular organism causes disease is determined by its natural distribution, its pathogenicity, and the state of the host defence systems. With common pathogens it is likely that immunity will develop early. The age at which this occurs can influence the outcome of infection. At extremes of age the immune

259

TABLE 9.1 Cases of infection reported to WHO from different geographical regions and countries Region or country Americas (all) USA* Europe

Malaria (1993) 984000 1217

50000

Poliomyelitis (1995)

Measles (1994/5)

Typhoid (1980/1)

Diphtheria (1994/5)

0

22988

42160

0 202

286

324

73729 23594

75203

245379

28418

1420 3791

7810

708

0 49483

312

Africa

2590000

643 411

S.E. Asia

3077000

2269

106773

329980

674000

226

110776

145

614

0

7763

265

13

E. Mediterranean

W. Pacific England and Wales

292000

1992

(Data supplied by Communicable Diseases Surveillance Centre, Colindale.) All figures are to nearest 100. *AII figures for USA are for 1995. AII figures for England and Wales are 1995 (except malaria - 1993)

TABLE 9.2 Deaths from infection in England in 1800 fall age groups) Cause

Number

Measles Scarlatina (scarlet fever) Smallpox Diphtheria Typhus Whooping cough Cholera Tuberculosis

9557 9 681 2749 5212 13012 8553 327 58564

system is either immature or in decline, and the patient is highly vulnerable. Paradoxically, during childhood many viral infections are silent or mild, but the same infection in an adult may be severe and complicated. This may reflect a difference in the vigour with which inflammatory responses are generated.

PATHOGENICITY

260

Microbial pathogenicity - that is, the potential of a microorganism to cause disease - is determined by its ability to enter the host and colonise, to penetrate tissues, to evade host defences and to damage tissues. The virulence of an organism is its degree of pathogenicity. An organism is described as highly virulent if a small number of microbes can cause severe disease. Skin is an effective barrier to most organisms and they usually enter either through a breach or via more delicate mucosal surfaces, such as those of the oropharynx and respiratory tract. Attachment to the cell membrane is essential for viruses, and bacterial colonization is also often facilitated by a specific attachment mechanism. Examples are Vibrio cholerae or Escherichia coll in the intestine, and

E. coli in the urinary tract. In each case a determinant on the bacterial surface interacts with a specific receptor on epithelial cells. Some pathogenic organisms cause disease at the site of colonization by producing toxins (V. cholerae),but most traverse the epithelium and their further success depends on their ability to avoid host defences. In the first encounter between human and organism innate immunity is the only defence. In the case of bacteria this centres on phagocytosis, and some bacteria are equipped with capsules that confer resistance to phagocytosis. Interferons play a major part in first-line resistance to viruses, but viruses differ both in their ability to stimulate interferon production and in their susceptibility to its effects. The mechanisms of disease provoked by bacteria are complex, but a distinction can be made between effects caused by exotoxins (tetanus or diphtheria toxins, for example) and those resulting from release of bacterial components such as endotoxins.

SUSCEPTIBILITY TO INFECTION

The effects of even virulent microbes are curtailed by an immune response. Infections therefore typically result from a small number of identifiable situations: • First encounter with a virulent organism, in sufficient numbers, entering by an appropriate route; • Breach of normal anatomical barriers, e.g. by surgery, injury or indwelling lines; • Impairment of innate immune responses; • Impairment of adaptive immune responses. When anatomical barriers are breached by surgery, tissues that are usually sterile are exposed to normal microbial flora at the adjacent site. Similarly, accidental injury often results in exposure to environmental organisms, such as Clostridium tetani present in the soil. Impairment of immune responses leads to susceptibility to infection by organisms of low pathogenicity as well as

9

FIG. 9.1 Incubation periods and periods of communicability of common infectious diseases

to virulent organisms. Infection by organisms of low pathogenicity is called opportunistic. Patients with impaired immunity may therefore present with frequent infections, infection caused by opportunist microorganisms, unusually severe disease caused by infection, or repeated infection by the same organism.

• Incubation period: time from contact with microorganisms to onset of symptoms (Fig. 9.1) • Prodromal illness: minor symptoms preceding the definitive illness • Overt disease • Resolution (convalescence).

This approach will often lead to a simple management plan. It may also be necessary to establish the source in the community. In some cases the doctor has a legal obligation to notify the disease (see Control of infection, p. 271).

This pattern is often influenced by the immune state of the patient and may be altered by drug therapy. Some symptoms, such as sore throat and fever, are more commonly associated with infection than with other disease processes. The pattern of fever may be helpful, but patients may say that they have had a fever when they have not measured their temperature. They may mean aching in the limbs, sweatiness feeling hot or cold, shivery, or true rigors. Of these, a rigor is a good indicator. This is a sensation of coldness accompanied by shivering (associated with a rise in temperature) followed by a hot sensation and marked sweating. Although no one symptom is specifically associated with infection there are many symptom complexes that are highly suggestive of an infective aetiology. For example, although headache has many causes, the combination of headaches with fever, photophobia and vomiting strongly suggests infection of the meninges (meningitis).

Is the disease caused by infection and, if so, where is the site of infection?

Examination The findings on examination may be helpful in three ways:

A careful history and examination will provide important indicators that the patient is suffering from infection.

• Some infections are associated with signs that are specific or which suggest only a short list of differential diagnoses. Examples are the rash of chickenpox or measles, neck stiffness caused by acute meningitis, or the characteristic cough of whooping cough. • More commonly a combination of signs is highly suggestive of an infection. For example, exudative tonsillitis, generalized lymphadenopathy and splenomegaly together suggest infectious mononucleosis; fever, relative bradycardia, splenomegaly and rose spots suggest typhoid. • Physical signs may indicate the site of infection. Consolidation in the lungs suggests pneumonia; loin tenderness is associated with pyelonephritis; inflammation,

PRINCIPLES OF DIAGNOSIS AND MANAGEMENT The principles guiding the clinician in the diagnosis and treatment of an infective illness can be considered by asking some simple questions: • Is the disease caused by infection and, if so, where is the site of infection? • What is the microorganism responsible? • Which drugs can eliminate the organism, and will treatment benefit the patient?

History Three aspects of the history are likely to be helpful: • Pattern of the illness • History of contact with patients with a similar illness, or types of behaviour that confer a risk of infection • History of travel to places where particular infections are known to be common. Illnesses caused by infection tend to have a form that is reproduced in most individuals and which is characterized by the following features:

261

volume of fluid permits. The yield will be higher and a Gram stain can be performed. • Think of culture before allowing a tissue specimen to be immersed in fixative such as formalin. • Ensure that specimens for culture are transported to the laboratory without delay. Microorganisms must be nurtured, and all too often they die waiting for transport, or are overgrown by contaminant organisms. Specimens for viral or fungal culture often require special collection techniques and the laboratory should be consulted before obtaining the specimen.

Which drugs can eliminate the organism from the infected site, and wifl treatment benefit the patient?

FIG. 9.2 111ln white cell scan in a paracolic abscess At 24 hours after the intravenous injection the activity is localised, indicating the accumulation represents an enclosed abscess rather than inflamed bowel. (Reproduced by kind permission of Dr Seth Saverymuttu.)

swelling and fluctuance beneath the skin suggest an abscess.

What is the microorganism responsible? A careful history and examination may point to a diagnosis of infection and can be specific for a particular infection. More frequently, infection is strongly suspected but the site and the organism responsible are unknown. Imaging techniques are often helpful in finding the site of infection, e.g. ultrasound or CT scanning can demonstrate an abscess in soft tissues. Gallium or technetium (99mTc) leukocyte scans frequently demonstrate a soft tissue abscess. In the latter technique the patient's leukocytes are labelled with 99mTc and reinjected (Fig. 9.2). Labelled leukocytes enter sites of acute inflammation, such as an abscess, surgical wounds and the bowel wall of patients with inflammatory bowel disease. Before appropriate treatment can be started the right samples must be collected and sent to the laboratory. The laboratory can give most help if the correct samples are taken and are accompanied by a brief summary of the patient's illness. The drug history is particularly important, as antibiotic in the sample may inhibit the growth of bacteria in vitro. There are simple rules for the collection of samples:

262

• Avoid contamination of the sample with irrelevant organisms by careful cleansing, e.g. cleanse skin with antiseptic before taking blood for culture, and wear gloves during venepuncture. • When infected fluid is to be cultured (e.g. pus) send a fluid sample in a sterile bottle rather than a swab, if the

Some guidelines for the choice of antimicrobial drugs are given on page 263. Whether treatment is to be started immediately or after the results of laboratory tests are available depends on the severity of the disease and the likely consequences of delay. There is virtue in waiting until the precise diagnosis is known if the clinical situation permits, and this applies particularly when antibiotics have been given before admission to hospital. Acute meningitis and septicaemia are obvious exceptions because the patient may die before a final result is available, but even then it is desirable that the correct specimens are obtained before treatment is started. Valuable preliminary information can be obtained by Gram staining and microscopy of specimens such as sputum or pus. Direct testing for microbial antigens may increase in importance for disease caused by a limited range of organisms, such as acute meningitis. Treatment is not always beneficial to the patient, even when an antimicrobial drug with appropriate activity can be delivered to the site of infection in adequate concentration. Faecal excretion of salmonellae may be prolonged by treatment with some antibiotics.

ANTIMICROBIAL AGENTS The choice of antimicrobial drugs is central to the management of infection. General measures should not be neglected: these include use of analgesics, attention to nutrition and hydration, and cleaning, debridement or drainage of infected areas. The term 'antimicrobial agent' can be applied to any substance that is active against microorganisms. Strictly, an antibiotic is a substance produced by a microorganism which inhibits the growth of another. In this chapter 'antibiotic' will be used (inaccurately) to refer to antibacterial drugs, whether synthetic or naturally produced. In this section the dose of antimicrobial is not normally given, as this often depends on the nature of the particular infection being treated, and the reader should refer to sections on the management of specific diseases. Only the more severe side-effects are noted here.

edge of antibiotics, including the known sensitivity of particular organisms to them. The latter are not constant, and so treatment is often started on the basis of an intelligent guess and modified later according to laboratory results. Bacterial sensitivity is tested using discs impregnated with antibiotic and measuring the size of the ring of inhibition on culture plates. Other factors are also important in the choice of antibiotic, such as absorption and penetration to the site of infection. Whenever possible the antibiotic chosen should have a spectrum of antibacterial activity limited to a few organisms, as the broader the spectrum the greater the chance of side-effects owing to disturbance of the normal microbial flora. The choice between a bactericidal and a bacteriostatic drug is not usually critical but it may be important when the organisms are fairly inaccessible, as in infective endocarditis, or when host defences are impaired (see Table 9.6). Combinations of antibiotics are used for one of three purposes:

Choice of appropriate drugs The selection of a suitable antimicrobial agent is fairly straightforward when the microorganism responsible is known (Table 9.3), but often this is not the case and a sensible guess has to be made as to the organisms responsible. In the case of a urinary tract infection in a previously healthy young woman, for example, faecal organisms such as E. coli would be the most likely pathogens, and this would guide initial therapy. The information required to make an appropriate choice can be summarized as follows: • • • • •

Probable organism(s) Probable antimicrobial sensitivities of the organisms Ability of the chosen drugs to reach the site of infection History of allergy to drugs Interaction of these drugs with any other drugs the patient is taking • Side-effects of the chosen drug • Cost of the drug.

• To obtain a spectrum of antibacterial activity that could not be achieved with a single agent (e.g. treatment of septicaemia when the causative organism is not known); • To take advantage of synergy between drugs (e.g. during the initial treatment of infective endocarditis); • To prevent the emergence of resistant organisms (e.g. treatment of tuberculosis).

An example of the process of decision is shown in Figure 9.3. This logical sequence assumes a considerable knowl-

TABLE 9.3 Common sensitivities of some bacteria to antibiotics Bacteria

1st choice

2nd choice

Streptococci (except anaerobic) Staphylococci Neisseria meningitidis Neisseria gonorrhoeae Gram-positive bacilli Escherichia coli (urinary tract) Pseudomonas aeruginosa

Penicillin

Erythromycin

Flucloxacillin Penicillin Azithromycin Penicillin Trimethoprim

Erythromycin Cefotaxime Penicillin

H. influenzae (respiratory infection) Gram-negative (septicaemia)

Route of administration The ideal drug would give adequate serum levels after oral administration, whatever the state of the patient. However, an ill patient often cannot swallow tablets and absorption from the gut may be unreliable. The choice between intravenous (i.v.) and intramuscular (i.m.) administration is determined by the form in which the drug is available and its pharmacokinetics, the discomfort of intramuscular administration, and practical issues such as accessibility of veins and other drugs being administered. Topical antibiotic treatment is sometimes useful in the eye (p. 304), on the skin (Ch. 10) or in the respiratory tract (e.g. in cystic fibrosis, see p. 651), but the development of resistance is a problem.

Amoxicillin

Piperacillin and gentamicin (usually requires a combination) Co-amoxiclav

Ceftazidime and gentamicin

Cefotaxime

Ciprofloxacin Piperacillin and gentamicin

Erythromycin

Duration of treatment The aim of treatment with antibiotics is to eliminate infec-

Antibiotic

Probability of sensitivity

Penicillin

>95%

+ +When meninges inflamed (iv administration)

Allergy

Low

Cefotaxime

>95%

+ + (iv administration only)

Allergy in 10% of penicillin-allergic subjects

High

1st choice in Western world if penicillin allergy

Chloramphenicol

>95%

+ + +(Oral or iv administration)

Very rare marrow aplasia

Low

1st choice in Third World country

Penetration

9

Toxic effects

FIG. 9.3 Choice of antibiotic in a patient with acute pneumococcal meningitis

Cost

Decision 1st choice if no allergy

263

tion, when this is achievable, or to reduce the number of organisms to a point at which host immunity can do the rest. Surprisingly little is known about the duration of treatment needed to achieve these aims in most infections, and there is considerable unsupported dogma regarding the importance of completing a prescribed course of antibiotics. The usual argument for finishing a course of treatment is that early discontinuation encourages the emergence of resistance. However, resistance rarely arises during treatment except when the bacterial population already contains resistant mutants, as in streptomycin therapy for tuberculosis. It is most likely to occur in chronic and heavy infections (e.g. Pseudomonas aeruginosa infection in cystic fibrosis), and with particular antibioticbacterial groupings, such as quinolones or fucidic acid in staphylococcal infections, cephalosporins in some Gram-negative infections, and rifampicin in infections by many bacterial species. There is good evidence about the duration necessary for treatment of tuberculosis, urinary tract infection, gonorrhoea and infective endocarditis, but virtually none about upper and lower respiratory tract infections, which account for 75% of community prescriptions. Opinions vary considerably about the appropriate time to treat in most other infections, and often the only advice available is to be guided by the clinical response.

MECHANISM OF ACTION OF ANTIBIOTICS

TABLE 9.4 Examples of major pathogens which have developed new resistance patterns

Antibiotics act in one of the following ways:

Organism

Resistance to

• Inhibition of bacterial cell wall synthesis (penicillins and cephalosporins); • Inhibition of particular aspects of bacterial cell function: protein synthesis (aminoglycosides, chloramphenicol, tetracyclines, macrolides, clindamycin); DNA replication (rifampicin, quinalones); microbial enzymes (sulphonamides, trimethoprim, isoniazid).

Enterococcus Strep, pneumonias N. gonorrhoeas Staph. aureus

Ampicillin/gentamicin Penicillin Penicillin Methicillin (flucloxacillin)

Resistance to antibiotics There are two major ways in which resistance occurs: • The bacterium produces an enzyme which inactivates the antibiotic. The best-known example is B-lactamase, which alters the B-lactam ring of penicillins and cephalosporins (Fig. 9.4). Aminoglycosides and chloramphenicol are other drugs broken down by bacterial enzymes. • The bacterium may alter such that the antibiotic is rendered ineffective. This occurs either by a change in the permeability of the bacterial cell wall or by an alteration at the target site of the antibiotic.

264

FIG. 9.4 Structure of B-lactam ring

Resistance tends to emerge by natural selection in a population of bacteria exposed to antibiotic. This may be acquired by mutation (leading to a single base alteration), but more often it is the result of incorporation of new

genetic material into the bacterial nucleus or the transfer of genetic material from other organisms in the form of extrachromosomal DNA as plasmids (episomes) and transposons. There is an increasing problem with emergence of resistance among bacteria (Table 9.4) which is largely related to the use of antibiotics acting as a selection pressure. In hospitals methicillin-resistant Staphylococcus aureus (MRSA) colonize the skin of patients admitted with other problems and can be difficult to clear. For the majority of patients these organisms are harmless, but they can cause deep infection in certain, groups, such as orthopaedic and cardiac surgery patients, and there is controversy about whether hospitals should isolate all colonized patients and attempt to clear their skin of infection, or just take steps to protect high-risk groups. On a wider scale, antibiotics introduced in poor communities for the treatment of common conditions become useless after a few years because resistance develops.

This has happened when trimethoprim was given for the treatment of dysentery in the past, and it is an emerging problem with antibiotics used to treat gonorrhoea and Strep. pneumoniae infections. It is most likely to happen when antibiotics are given in inadequate doses in a random manner, which happens when they are available over the counter. Antibiotic policies have been devised to avoid the problem but they are most easily applied in hospitals, and even there they are not wholly successful. So far pharmaceutical companies have been able to produce new compounds to keep ahead of the problem, but it is important that we learn more about the ways in which resistance emerges if it is to be controlled.

ANTIBIOTICS IN COMMON USE B-Lactams

These are the most commonly used antibiotics and have in common the B-lactam ring (Fig. 9.4). There are two major groups, the penicillins and cephalosporins. Penicillins Benzyl penicillin or penicillin G is a highly potent bactericidal natural penicillin which was the forerunner of all the other penicillins. Long-acting forms, procaine and benzathine penicillins, allow longer intervals between doses. Its disadvantages are that it can only be given parenterally, it has little activity against Gram-negative organisms, and many staphylococci are resistant to it. Semisynthetic and synthetic penicillins have been developed to circumvent these problems. Doses are now usually expressed in mg (600mg = 106 units = 1 megaunit). Standard adult dose: 300 mg 6-hourly; high dose 600 mg 4-hourly. Phenoxymethyl penicillin is a natural penicillin which is acid resistant and can therefore be taken by mouth for the treatment of minor infections with streptococci which are highly sensitive. Absorption is variable, and so it should not be used for more serious infections. Standard adult dose: 500 mg qds. Broad-spectrum penicillins, such as ampicillin and amoxicillin, have useful activity against Gram-negative organisms, excluding Pseudomonas species. They are well absorbed from the intestine. Standard adult doses: amoxicillin 250 mg tds; ampicillin 0.25-lgqds. B-Lactamase-resistant penicillins, such as flucloxacillin and methicillin, are used mainly to treat penicillinaseproducing staphylococci. Standard adult dose: flucloxacillin 250 mg qds. Co-amoxiclav is a combination of amoxicillin and the B-lactamase inhibitor clavulanic acid, which makes the amoxicillin active against resistant strains of Staph. aureus, H. influenzae, E. coli and many Bacteroides and Klebsiella species. Standard adult dose: 375 mg (containing amoxicillin 250 mg) tds. Antipseudomonal penicillins: the ureidopenicillins (azlocillin and piperacillin) are distinguished by their activity against Pseudomonas species and have a broader

spectrum of activity against Gram-negative organisms. Standard adult doses: azlocillin 2 g 8-hourly; piperacillin 150mg/kg/day in divided doses, up to 16g daily in severe infections. Toxicity. The most common problem is hypersensitivity causing rash and anaphylaxis (see p. 96). This is an allergic reaction which develops 10-14 days after first exposure, but much sooner on re-exposure. Its severity varies from mild to fatal according to the reactivity of the patient, the dose of drug and the route of administration. There is crossreactivity between penicillins, and to some extent with cephalosporins. Ampicillin is particularly likely to cause rash in patients with infectious mononucleosis.

9

Cephalosporins Cephalosporins have a similar structure and mode of action to penicillins but modifications to their structure have produced a range of drugs with a broader spectrum of activity. Examples are cefaclor, which has a broad spectrum of activity, including some against staphylococci, and can be taken orally (standard adult dose: 250 mg tds); cefotaxime, a parenteral drug with a very broad spectrum excluding Pseudomonas (standard adult dose: 1g 12hourly, or up to 4 g 8-hourly); and ceftazidime, similar to cefotaxime but including Pseudomonas activity (standard adult dose: 1 g 8-hourly, or up to 2g 8-hourly). Ceftriaxone has a similar spectrum of activity to cefotaxime but is unusual in that it is given only once a day (standard adult dose: 1 g daily, or up to 4g daily; doses above 1 g given in several sites). These drugs are expensive and their use should be part of a defined antibiotic policy. Their sideeffects are similar to those of penicillins. Other B-lactams Imipenem is a carbapenem which differs from other (3lactams in its remarkable spectrum of antibacterial activity and resistance to hydrolysis by most B-lactamases and other enzymes. It is administered in combination with cilastatin, which inhibits its renal metabolism by inhibiting a kidney enzyme (standard adult dose: 500mg 12-hourly, up to 750mg 12-hourly). CNS side-effects such as convulsions and confusion limit its usefulness, and meropenem is less likely to cause such toxicity (standard adult dose: 500mg 8-hourly, up to 1 g 8-hourly). These are reserve drugs, only prescribed when conventional drugs have failed. Aztreonam is a monolactam with a spectrum of activity limited to some aerobic Gram-negative bacteria (standard adult dose: 2g 12-hourly). Macrolides Erythromycin is an extremely useful drug with a macrolide structure. It can often be used as an alternative to penicillin when the patient is allergic. In addition it is active against Haemophilus influenzae, Mycoplasma, Legionella, Bordetella pertussis and Campylobacter (standard adult dose: 250mgqds). Newer macrolides such as clarithro-

265

mycin and azithromycin are less likely to cause gastrointestinal side-effects and can be given in shorter courses than erythromycin (standard adult doses: clarithromycin 250 mg bd; azithromycin 500 mg daily for 3 days, or as a single 1g dose treatment for genital chlamydial infection). Chloramphenicol Chloramphenicol is a cheap drug, well absorbed and distributed, with useful activity against Gram-positive and negative cocci and H. influenzae. Its major drawback is the very rare but usually fatal induction of marrow aplasia. It is also difficult to use safely in babies. Standard adult dose 50mg/kg/day in four doses. Tetracyclines Tetracyclines were widely used to treat exacerbations of chronic bronchitis until resistance became common. Now their main use is for the treatment of chlamydial infection, but they are also active against Mycoplasma, Brucella, rickettsiae, Coxiella and Borrelia. They should not be used in pregnant women and children under 9 years old because they cause staining of teeth and sometimes dental hypoplasia. Doxycycline is less prone to cause diarrhoea than other tetracyclines but occasionally causes oesophageal ulceration. Standard adult dose: 200mg on day 1, then l00mg daily. Aminoglycosides Aminoglycosides such as gentamicin, amikacin and tobramycin are used in serious Gram-negative infections. They are ototoxic and nephrotoxic. Blood levels have to be monitored carefully, especially when there is renal impairment, as they are excreted purely by the kidneys. Peak blood levels are measured 15 minutes after intravenous or 1 hour after intramuscular administration, and trough levels immediately before the next dose. Standard adult doses: gentamicin 2-5mg/kg/day in two doses; amikacin 15mg/kg/day in two doses; tobramycin 3 mg/kg/day in three doses. Sulphonamides and trimethoprim Sulphonamides were the first synthetic antimicrobial drugs, but their use is now limited by the development of resistance and by the frequency of troublesome sideeffects such as nausea, rash (Stevens-Johnson syndrome at worst) and bone marrow toxicity. Sulfamethoxazole is still widely used in combination with trimethoprim (cotrimoxazole; standard adult dose 960mg bd, but see p. 344 for treatment of Pneumocystis carinii infection) because they were thought to be synergistic, but it now seems that trimethoprim alone is just as effective in the treatment of

1

266

MCQ 9.1

Gram-negative urinary infection. Trimethoprim can cause thrombocytopenia and megaloblastic anaemia. Standard adult dose 200mgbd. Clindamycin Clindamycin is a lincosamide with activity against Grampositive organisms, including resistant staphylococci, but also against anaerobic organisms. It achieves good penetration into bone and can be used to treat osteomyelitis. Common side-effects include rash and diarrhoea, rarely with pseudomembranous colitis. Standard adult dose 300mgqds. Metroindazole Metronidazole is an excellent drug for the treatment of anaerobic infection. It is cheap when administered orally or rectally, and the more expensive intravenous preparation is rarely indicated. It is also the first choice in treatment of Giardia lamblia, amoebic disease and Trichomonas vaginalis. Serious side-effects are rare, but nausea and abdominal pain are common. Standard adult dose 400mgtds. Sodium fusidate Sodium fusidate is highly bactericidal against staphylococci, including penicillin-resistant ones, but resistance develops during treatment unless another drug is given concurrently. It is reserved for treatment of serious infections such as septicaemia and osteomyelitis. It is necessary to monitor for hepatotoxicity. Standard adult dose 500 mg tds. Vancomycin Vancomycin is reserved for serious penicillin-resistant staphylococcal infection or for the treatment of pseudomembranous colitis. Serum levels must be monitored, as it is ototoxic and nephrotoxic. Standard adult dose 1g infused over 100 minutes 12-hourly, or 125mgqds orally for pseudomembranous colitis. Teicoplanin is a similar drug which is slightly less toxic. Standard adult dose 400 mg 12-hourly for three doses, then reduce to 200400 mg daily. Quinolones Nalidixic acid has been used exclusively in the treatment of urinary infection by Gram-negative organisms, as adequate levels are only found in the urine. Standard adult dose 1 g qds. Resistance to this group of drugs is rare, as they are not susceptible to plasmid-mediated resistance. The newer generation. of 4-fluoroquinolones includes ciprofloxacin, which is well absorbed from the intestine and widely distributed. Standard adult dose 500 mg bd. This group contains the only orally administered drugs active against Pseudomonas, and they have become increasingly important in the management of gut infections and severe systemic infections, including typhoid. Major toxicity is uncommon.

Rifampicin Rifampicin, a synthetic derivative of rifamycin, is central to the treatment of tuberculosis but is also very active against staphylococci, streptococci and E. coli. Its major toxicity is to the liver. Standard adult dose 300 mg bd, but see p. 645 for the treatment of tuberculosis. Quinuprastin/dalfoprastin Quinuprastin/dalfoprastin is a new streptogramin recently licensed in the UK and the USA which is active only against Gram-positive bacteria, including methicillinresistant Staphylococcus aureus, vancomycin-resistant Enterococcus faecium (but not E. faecalis) and penicillin/ macrolide-resistant pneumococci, and its use should be restricted to treatment of such infections. Administration is recommended through a central venous catheter, as peripheral lines tend to develop complications. Linezolid Linezolid, a new oxazolidinone antibiotic is active against Gram-positive bacteria which are resistant to other antibiotics. It can be given orally or intravenously (600 mg every 12 hours by either route). 1

ANTIFUNGAL DRUGS Most fungal infections of the skin are treated with topical preparations, but cuticular infection requires prolonged systemic therapy. Visceral infections also require systemic therapy. Amphotericin B and nystatin Amphotericin B and nystatin are polyene antibiotics produced by Streptomyces species. They exert a fungistatic effect by binding to sterols in the fungal cell membrane, causing leakage of cell components. Amphotericin B acts synergistically with the unrelated compound flucytosine, probably by increasing permeability to flucytosine. Amphotericin B can be administered intravenously, after which there is wide tissue distribution but poor penetration to brain and meninges. The plasma half-life is 24 hours and it is excreted very slowly in the urine, 40% of the infused dose appearing in the urine during the following week. Amphotericin B has a broad spectrum of activity and is used against the agents of candidiasis, aspergillosis, histoplasmosis, blastomycosis, coccidioidomycosis, cryptococcosis and some organisms causing mucormycosis. Minor symptoms following infusion are common, and a wide variety of symptoms may be unpleasant enough to require suppression with analgesics, antiemetics etc. More serious side-effects occur when there is impairment of renal function, which is exacerbated by the resultant rise in plasma drug levels. Renal and haematopoietic toxicity are the main problems. Therapy has to be interrupted to allow a fall in plasma level. It is given in a dose of 0.25 mg/kg body weight

intravenously over 6 hours, increasing to 0.8 mg/kg according to tolerance. Treatment is withdrawn intermittently as required to maintain the plasma level at l-1.5mg/L. A less toxic liposomal form of amphotericin B is expensive and its use is limited to febrile neutropenic patients and to those unable to tolerate the native form. Standard adult dose of liposomal amphotericin: Img over 10 minutes test dose, then 3 mg/kg/day in one dose. Neither amphotericin B nor nystatin is absorbed from the intestine but they are useful as topical or mucosal preparations, particularly for treatment of Candida infections. Nystatin is available as a suspension or pastilles for oral infection, cream and pessaries for vaginal infection, and cream or ointment for skin infections.

9

Flucytosine Flucytosine is a synthetic fluorinated pyrimidine which enters fungal cells and inhibits metabolism by interfering with DNA and RNA synthesis. It can be given in combination with amphotericin B for treatment of systemic fungal infections, but is rarely used now. Standard adult dose 200 mg/kg/day in four doses, each infusion over 20-40 minutes. Imidazoles Ketoconazole, miconazole and clotrimazole are imidazole drugs. Most are available only topically and have activity against dermatophytes and Candida. Imidazoles have no place in the treatment of systemic fungal infections. The mode of action is not fully understood, but imidazoles interfere with synthesis of sterols and other fungal membrane lipids, resulting in defective cell wall synthesis. In vitro they enhance killing of Candida by leukocytes. Ketoconazole is an oral preparation which is useful for treatment of severe mucosal infections, such as oesophageal candidiasis which is not responsive to topical treatment. The major toxicity is drug-induced hepatitis, which can progress to irreversible hepatic failure. Standard adult dose 200 mg daily. Triazoles Triazoles are a new group of antifungal agents, including fluconazole and itraconazole, which are useful as oral agents for treating systemic mycoses. Both are effective in treating oesophageal candidiasis using very short courses, e.g. a single dose of fluconazole 150 mg. Prolonged treatment in patients with immunodeficiency often results in emergence of resistant strains of Candida, but at present cross-resistance between itraconazole and fluconazole is uncommon. Fluconazole (but not itraconazole) penetrates into the CSF and is useful for treatment of cryptococcal meningitis (dose 400mg, then 200mg i.v. daily). Sideeffects are rare but there is a small risk of hepatotoxicity. Standard adult dose of itraconazole: 100 mg daily. Griseofulvin and terbinafine Griseofulvin, produced by Penicillium griseofulvum, is

267

absorbed from the gastrointestinal tract and deposited in keratin, where it inhibits the growth of dermatophytes. It is relatively non-toxic and can be used in extensive dermatophyte infections which are unresponsive to topical treatment. It is particularly useful for nail and scalp infections (p. 389), but has to be given for weeks or months. Serious side-effects, such as bone marrow and neurological disorders, have been reported occasionally after griseofulvin use. Standard adult dose: 500mg daily. Terbinafine, an allylamine, can also be used for these infections and may be effective after shorter courses. Standard adult dose: 250 mg daily. 1

ANTIVIRAL TREATMENT Considerable progress has been made in synthesizing agents which specifically inhibit an essential step in virus replication without causing host toxicity. This has been achieved: • By using an agent which is activated by the virus, thus localizing its action within infected cells; • By using an agent which inhibits virus metabolism specifically. Usually this is achieved when the agent inhibits a viral enzyme. Drugs have been developed which are active against herpes viruses, human immunodeficiency virus, hepatitis B and C, influenza viruses, enteroviruses, respiratory syncytial virus, and some viruses which cause viral haemorrhagic fever.

Drugs active against herpes group viruses Aciclovir, the most widely used agent currently available, produces its specific effects using both of these mechanisms. It is inactive until phosphorylated to a triphosphate by viral thymidine kinase in infected cells. In this form it produces its main antiviral action by inhibiting viral DNA polymerase, essential for synthesis of viral DNA. Aciclovir triphosphate is a more powerful inhibitor of viral than of host DNA polymerases. Aciclovir seems to be a safe drug but is expensive. Its use is restricted to situations in which it is clearly beneficial (Table 9.5). Newer drugs such as famciclovir and valaciclovir are similar in their spectrum of activity and toxicity but differ in absorption, so that famciclovir can be given less often than aciclovir, and valaciclovir achieves a higher serum concentration than aciclovir after oral administration. During acute viral infections much of the inflammation is caused by the host response in the later part of the illness, 1

268

MCQ 9.2

TABLE 9.5 Treatment of herpes virus infections with aciclovir Virus

Infection

Route

Herpes simplex virus

Encephalitis Neonatal infection Primary genital infection Keratoconjunctivitis Prophylaxis: bone marrow transplantation Shingles in immuncompromised patients Chickenpox in immunocompromised children Chickenpox in neonates Chickenpox in adults Possible Chickenpox pneumonia

IV IV 0 T 0 IV IV IV IV 10

Varicella zoster

IV = intravenous; 0 = oral; T = topical

and antiviral treatment must be started early to be effective. Antiviral treatment does not affect virus integrated into the host genome, and treatment of the primary herpes simplex infection does not influence subsequent viral reactivation. Ganciclovir, an antiviral drug licensed for treatment of cytomegalovirus (CMV) infection, is activated by phosphorylation in a similar way to aciclovir, but depends on human cellular enzymes for this process (Fig. 9.5). Because phosphorylation occurs in infected and normal cells, toxicity is greater than that of aciclovir and it causes bone marrow suppression. Foscarnet, a phosphonate, also inhibits CMV DNA polymerase and has a different toxicity from ganciclovir.

Antiretroviral drugs There are three main classes of antiretroviral drugs: nucleoside and non-nucleoside inhibitors of HIV reverse transcriptase, and a third class that inhibit HIV protease. At least three drugs are used in combination to limit the development of resistance and, once started, treatment is normally maintained for life (see Ch. 11).

Ribavirin (tribavirin) Ribavirin is another nucleoside analogue that is active against a wide spectrum of DNA and RNA viruses, but its clinical use is confined to treatment of severe respiratory syncytial virus infection causing bronchiolitis, some of the viral haemorrhagic fevers such as Lassa fever, and hepatitis C virus infection in combination with interferon-a. For RSV infection it is given by inhalation.

Interferons Interferon-a, which is produced by T lymphocytes, has been shown to inhibit replication of hepatitis B and hepatitis C viruses, possibly by inhibiting virus assembly. It has many side-effects, including flu-like syndrome early in

9

FIG. 9.5 Mechanisms of action of ganciclovir Virus replication in host cells induces the enzymes (guanosine kinase and phosphorylase). Because the same enzymes are involved in phosphorylation of ganciclovir, infected cells have higher concentrations of ganciclovir triphosphate. This is the activated form, which blocks viral replication,

therapy, lethargy throughout treatment, bone marrow toxicity and depression, and is only effective in a small proportion of patients, but it is the only substance known to have any influence on these infections at present.

FURTHER READING ON ANTIMICROBIAL AGENTS

TABLE 9.6 Some microorganisms causing infection in immunocompromised patients Type

Microorganism

Remarks

Deficiency

Bacteria

Gram-positive, e.g. Pneumococcus Gram-negative, e.g. H. influenzae Mycobacteria, e.g. avium intracellulare

Encapsulated

Ab/Neut/S

Encapsulated

CMI

Both tuberculous and non-tuberculous Reactivation

CMI

Encephalitis Virulent Septicaemia/ pneumonia Mucosal Systemic Meningitis Pneumonia Cerebral Gut Gut Systemic Systemic

Ab Ab/CMI Ab/Neut

Lambert H P, O'Grady F, Finch R, Greenwood D 1996 Antibiotic and chemotherapy, 7th edn. Edinburgh: Churchill Livingstone.

INFECTION IN THE IMMUNOCOMPROMISED PATIENT An immunocompromised patient is defined as one in whom any part of the innate or adaptive immune response is defective. Defects of the immune system are physiological (neonates and the aged), and either primary or secondary. Primary immune deficiency diseases are rare and an account is given in Chapter 4. Secondary immune deficiency is more common and of increasing importance. Causes of secondary immune deficiency are: • Malignant disease and its treatment; • Drug-induced chronic immune suppression, e.g. in organ transplantation or chronic inflammatory diseases, such as systemic lupus erythematosus (SLE); • Infection: the most severe infection is HIV (Ch. 11), but many other viruses cause a milder deficiency of cellmediated immunity (CMI); • Splenectomy; • Chronic diseases, e.g. chronic renal failure, malnutrition, protein deficiency, diabetes mellitus. These conditions cause varying combinations of neutropenia and impairment of humoral and cellular immunity. Deficiencies of antibody or neutrophils are associated with infection by encapsulated bacteria, yeasts and fungi. Defi-

Viruses

Yeasts Fungi Protozoa

Helminths

Herpes, e.g. herpes simplex virus, varicella zoster virus, cytomegalovirus Echo Live vaccines, e.g. measles Candida Aspergilius Cryptococcus Pneumocystis Toxoplasma Cryptosporidium Giardia lamblia Malaria Strongyloides

CMI

CMI CMI CMI CMI/Ab CMI CMI CMI/Ab Ab/S CMI

Ab = antibody; Neut = neutrophil; CMI = cell-mediated immunity; S = post-splenectomy

cient CMI causes infection by a wider range of pathogens, including Gram-positive and -negative bacteria, viruses, yeasts, fungi, protozoans and helminths (Table 9.6). In all these situations immunization with live viral vaccines is dangerous, as the attenuated organism may behave as a pathogen. The pattern of infection in immunocompromised

269

SUMMARY 1 Management of the neutropenic patient with fever

TABLE 9.7 Organisms causing pneumonia in immunocompromised patients

• Take samples for culture: blood, urine, stool, i.v. line tip, etc. as appropriate • Start treatment after taking samples, guided by any previous culture results • Examples of empirical treatment regimens: - gentamicin plus piperacillin or ceftazidime - If no response within 48 hours, switch to - vancomycin If no response, or recurrence, particularly if previous antibiotic courses given - amphotericin B or fluconazole for Candida septicaemia • If chest X-ray shows pulmonary infiltrates, add erythromycin initially for atypical pneumonia. Consider a wide range of organisms (including M. tuberculosis) and viruses (e.g. CMV) if there is no response to antibiotics

Bacteria Gram-positive, e.g. Strep. pneumoniae Gram-negative, e.g. H. influenzae, Klebsiella pneumoniae Viruses Cytomegalovirus, herpes simplex, varicella zoster Measles (giant cell pneumonia) Protozoa Pneumocystis carinii Fungi/yeasts Candida Aspergillus

NOSOCOMIAL INFECTION patients is unusual in several respects. Bacterial infections in splenectomized patients are sudden in onset and rapidly progressive, fulminant pneumococcal septicaemia being the most common. Cellular immune deficiency causes viral infection to be more severe, prolonged and widely disseminated than is usual. Fever in the neutropenic patient (see Summary 1) During periods of neutropenia the onset of fever should be taken to indicate serious infection unless there is an obvious alternative explanation. 'Blind' treatment with antibiotics is started immediately after the collection of specimens, including blood cultures. Blood cultures are negative in over half of such fever episodes, but the choice of antibiotics is guided by positive results in which Gram-negative bacteria, including Pseudomonas species, predominate, and a combination of a broad-spectrum penicillin with activity against Pseudomonas (e.g. piperacillin) and gentamicin is often used (p. 265). New pulmonary infiltrates on chest radiograph New pulmonary infiltrates may be due to infection with a wide range of possible pathogens (Table 9.7). Sputum induction, bronchoscopy and lavage, and transbronchial biopsy may provide diagnostic information. Invasive procedures have a risk in neutropenic and thrombocytopenic patients.

270

Reactivation of herpes viruses Patients with impaired CMI may suffer severe herpes virus infection. Shingles (varicella zoster) in patients with Hodgkin's disease tends to involve more than one dermatome, and there may be dissemination of virus causing widespread chickenpox lesions on the skin and occasionally in viscera. Within affected dermatomes lesions are dense and heal slowly. Reactivation of herpes simplex virus (HSV) may cause extensive perioral ulceration and occasionally visceral dissemination. CMV reactivation causes pneumonia in renal transplant recipients.

Hospital patients may be particularly susceptible to infection and hospital bacteria are often resistant to common antibiotics. Susceptibility is increased in postoperative patients, patients with bums, patients on ventilators, in malnourished patients, and in disease states that compromise immunity specifically. Hospital-acquired infections are known as nosocomial and are a particular problem on intensive treatment units. Nosocomial infection is an increasing problem and the resistance pattern of bacteria is constantly changing. The most important factors influencing the likelihood of cross-infection are the use of handwashing, the ratio of nurses to patients, the number of visits from doctors and the space between patients. The bacteria associated with particular situations can sometimes be predicted, e.g. staphylococci and streptococci are common causes of surgical wound infections, and Pseudomonas frequently infects burns. In ventilated patients difficulties arise in deciding whether bacteria such as Pseudomonas are merely colonizing the respiratory mucosa or whether they are pathogenic. Antibiotic treatment of colonising organisms is undesirable since the local flora then change to more resistant organisms. With increasing use of antibiotics there has been a resurgence of outbreaks of nosocomial infection with methicillin (and flucloxacillin)-resistant Staphylococcus aureus (MRSA). MRSA colonize the skin and nose of normal or diseased subjects, but they are particularly persistent in patients with any form of skin lesion. In most cases MRSA do not cause disease, but occasionally vulnerable patients can develop septicaemia or local suppuration which necessitates the use of vancomycin. Because MRSA spreads rapidly from person to person it is important to isolate infected patients from other vulnerable patients, and to screen staff for carriage. Colonized subjects are treated with topical antiseptic preparations and, in the case of staff, removed from work until repeated swabs are clear. Vancomycin-resistant staphylococci have been isolated and the increasing use of this antibiotic has led to a problem with resistant enterococci. There is some evidence

that antibiotic cycling in intensive care units - that is, changing the first-choice antibiotics every few months limits the rate of increase in resistance among common nosocomial bacteria. For example, resistance to gentamicin in enterobacteriaceae such as Klebsiella pneumoniae and Serratia marcescens increases when gentamicin is used, but decreases when amikacin is substituted. However, there is no consensus on the best policy at present. The use of antibiotics must be controlled both in hospitals and in the community. Most hospitals have antibiotic control policies which recommend the antibiotic to be used in a particular situation and restrict the administration of some drugs to special situations. Policies for antibiotic prophylaxis before surgery need to be reviewed regularly to keep ahead of the development of bacterial resistance.

CONTROL OF INFECTION Containment of outbreaks of infection is required when susceptible subjects may develop a serious illness if they become infected. It is not always desirable to contain infection. Chickenpox, a mild disease in normal children, is usually severe in adults and so outbreaks among schoolchildren should not be controlled. In a hospital where non-immune patients are liable to develop severe manifestations, however, it is important to prevent crossinfection to non-immune individuals (about 10% of the adult population). Isolation of patients in single rooms greatly facilitates control of infection. Other measures are determined by the mode of transmission and the degree of infectivity of the organism. Transmission is by: • Respiratory secretions (e.g. measles, pulmonary tuberculosis) • Excretions and secretions (e.g. gastroenteritis, hepatitis A, typhoid) • Skin contact and fomites (e.g. MRSA). Respiratory transmission can be prevented by having the patient and/or the attendants wear appropriate masks. Organisms transmitted in respiratory secretions usually infect others through the upper or lower respiratory tract, and gowning and handwashing are of minor importance. When excretions and secretions are infectious the most important containment measure is handwashing. When close contact is necessary with a highly infectious patient, such as a patient excreting Shigella in the faeces, gowns and gloves are worn. Special arrangements are needed for the disposal of excreta, contaminated dressings and contaminated clothing and bedding. When microorganisms are found in significant number on the skin, in fomites and in dust around the patient, as when MRSA colonize patients in hospital, additional precautions must be taken, including the use of overshoes together with gowns and gloves.

TABLE 9.8 Notifiable diseases under the Public Health (Infectious Disease) Regulations 1988 (UK, October 1988) Acute encephalitis Acute poliomyelitis Anthrax Cholera Diphtheria Dysentery (amoebic or bacillary) Food poisoning Leptospirosis Malaria Measles Meningitis meningococcal pneumococcal Haemophilus influenzae viral other specified unspecified Meningococcal septicaemia (without meningitis) Mumps

9

Ophthalmia neonatorum Paratyphoid fever Plague Rabies Relapsing fever Rubella Scarlet fever Smallpox Tetanus Tuberculosis Typhoid fever Typhus fever Viral haemorrhagic fever Viral hepatitis Hepatitis A Hepatitis B Hepatitis C other Whooping cough Yellow fever

AIDS is not a statutorily notifiable disease but doctors are urged to participate in a voluntary confidential reporting scheme Source: PHS Communicable Disease Surveillance Centre

In many countries some infections are notifiable (Table 9.8), that is, the medical attendant is obliged by law to report the case to the authority that controls the area in which the patient lives so that steps can be taken to trace contacts. Notification also allows surveillance of the incidence of these infections. Most patients have ceased to be an infectious risk by the time they are discharged from hospital, but intestinal pathogens often remain in the stools for several weeks. Carriers of such pathogens are advised to avoid handling food for others, to wash hands regularly and to avoid towel sharing. Following notification, stool samples will be checked by the local authority until they are pathogen free.

IMMUNIZATION Active immunization Active immunization is the administration of antigen to induce a protective immune response in the host. Antigens used include crude preparations of killed organisms, extracellular products of organisms, and live attenuated organisms. The aim is to induce an optimal response at the site where infection normally occurs, but it has proved difficult to achieve this at mucosal surfaces. Currently, oral poliomyelitis vaccine is the most successful intestinal vaccine because the attenuated virus induces a prolonged and renewable response in the gut. Oral vaccines have been developed against typhoid and shigel-

271

Contraindications to immunization

TABLE 9.9 UK immunization schedule (2001) Age

Immunization

Type of vaccine

Route

2 months

Diphtheria Tetanus Pertussis Polio H. influenzae type b Meningococcus type C As above Measles, mumps, Rubella (MMR)* Diphtheria Tetanus Polio MMR BCG Tetanus Polio Diphtheria Rubella

Toxoid Toxoid Killed Live attenuated Capsular antigens Conjugate vaccine As above Live attenuated

IM or SC IM or SC IM or SC 0 IM or SC IM As above IM

As above

As above

Live attenuated

ID

As above

As above

3 and 4 months 12-18 months 5 years

10-14 years School-leaving Adults

Diphtheria Polio Tetanus Hepatitis B Influenza

For seronegative women For non-immunized " Selected groups

IM

*MMR introduced into UK in October 1988 Detained in UK until MMR established IM = intramuscular; SC - subcutaneous; 0 = oral; ID = intradermal

losis but they are less effective. Oral cholera vaccine has been developed using the adjuvant properties of cholera toxin B subunit, and tests in Bangladesh have yielded promising results. In contrast, the cholera vaccines available using heat-killed organisms are ineffective. A similar local immune system operates in the respiratory tract and further studies may help to develop local vaccines against respiratory pathogens. The immunization schedule in the UK in 2001 is shown in Table 9.9. Mumps vaccine was introduced as part of the routine childhood immunization schedule in 1988. The policy of immunizing only females against rubella has also been changed, as there is still a notable incidence of congenital rubella. BCG (bacille Calmette-Guerin) immunization for protection against tuberculosis is more important in children of racial groups with a high incidence of the disease, and it may be given to neonates who are at risk of developing meningitis. In North America the incidence of tuberculosis was not high enough to justify its routine use. In contrast, BCG vaccination of neonates is a common practice in the tropics.

1

272

MCQ 9.3

• Live vaccines should not be given to anyone with untreated malignant disease or with any form of immunodeficiency, including those receiving immunosuppressive therapy, or to pregnant women. • Live vaccines should not be given within 3 weeks of another injected live vaccine. • Coincidental severe febrile illness may limit the response to immunization. Immunization should therefore be postponed, but not forgotten. • Significant documented allergy to any component of the vaccine may be a contraindication to its use. • Rubella vaccine should not be given in pregnancy, and pregnancy should be avoided for 1 month after immunization against rubella. 1

Passive immunization Passive immunization is the administration of antibody preparations to prevent or ameliorate infection. Pooled human immunoglobulin is the crudest form, and was commonly used to protect travellers against hepatitis A virus infection. More defined preparations of hyperimmune globulin contain high levels of antibodies to particular organisms, usually viruses, which are not present in sufficient concentration in pooled human immunoglobulin to provide protection. Examples are hepatitis B immunoglobulin given following needle-stick injuries, varicella zoster immunoglobulin used in neonates born to mothers who have recently developed chickenpox (p. 284), and tetanus immunoglobulin used for patients at risk of developing tetanus. Most current antibody preparations are derived from human serum. Passively administered antibody has a short half-life, giving protection for only a few weeks. Treatment has to be given soon after exposure to be effective.

PREVENTIVE MEASURES FOR OVERSEAS TRAVEL OR RESIDENCE The question that is most commonly asked prior to overseas travel to or residence in the tropics or subtropics is 'what vaccines are needed?' In fact, the main disease threats do not come from vaccine-preventable conditions, but from trauma, alcohol-related illness, sun exposure, unprotected sexual contacts and infectious diarrhoeal disease. Malaria is the main threat that is amenable to preventive measures.

Protection of skin The skin must be protected against biting insects. This is important not only for the prevention of malaria, but also for protection against viral, rickettsial, protozoal and filarial diseases transmitted by insect vectors. Exposure to ticks occurs mainly in rural bush areas and

so appropriate clothing and footwear is needed to prevent ticks getting on to the skin. Evening and night hours are times for maximum frequency of biting by many insects, and at these times the skin should be covered as much as possible wearing shirts with long sleeves and long trousers or skirts. Insect repellent (diethyltoluamide (DEET)containing compounds with formulations designed to have a long retention time on the skin) is applied to exposed areas, including the forearms, lower legs, ankles and feet, and face and neck. Elasticated wrist and ankle bands can be impregnated with DEET and worn for the same purpose. Mosquito nets should be used in malarial areas if doors and windows are not screened. Impregnation of bed nets and curtains with permethrin enhances protection through its residual insecticidal action. It is also useful to spray bedrooms with insecticide last thing at night. Schistosomiasis is acquired by contact between skin and slow-flowing or still water containing cercariae. Swimming and paddling in fresh water are best avoided in endemic areas.

Protection against malaria Measures to prevent mosquito bites (see above) and chemoprophylaxis against malaria are the sole means of preventing malaria in the traveller. The traveller must be told that there is no vaccine against malaria and that none of the vaccines received prior to travel is of any benefit in this regard. Details of chemoprophylaxis against malaria are given in Table 9.36, p. 352.

Immunization for travellers or overseas residents A range of vaccines is available and the commonly used ones are listed in Table 9.10. Poliomyelitis is still common in many areas of the world, and so adequate vaccination should be assured in all travellers. Meningococcal vaccine is most often recommended for backpackers in subSaharan Africa. It is occasionally a requirement for travel to specified countries, e.g. Saudi Arabia for the Hajj pilgrimage. Yellow fever vaccination is essential for those visiting tropical Africa, the Caribbean and South America. Oral poliomyelitis and yellow fever vaccines should be given either together or separated by an interval of 3 weeks. Active hepatitis A vaccination is available and is usually used in preference to passive immunization, although it is more expensive. Hepatitis B vaccination is particularly recommended for those at particular risk, i.e. healthcare and hospital laboratory workers, although increasingly backpacking travellers are having this vaccine. Pre-exposure rabies vaccination is needed by those who may have any contact with animals, but again increasing numbers of tourists undertaking overland travels are having pre-exposure vaccination. The intradermal route allows more economical use of the vaccine. Tetanus vaccination is needed for any who may be at particular risk of trauma, e.g. military personnel, construction workers etc.

TABLE 9.10 Immunization for travellers to the tropics

9

Isolated visit to main tourist areas Typhoid Polio Hepatitis A Yellow fever (for some locations) Extensive travel (backpacking, trekking, gap-year students) Typhoid

Polio Hepatitis A & B Rabies Meningococcal disease A & C (sub-Saharan Africa) Yellow fever (West Africa, Central & South America, Trinidad) Frequent visitor (business or holiday) or resident Typhoid Hepatitis A & B Rabies Yellow fever (West Africa, Central & South America, Trinidad) Meningococcal disease A & C Note: these groupings of vaccines are guidelines only and need modification for individuals in relation to their potential exposure during proposed travels. Although on a risk/cost analysis some of these recommendations may seem excessive, they give the traveller coverage for vaccine-preventable conditions endemic in many areas. All travellers should be up to date with childhood vaccines.

Typhoid vaccination is a sensible precaution for many areas, but those vaccinated should be told that having the vaccine does not mean that they are 100% protected and can be careless about sources of water and food. Although the whole-cell killed vaccine has the greatest efficacy it also has the greatest incidence of side-effects and is not now obtainable. The choice lies between two vaccines: (a) the Ty21a oral vaccine containing the epimerase-deficient strain of S. typhi, which has the appeal of stimulating an immune response at the site of entry of the natural pathogen, and (b) the vaccine based on the Vi antigen of S. typhi. Cholera vaccination is not effective.

Protection of the gut Gastrointestinal infection is the most common affliction faced by the traveller. Boiled water used to make tea and coffee or left to cool in a covered, clean container is safe to drink. Bottled drinks are safe. Water from the tap or water used to make ice cubes may not be safe. Tincture of iodine (2%) can be used to sterilize water, adding 0.5mL per litre and allowing it to stand for 30 minutes. Iodine is more potent than chlorine-based sterilizing tablets.

SYNDROMES OF INFECTION SEPTICAEMIA Septicaemia is uncontrolled bacterial infection in the blood. Unlike septicaemia, transient bacteraemia is fairly common

273

in normal subjects. Septicaemia is a much more serious condition which is likely to lead to death if untreated.

Aetiology A wide range of bacteria can cause septicaemia, but they can be divided into those that do so as part of a primary infection by a single organism acquired in the community, and those that cause septicaemia as a complication of another process, such as abdominal sepsis following surgery, when there is often a mixed infection. Community-acquired infections are usually recognized as part of a specific illness, as when Strep, pneumoniae septicaemia complicates pneumonia, meningococcaemia with meningitis, or E. coli septicaemia with pyelonephritis. The other kind of septicaemia is usually seen in hospitals, often in a debilitated patient, and the likely bacteria can often be predicted from the clinical situation.

Pathogenesis Whole bacteria and substances released by bacterial lysis trigger a complex series of events. A major component is endotoxin, the lipopolysaccharide component of Gramnegative bacterial cell walls, of which lipid A is thought to be responsible for most of the toxic effects. Lipid A can activate complement and Hageman factor and can cause the release of cytokines, including tumour necrosis factor. Important pathological consequences ensuing from activation of the potent inflammatory pathways are hypotension, increased vascular permeability and consumptive coagulopathy. Less attention has been paid to the effects of Grampositive organisms, but activation of inflammatory processes can certainly be triggered by cell membrane components of these organisms. An example is alternate pathway complement activation (Fig. 4.1, p. 80) by teichoic acid found in Staph. aureus membranes.

negative septicaemia, but also occurs with Gram-positive septicaemia. For this reason the general term septicaemic shock is usually used. Prolonged hypotension leads to acute renal failure, and disseminated intravascular coagulation causes bleeding. Acute respiratory distress syndrome (p. 697) frequently complicates hospital-acquired septicaemia. The diagnosis is based on clinical assessment, but blood cultures, urine cultures, wound swab and any aspirate from an infected site should be obtained at the time the patient is seen, so that antibiotic treatment can be started without undue delay. The antibiotic regimen can be changed if necessary when culture results are available and the response to treatment has been assessed. Blood cultures should be taken whether the patient is febrile or not.

Management Treatment is urgent because of the high mortality. Antibiotics and circulatory support are the mainstays of management. The use of plasma expanders and selective vasoactive drugs is discussed on page 731.

Antibiotics The initial choice of antibiotics is based on a logical guess at the bacteria responsible (p. 263). As few antibiotics as possible are given to achieve the necessary cover. At present a common choice is a combination of a broadspectrum penicillin, such as ampicillin, with an aminoglycoside such as gentamicin. One problem is that many patients are in incipient renal failure, which enhances the toxicity of gentamicin. Newer drugs, such as cefotaxime and piperacillin, are active against Gram-positive and Gramnegative bacteria and lack the toxicity of gentamicin, but they are more expensive. If anaerobic septicaemia is suspected metronidazole is often used, and for hospitalacquired staphylococcal infection a B-lactamase-resistant penicillin, such as flucloxacillin, is added. Pseudomonas infections are treated with a combination of drugs, e.g. azlocillin and gentamicin. 1

Clinical features and diagnosis The manifestations depend on the age of the patient, the illness being more featureless and more lethal at extremes of age. Signs are usually non-specific, but there are signs in the skin suggestive of the diagnosis, such as purpuric lesions associated with meningococcaemia and pustular lesions in staphylococcal septicaemia. Usually there is a rapid deterioration in the patient's condition, including a change in mental state. Fever is often present, with tachycardia and a bounding pulse, but as the condition progresses the patient becomes hypothermic with a rapid thready pulse and peripheral vasoconstriction. The latter situation is often considered to be typical of Gram1 274

MCQ 9.4

PYREXIA OF UNKNOWN ORIGIN When a patient has an elevated body temperature for a period of longer than 3 weeks without a cause being found the diagnostic problem is called pyrexia (or fever) of unknown origin (PUO). A febrile patient may lose the normal diurnal variation in body temperature or may sometimes have an exaggerated diurnal pattern, with a rise during the night causing drenching night sweats (common in tuberculosis and lymphoma) followed by a rapid fall. Many patterns of fever are described but they are rarely helpful in diagnosis.

Causes of PUO The most common causes of PUO vary according to the

age of the patient, and published incidences vary with the type of institution at which the problem is studied. Infection is usually the most common cause in children, whereas infection and neoplastic disease are equally important in adults. Connective tissue disease can present with fever at any age: Still's disease and SLE are commoner causes in children and young adults, and polymyalgia and giant cell arteritis more frequent in the elderly. Infection Most bacterial infections cause an illness which is clinically manifest within 3 weeks, but there are some exceptions. Bacterial endocarditis can cause low-grade illness, particularly in the elderly. Most cases can be diagnosed by blood culture, but occasionally blood cultures are repeatedly negative. Infection in the biliary tract may be associated with only minor clinical signs and little abnormality of liver function tests. Intracellular bacteria, such as Brucella and some Salmonella species, can cause recurrent septicaemia, but this can usually be diagnosed by blood culture. Deep-seated abscess is a possible cause of PUO, and modern imaging techniques are helpful in locating the collection (p. 262). Tuberculosis is still a major cause of PUO in developing countries and in some temperate countries such as the UK. Often the granulomatous lesions of tuberculosis are too small to cause pulmonary radiological abnormalities, and the site of infection is often outside the lungs. In recent years tuberculosis has become a common feature of HIV infection. Viral infections rarely cause PUO (within the definition), with the important exception of HIV. The history should therefore enquire into sexual practices, intravenous drug abuse and transfusions. Fungal infections rarely cause PUO without a predisposing cause, but they may do so as a complication of immunosuppressive disease or therapy. An exception is histoplasmosis.

Neoplasia Lymphomas frequently produce a prolonged period of fever before any other manifestation emerges. Leukaemias can also cause PUO, but this is sometimes caused by a complicating infection. Many occult solid tumours can present as PUO, among them hypernephroma, pancreatic carcinoma, intestinal tumours, bronchial carcinoma, ovarian tumours and sarcomas. Fever is particularly likely when there are hepatic metastases. Atrial myxoma is a rare benign tumour (p. 582) which occasionally produces an illness similar to bacterial endocarditis. Connective tissue (autoimmune) disease Most diseases in the category of connective tissue disease can present with PUO before the characteristic disease pattern emerges. SLE, rheumatoid arthritis and Still's disease (including the adult form), polyarteritis nodosa, temporal arteritis and polymyalgia rheumatica should be considered depending on the clinical setting.

Miscellaneous Granulomatous conditions. Granulomatous conditions include sarcoidosis, granulomatous hepatitis and Wegener's granulomatosis. Sarcoidosis often lacks overt signs in the early phase of the disease and should be considered, particularly in ethnic groups with a high prevalence, such as West Indians. Factitious. One of the most odd forms of human behaviour is the habitual feigning, or the deliberate inducement, of illness. Factitious fever is caused in two ways:

9

• In the first the fever disappears when the temperature is measured by someone other than the patient, with the patient under continuous observation. The patient is either saying that fever exists when it does not, or is manipulating the thermometer. • In the other, more serious form the patient induces true fever by self-injury or self-infection. The methods used are diverse, but faeces are the usual source of organisms and infection is induced by rubbing or injecting organisms into the skin or blood. Suspicion is aroused when mixed and varying organisms which could be faecal are found in an unexpected site, or when infection persists despite treatment that would normally be effective. The condition is more common in young females than in males, and the deceit may be helped by some medical knowledge. There are sometimes personal problems, which suggest that this behaviour is a plea for attention or help. When suspicion grows that fever is factitious it is wise to avoid directly confronting the patient with the possibility. Usually it will be denied, and sometimes confrontation can precipitate a psychiatric crisis, even resulting in suicide. It is preferable to attempt to solve the associated problems if they can be identified and to create a situation in which it is difficult for the patient to continue to provoke sepsis, e.g. by removing intravenous catheters or covering inflamed sites. The patient should be made to realize that the manoeuvres are understood, if possible without a direct accusation. The best efforts to resolve factitious fever often fail, and patients either return intermittently or present themselves to another medical team. Bowel disease. Inflammatory bowel disease is usually associated with incriminating symptoms such as weight loss and diarrhoea, but these may be slow to emerge. Whipple's disease, caused by the periodic acid-Schiff-positive Grampositive bacterium now known as Tropheryma whippelii, usually presents with polyarthralgia and diarrhoea in middle-aged men, but initially there may be only PUO and abdominal pain (see p. 792). Intestinal lymphoma may also present with the latter symptoms. Drug reaction. Once thought of, drug reaction is a straightforward cause of PUO. The possibility that the fever is caused by a drug is tested by withdrawal. Multiple pulmonary emboli. When frequent small embolic events occur, pyrexia may be the sole manifestion before the pulmonary circulation is greatly compromised.

275

SUMMARY 2 Causes of pyrexia of unknown origin in UK Frequency is given in parentheses • • • • •

Infection (30%) Neoplasia (25%) Connective tissue (autoimmune) disease (20%) Miscellaneous (12%) Undiagnosed (13%)

Other causes. Other occasional causes of PUO are alcoholic hepatitis, familial Mediterranean fever and occult haematoma. There are some young women whose normal evening temperature runs above the normal range by up to 0.5°C. This usually comes to light after an infection with slow recovery. Once it is established that there is no underlying pathology the patient can be reassured. Tropical causes are discussed below.

Diagnosis A careful history and examination of the patient is essential, and it is important to return to the patient repeatedly to go over parts of the history and to look for emergent physical signs. Simple screening tests which may need to be repeated at intervals include full blood count, ESR and CRP, routine biochemistry of blood and urine, cultures of blood, urine and other relevant samples, and chest radiograph. Additional investigations are guided by the clinical findings. There are two approaches. One is to perform tests which are more or less specific for particular diseases, e.g. exclude autoantibodies in autoimmunity or antibodies to particular pathogens. The other is to investigate a particular site which seems to be generating symptoms or signs. The latter is more often helpful, and it is in this area that advances in imaging techniques change the methods for diagnosing PUO. Ultrasound is particularly useful for abdominal examination, both in the detection of solid tissue abnormalities, such as tumours of the kidney or pancreas, lymph node enlargement or infiltration of the liver, and in locating collections of pus (subphrenic, perinephric, intrahepatic etc.). CT scanning supplements, and is in some situations superior to, ultrasound, for example when intestinal gas limits the views on abdominal ultrasound. MR scanning is often more sensitive than CT and is clearly superior for detecting lesions in soft tissue and bone. CT scanning and ultrasound are used to guide the needle during tissue biopsy. The use of 99mTc-labelled leukocytes is described on page 262.

MCQ 9.5

276

Even if no localized lesion has been found by imaging techniques, certain biopsy procedures are sometimes rewarding. If the liver enzyme tests are abnormal percutaneous liver biopsy should be considered, as diffuse granulomatous or infiltrative diseases such as sarcoidosis or lymphoma may be detected. Marrow aspiration and biopsy may reveal cryptic tuberculosis, or lymphomatous or carcinomatous infiltration. The marrow and liver biopsy should be cultured for Mycobacterium tuberculosis, bacteria including Brucella fungi and, where relevant, leishmania. Patience and persistence are essential in managing PUO. Neoplasia is the major cause of death among PUO patients, and so the exclusion of treatable neoplastic disease such as lymphoma is a priority. In some cases there is evidence to support a diagnosis of bacterial infection although the infection cannot be defined, and it may be justified to give blind antibiotic treatment. This should be aimed at the type of infection suspected. For example, a Staph. aureus osteomyelitis would require two antibiotics, such as flucloxacillin and fucidic acid, whereas an appendix abscess could be treated with ampicillin, gentamicin and metronidazole to cover E. coli, enterocci and anaerobic bacteria. In other circumstances a therapeutic trial of treatment for M. tuberculosis may be required. In most studies about one-fifth of cases remain unsolved despite the best efforts. 1

The approach to the patient with fever from or in the tropics Febrile illness is common in the tropics, or after travel to the tropics. The causes encompass infectious and noninfectious causes. Infections include those peculiar to tropical regions and those common worldwide (such as urinary tract infection, pneumonia) (Table 9.11). Clinical features A detailed history is essential. This should include accurate recording of the dates of travel, duration of stay and countries visited, with information on the places involved (cities, rural areas, beach resorts etc.) and the reasons for travel. Information about sexual contact during travel is also important. The history should also include details of vaccination prior to travel and malaria prophylaxis,

TABLE 9.11 Leading causes of imported fever (Hospital for Tropical Diseases) Malaria P. fdlciparum P. vivax P. ovale Viral (presumed) Dengue (proven) Bacterial gut infections

42% 57% 9% 4% 25% 6% 5%

9

TABLE 9.12 Incubation periods for some tropical diseases Type of organism

Short (less than 2 weeks)

Intermediate (2-6 weeks)

Long (more than 6 weeks)

Viruses

Arbovirus infections (e.g. Dengue) Marburg virus disease

Poliomyelitis Hepatitis A virus Arenavirus infection Ebola virus infection Lassa fever virus infection Haemorrhagic fever with renal syndrome

Hepatitis B Non-A, non-B hepalitis Rabies HIV

Q fever

Rickettsia Bacteria

Typhoid

Typhoid Brucellosis Leptospirosis Relapsing fever Lymphogranuloma venereum

Protozoa

Trypanosoma cruzi infection

Malaria Toxoplasmosis Trypanosoma rhodesiense infection

Helminths

with some assessment of the regularity of taking prophylaxis. The chronology of events should be recorded, particularly noting the interval between the end of the patient's travels and the onset of illness where this has occurred after leaving the tropics. This gives an idea of the incubation period of the disease (Table 9.12). Physical examination should pay attention to the presence of skin lesions, oral lesions, jaundice, adenopathy, hepatosplenomegaly, and involvement of the nervous system. After initial assessment it is vital to reassess the patient at frequent intervals, looking for the appearance of diagnostic symptoms and signs. Antimalarial chemoprophylaxis should be discontinued while investigation is proceeding as it may suppress parasitaemia and prevent diagnosis of malaria.

Visceral leishmaniasis Trypanosoma gambiense infection

Schistosomiasis (Katayama syndrome)

SUMMARY 3 Causes of imported fever Malaria Dengue Other arboviral infections Tuberculosis AIDS related infections Typhoid Rickettsial infections Amoebic liver abscess Visceral leishmaniasis Viral haemorrhagic fevers* * While numbers of cases of VHF are very low it is essential to consider this group of conditions in any patient who has visited an endemic area, is febrile and has a low platelet count. See risk assessment, p. 298.

Investigations The investigations performed will depend on the clinical diagnosis or the differential diagnosis. • Blood films should be examined daily for malaria parasites in any febrile patient who has been to a malarial area until either a diagnosis is made or the fever resolves. Blood films may also be examined for trypanosomes in a patient who has been in rural east or west Africa. A full blood count is useful for looking for leukocytosis (suggesting bacterial infections), leukopenia (typhoid, viral infections), pancytopenia (visceral leishmaniasis) or eosinophilia (helminthiasis: schistosomiasis, fascioliasis, visceral larva migrans, clonorchiasis, Bancroftian filariasis). • A Mantoux test should be done early on. • Blood cultures should be taken and, if brucellosis is a possibility, kept under observation for up to 6 weeks, as the organism grows slowly.

• Serological tests that may be useful include those for syphilis, leptospirosis, rickettsial diseases, brucellosis, amoebiasis, leishmaniasis, trypanosomiasis, toxoplasmosis, filariasis and schistosomiasis. They should be used selectively in relation to the clinical presentation and the exposure. A serum sample should be taken on admission and saved, and a second sample taken 10 days later. This pair of sera can then be used to look for a change (usually a rise) in titre. • Imaging procedures should include a chest X-ray and the scanning procedures mentioned under PUO are used as clinically indicated. • Liver and marrow biopsy (see PUO above) are also performed where there is a clinical indication, with samples cultured for routine organisms, brucella, fungi and parasites, in addition to histological examination of sections and microscopy of smears for tubercle bacilli and leish-

277

mania. Bone marrow culture may yield S. typhi in a patient who has previously taken antibiotics.

Management Ideally treatment will follow from the specific microbiological diagnosis of the cause of fever. Occasionally it is necessary to treat the patient for a disease suspected clinically but not proven by available laboratory tests. This may be treatment for malaria, or typhoid or rickettsial infection, or tuberculosis in fevers of longer duration. 1 12

GASTROENTERITIS Gastroenteritis is defined as an acute disease with diarrhoea and/or vomiting and a variable degree of systemic illness. 'Food poisoning' is a term used to describe a selflimiting form of gastroenteritis caused by eating contaminated food.

Aetiology A large variety of microorganisms can cause gastroenteritis (Table 9.13) but in over half of all cases the causative organism cannot be identified. Salmonellae and Campylobacter species are among the most frequent causes. In children the spectrum of organisms is slightly different and rotavirus is the most frequently recognized agent.

Epidemiology Gastroenteritis is a common infection in all parts of the world, occurring sporadically or in epidemics. Epidemics occur when a common source of food or drink is contaminated with micro-organisms and when infection spreads from person to person by the faecal-oral route. For example, Salmonella enteritidis frequently colonize the gut of chickens bred in battery conditions and may infect the oviduct and eggs of hens in batteries. During evisceration in the factory the meat is contaminated and the salmonellae survive freezing and thawing. If such chickens are not thoroughly cooked, salmonellae are ingested and colonize the intestine. During the last few years there have been outbreaks of diarrhoea due to E. coli O:157, which in patients of all ages can go on to cause a haemolytic uraemic syndrome with a significant mortality. This has occurred in retail food outlets where both cooked and uncooked meats are on sale, with the opportunity to contaminate cooked meats with E. coli O:157 from uncooked meat. Some forms of gastroenteritis show striking seasonal variations in frequency but these are usually significant only in temperate areas. Rotavirus is the most common

1

278

Case 9.1

2

MCQ 9.6

TABLE 9.13 Common causes of gastroenteritis Cause

Clinical presentation

Invasive bacteria Shigella species Salmonella species Campylobacter jejuni

Dysentery Diarrhoea, vomiting Diarrhoea predominantly

Toxin-producing bacteria Enteroxin E.coli(ETEC) Vibrio cholerae

Secretory diarrhoea Secretory diarrhoea

Neurotoxin Staphylococcus Bacillus cereus Clostridium botulinum

Vomiting (short incubation) Vomiting (short incubation) Vomiting, diarrhoea, neurological

Cytotoxin Clostridium difficile

Bloody diarrhoea, haemolytic uraemic syndrome Watery diarrhoea

Tissue-damaging viruses Rotavirus Norwalk virus Adenovirus Enterovirus

Infantile diarrhoea Winter vomiting Diarrhoea, vomiting Diarrhoea, vomiting

Protozoa Cryptosporidium Entamoeba histolytica Giardia lamblia Cyclospora cayetanensis

Diarrhoea Dysentery Chronic diarrhoea Diarrhoea

E. coli 0:157

cause of infantile diarrhoea worldwide. In Britain and the USA its peak incidence is in 2 or 3 months during the winter, but in tropical climates there is little seasonal variation. Outbreaks of food poisoning caused by Salmonella, Campylobacter or toxin-producing Staph. aureus are more common in summer.

Pathogenesis Organisms must evade the host defences and colonize the intestine before they can cause disease. Gastric acid is an important defence, and patients with a partial gastrectomy are susceptible to intestinal infection. Some bacteria have on their surface determinants that bind to epithelial receptors. Examples are Vibrio cholerae and some strains of E. coli. The known mechanisms by which microorganisms cause diarrhoea can be divided into two groups: toxin secretion and mucosal damage. These are not mutually exclusive, and for many organisms the precise mechanism is unknown. Toxin secretion Three different mechanisms are recognized by which toxins can cause intestinal pathology. Cholera is a good example of small intestinal toxin-mediated secretory diarrhoea (see

9

CASE STUDY 9.1 ACUTE INFECTIVE DIARRHOEA A 21-year-old female student was travelling in rural Nepal staying in guest houses. Overnight she developed fever, chills and generalized body aches and pains. The next day she was anorexic and developed acute watery diarrhoea, but after 12 hours started to get severe griping abdominal pains which were particularly severe in the left lower quadrant. The character of the diarrhoea changed to the passage of frequent small volume stools, with exacerbation of pain. After 12 hours the stools became uniformly bloodstained. She also described the sensation of wanting to empty the rectum with extreme pain in the anus and relatively little blood-stained material coming out (tenesmus). She was wakened from sleep numerous times in the first 24 hours of the illness and sought medical attention at a local hospital. On examination she looked unwell. Pulse 110 regular, BP 90/60, temperature 38.4°C. There

Questions 1. What is the differential diagnosis? 2. What investigations should be performed? 3. What are the main points of management?

was tenderness over the left lower quadrant of the abdomen. She passed a small amount of blood stained watery stool shortly after admission. Comment This history is suggestive of acute invasive bacterial infection of the gut. It is typical of shigellosis, with the initial fever, followed by acute watery diarrhoea and then dysenteric diarrhoea, passing small volumes of stool with pain and tenesmus. However, the distinction between shigellosis, Campylobacter infection and salmonellosis is clinically unreliable. The marked systemic upset initially would be against giardiasis, cyclospora (first described as a cause of diarrhoea in Asia among western travellers in Nepal), cryptosporidium, toxigenic E. coli and cholera. Amoebic dysentery begins more insidiously and is gradually progressive. Investigations must include examination of the stool for amoebic trophozoites, and culture for Shigella and Salmonella. The finding of amoebic cysts in her stool is irrelevant, as motile trophozoites with ingested red cells would be evident in faecal smears in invasive amoebiasis. Microscopy of the stool would have shown large number of pus cells, which do support a diagnosis of invasive bacterial infection.

p. 322), and enterotoxigenic strains of E. coli secrete a heatlabile enterotoxin which acts in a similar way. In staphylococcal food poisoning neurotoxin-secreting staphylococci multiply in food, and the toxin causes vomiting and diarrhoea within hours of ingestion. Staphylococci are not detected in the stools of patients with this type of food poisoning. Bacillus cereus, which contaminates reheated rice, causes vomiting by a similar mechanism. Cytotoxins, such as verotoxin produced by E. coli O:157 and toxin B of Clostridium difficile, cause tissue damage. Mucosal damage Shigellae invade the colonic mucosa and cause local inflammation, and so the faeces contain blood and inflammatory cells. Impairment of water absorption from the colon and exudation of protein-rich fluid through the

Since amoebic cysts were seen in a faecal smear, she was treated with metronidazole as well as norfloxacin for shigellosis. She was also given intravenous fluids as she could not initially tolerate ORS, but within 36 hours of starting treatment the fever regressed and the diarrhoea started to improve, though after finishing 7 days' treatment with both drugs bowel actions remained abnormal both in frequency and the consistency of the stools. No more blood was seen for the remainder of her stay in Nepal or after her return to the UK. The management was appropriate, though one could argue that there were sufficient clues to suggest that norfloxacin alone would have been enough. With marked systemic upset and dysenteric stools there are clear indications to give antibiotics. Rehydration by the oral route was first attempted, with recourse to intravenous fluids when that failed. The persisting abnormalities of bowel habit after effective treatment are an indication of postinfective intestinal irritability, but it is essential to ensure that there is no persisting infection, such as giardiasis, and that there is no possibility of non-specific inflammatory bowel disease which, some reports indicate, may be precipitated by acute gastrointestinal infection.

inflamed and ulcerated mucosa are the cause of fluid stools. Salmonellae, Campylobacter and rotavirus are also invasive and cause tissue damage, but the precise mechanisms are not understood. Damage to the brush border of the enterocytes causes temporary disaccharidase deficiency, which may prolong the diarrhoea even after the pathogen has been eliminated.

Clinical features Most episodes of gastroenteritis are mild and self-limiting. The incubation period is usually short, ranging from a few hours with staphylococcal food poisoning up to 5 days with Campylobacter infection. Vomiting, diarrhoea and cramping abdominal pains are typical features. Large volume stools suggest a small bowel diarrhoea. The stools are fluid

279

without blood or pus in toxin-mediated disease (e.g. cholera), but when invasion and mucosal damage occur in Salmonella, Shigella and Campylobacter infections blood and mucus may be seen in the stool. The appearance of the stools is an unreliable guide to the type of organism. The major consequence of gastroenteritis is dehydration with electrolyte imbalance. Babies are more vulnerable and can rapidly become severely dehydrated. It is essential that the clinician can make a reasonably accurate assessment of the degree of dehydration and Table 9.14 gives some guidelines. Fever and signs of systemic illness suggest invasive pathogens. Some invasive organisms can cause septicaemia (e.g. Salmonella typhimurium), and some shigellae produce neurotoxins which enter the circulation in the absence of septicaemia and, on reaching the central nervous system (CNS), cause headache and meningism. The illness is usually over within 7 days, but symptoms are liable to persist for longer with shigellae and Campylobacter. Persistent diarrhoea suggests Giardia infection, especially if steatorrhoea is present. Although amoebiasis is commonly thought of as a cause of acute watery diarrhoea in the tropics it is in fact relatively uncommon, causing only 6% of cases in a large study from Bangladesh. Usually the diarrhoea has an insidious onset and gets worse gradually, though all ranges of severity are possible, from the passage of a little bloodstained mucus with stools to a severe illness every bit as bad as ulcerative colitis.

Diagnosis The diagnosis is usually clear, but problems arise when there is vomiting in the absence of diarrhoea; when abdominal pains are accompanied by localized tenderness, suggesting pelvic inflammatory disease or appendicitis; and when blood in the stools is taken to indicate inflammatory bowel disease. A positive diagnosis is established by identification of the organism in stools by microscopy, antigen assay (for rotavirus) or culture.

Management The management of diarrhoea is concerned mainly with the management of dehydration. After an assessment of the degree of dehydration using the criteria set out in Table 9.14, a plan of management is determined. Where there is no dehydration (<5% body weight lost) oral fluids are given freely and normal feeding is encouraged. When there is some dehydration (5-10% body weight lost) the deficit should be replaced using oral dehydration solution (ORS; Table 9.15) 80ml/kg over 4 hours. In addition to ORS, water and other non-electrolyte fluids should be 1

280

MCQ 9.7

TABLE 9.14 Assessment of dehydration A No dehydration

B Some dehydration

C Severe dehydration

Condition

Alert

Eyes

Normal

Restless*/ irritable* Sunken

Tears Mouth and tongue Thirst Drinks Skin turgor

Present Moist

Absent Dry

Lethargic*, unconscious* Very sunken and dry Absent Very dry

None Normally Normal

Thirsty* Eagerly* Reduced*

Drinks poorly* Not able to drink* Markedly reduced*

Features in bold with * are key signs. If 2 signs from column C including 1 key sign are present then severe dehydration is present; if 2 signs from column B including 1 key sign are present, there is some dehydration. (After A Manual for the Treatment of Diarrhoea, World Health Organisation 1990.)

TABLE 9.15 Composition of oral rehydration solution (WHO formulation) Sodium chloride Trisodium citrate, dihydrate Potassium chloride Sucrose Made up to 1 litre with clean water

3.5 g/L 2.9 g/L 1.5 g/L 40 g/L

given to provide water for insensible losses. After each bowel action an additional 400 mL of ORS should be given to replace continuing losses. Oral rehydration solution very effectively replaces lost fluid despite the presence of an enterotoxin-mediated secretory process in the gut, but it does nothing to reduce the volume of diarrhoeal stools produced. The use of rice flour as the source of glucose in cereal-based ORS solutions not only replaces lost fluid but also reduces the volume of faecal effluent. It has the further advantage of being widely available in areas of the world where infectious diarrhoeal diseases are common, but preparing cereal-based ORS requires heating the solution to make a palatable gruel. When there is severe dehydration (>10% body weight lost) intravenous access should be established immediately and Ringer's lactate (or normal saline if that is not available) infused at 30mL/kg over 30 minutes, and then 70mL/kg over 2 hours. Thereafter, continuing losses should be replaced orally if the patient is able to drink, or intravenously if not. In all patients with diarrhoea it is essential that their progress is assessed every hour, so that if symptoms and signs are deteriorating on an ORS regimen they can be

switched to intravenous fluids. Vomiting may prevent successful ORS therapy, but first the administration of frequent small volumes of fluid by mouth should be tried. When initial rehydration has been achieved, continuing faecal losses should be replaced by ORS, 400 mL after each bowel action, and feeding begun as soon as the patient wants to eat. Bowel sedatives have little place in the management of intestinal infections and they may be harmful, precipitating toxic dilatation in elderly and young patients in infections with the invasive bacterial pathogens. Antibiotics are normally not required, but they are needed when there is profuse watery diarrhoea, marked systemic upset and obviously dysenteric (bloodstained) diarrhoea. Empirical treatment with ciprofloxacin 500 mg twice daily for 5 days is usually given. In addition, the following situations merit antibiotic treatment: • • • • • • • •

Septicaemic Salmonella infection (ciprofloxacin) Severe shigellosis (ciprofloxacin) Severe Campylobacter infection (erythromycin) Yersinia (doxycycline) V. cholerae (doxycycline) Entamoeba histolytica (tinidazole, metronidazole) Giardia lamblia (tinidazole, metronidazole) Cyclospora cayetanensis (co-trimoxazole).

Problems associated with use of antibiotics are exacerbation of diarrhoea, prolonged carriage of some organisms (e.g. salmonellae) and development of resistance to multiple antibiotics (e.g. shigellae). There is evidence to suggest that antibiotic treatment in E.coli O:157 infection may increase the risk of renal complications by promoting toxin release. 1

FURTHER READING ON GASTROENTERITIS A manual for the treatment of diarrhoea 1990. Geneva: World Health Organization Programme for the Control of Diarrhoeal Diseases.

ACUTE UPPER RESPIRATORY INFECTION

TABLE 9.16 Causes of upper respiratory infection Disease

Causative organism

Common cold and pharyngitis

Rhinovirus Coronavirus Influenza virus Parainfluenza virus Respiratory syncytial virus Adenovirus Unknown

Pharyngitis/tonsillitis

As above, plus: Epstein-Barr virus B-haemolytic streptococcus group A Mycoplasma pneumoniae Mixed anaerobes and spirochaetes* Corynebacterium diphtheriae* Neisseria gonorrhoeae*

Stomatitis

Candida albicans Herpes simplex virus Coxsackie A and other enteroviruses Mixed anaerobic bacteria

9

*Less common

treatment of choice for streptococcal pharyngitis, with erythromycin for penicillin-allergic subjects. Ampicillin and related drugs should be avoided, as they often cause a rash in patients with infectious mononucleosis. Upper respiratory infections are frequently complicated by otitis media in children and, less often, by sinusitis in adults or children. These are usually caused by secondary bacterial infection, and it is therefore important to recognize them and treat with antibiotics. Common bacterial causes of otitis media include Strep. pneumoniae, H. influenzae (mainly untypable), group A B-haemolytic streptococci and Branhamella catarrhalis. Many H. influenzae are now resistant to amoxicillin, and co-amoxyclav can be used instead. The bacteria that cause acute sinusitis are similar to those incriminated in otitis media, but anaerobes, Gramnegatives and Staph. aureus should also be considered. Upper respiratory anaerobic bacteria are usually penicillin sensitive, in contrast to those found in the large bowel.

Common cold and pharyngitis The common cold has an incidence of 1-4 episodes per person per year. Large gaps remain in our knowledge of pathogenesis, although it is clear that many viruses can cause the same clinical entity. There is considerable overlap between organisms causing colds and those causing pharyngitis (Table 9.16). Pharyngitis is more important, as it is often part of a more systemic illness. The majority of episodes of pharyngitis are caused by viruses. Antibiotic treatment should be reserved for culture-proven cases of streptococcal pharyngitis. In the developed world, group A streptococcal pharyngitis is now rarely complicated by rheumatic fever or glomerulonephritis. Penicillin, as a single i.m. dose or orally for a minimum of 10 days, is the

Acute stomatitis In contrast to pharyngitis, the microbial cause of acute stomatitis can often be diagnosed from the appearance of lesions in and around the mouth. Candida is a common pathogen, particularly in debilitated patients, patients receiving broad-spectrum antibiotics and patients with HIV infection. White plaques occur on an erythematous background. The distribution is wide, including the buccal mucosa. Ulcers are usually superficial and not easily confused with primary herpes simplex infection. The latter is often associated with perioral clusters of vesicles. Severely painful ulceration of the mouth and lips in a young immunocompetent patient is usually due either to

281

Stevens-Johnson syndrome or to primary herpes simplex infection. Coxsackie virus infection may cause small, scattered vesicles and ulcers, but they are more sparse than those of herpes simplex and less painful. Noninfective causes of ulcers must also be considered, such as aphthous ulcers and, rarely, Behget's disease (p. 1153). Infections in the subcutaneous tissue spaces of the

CASE STUDY 9.2

CHRONIC FATIGUE

A 34-year-old female nursery school teacher in a stable relationship complained of recurrent episodes of sore throat, the most recent of which was 2 months ago. She had had difficulty throwing off 'flu and other infections for 3 years and felt tired all the time. She had pains in the neck and the 'glands' in her neck were going up and down. Her muscles ached during and after exertion and she became fatigued sooner than she would have expected. She denied depression and had no problems at work. There were no abnormalities on examination and, in particular, no lymphadenopathy but minimal tenderness in the anterior cervical region. Muscle power was normal and there was no convincing muscle tenderness. Her affect was normal but she was clearly tired of her symptoms, sometimes sighing while describing them, and she was anxious about an underlying persistent disease of a 'weak immune system'. Questions

1. What is the differential diagnosis? 2. What further investigations would you perform? 3. What would be your approach to management? Discussion

The GP had checked her full blood count repeatedly so we know that she did not have neutropenia or lymphopenia. Examination of the pharynx showed normal tonsils but it would be helpful to see her during an episode of sore throat to exclude 282

face and neck often originate from periodontal infection. Ludwig's angina is an anaerobic bacterial infection of the submandibular and sublingual spaces which causes firm swelling of the tissues below the chin. These may press inwards and compromise the airway. There is accompanying fever and tachycardia and the patient requires urgent antibiotic treatment, usually with penicillin and metronidazole.

recurrent tonsillitis. Persistent pharyngeal infection with group A Phaemolytic streptococcus may cause recurrent sore throat but it does not cause tiredness between episodes. One of her episodes of sore throat could have been infectious mononucleosis caused by EBV infection (or, less commonly, CMV or toxoplasma infection); this usually causes one continuous episode of illness and recurrence is rare beyond 12 months after the initial episode but it can cause tiredness persisting for up to 6 months. Many patients are anxious that recurrent upper respiratory infections indicate a disorder of the immune system. This is not so in the absence of recurrent candidiasis and it is not necessary to undertake immunological investigations to prove it but reassurance is often helpful. Simple causes of tiredness such as anaemia and hyper- or hypothyroidism can be excluded by examination and confirmed by appropriate tests. Rather than requesting a wide range of investigations, it is advisable to target them at excluding diseases which have been a particular cause of anxiety to the patient if the anxiety can not be allayed by reasonable discussion. Assessment by a sympathetic psychiatrist can be helpful and should be regarded as a part of the overall investigation of patients with very prolonged symptoms. Chronic fatigue syndrome is diagnosed when the combination of symptoms illustrated by the above patient is present with negative investigations for organic disease. Many patients have the

somatic symptoms of depression but deny the state of depression or attribute it to the presence of physical symptoms. Sometimes anxiety seems to be a more important factor. The term myalgic encephalomyelitis (ME) should be avoided since none of these patients has encephalomyelitis. The patient was reassured by a wide range of normal investigations (including FBC, ESR, CRP, U&E, LFTs, thyroid function, auto-antibody screen, CXR, negative IgM antibodies to EBV, CMV and toxoplasma and negative throat swab). She was seen by a psychiatrist who found no evidence of depression but she discussed some family stress at Christmas and uncertainty about her career; when she was on holiday at Christmas, there was a family row in which she acted as mediator between her brothers and her mother which she found very stressful. Later she was able to resolve this but she missed her holiday because of the stress and her illness (flu). At work she had been considering whether she should go part-time because of her symptoms or aim for promotion to a deputy school headship. Her symptoms had been improving slowly and she now felt optimistic with the summer holidays looming. She was offered counselling which she attended and found useful. She also undertook a graded exercise programme and at review sessions she was encouraged to work through her fatigue. By end of the year, she was working full-time in the same job and coping despite some persistent tiredness.

SEXUALLY TRANSMITTED DISEASES Sexually transmitted diseases (STDs) are infections commonly transferred through sexual contact. They are discussed in detail in Chapter 11.

CHRONIC FATIGUE SYNDROME (POSTVIRAL SYNDROME, MYALGIC ENCEPHALOMYELITIS) This syndrome is seen most often in young adults. There is a prolonged period of lethargy, often with muscle pains and fatiguability, usually after an episode reminiscent of a viral illness but sometimes in the absence of an evident acute illness. The problem causes great distress to patient and relatives because an explanation is lacking and there is prolonged inability to work or play at expected levels, often coupled with an unreasonable need for sleep. Doctors are frustrated by their inability to establish a positive diagnosis or to give any specific treatment, and the patient often turns to alternative medicine.

Aetiology The syndrome is heterogeneous and no single explanation will fit all patients. The term 'myalgic encephalomyelitis' is misleading as there is no evidence of encephalomyelitis and myalgia may not be present. A number of infections have been associated with this syndrome and they are not all viral. Infectious mononucleosis is the best known and studied, but CMV infection and toxoplasmosis have been suspected, and it is likely that many other organisms can act as triggers in susceptible individuals. Enteroviruses, particularly of the Coxsackie group, have also been suspected but

evidence is inconclusive. Often the only evidence for infection at the onset is a history of febrile illness, perhaps with pharyngitis or lymphadenopathy, and a full blood count may have revealed some atypical mononuclear cells. The relationship of postviral syndrome to depression is complex. There is no doubt that some patients develop the symptoms and signs of depression and benefit from antidepressant treatment. More commonly the patient becomes depressed as a result of the illness but lacks many of the features of primary depression. In this situation antidepressant therapy is unlikely to be beneficial.

9

Clinical features The central symptoms are profound lethargy and easy fatiguability. Often these are accompanied by muscular pains, headaches and symptoms suggestive of fever, such as sweating. This is usually mild, and when the temperature is recorded it is normal or only minimally elevated in the evenings. Exertion exacerbates tiredness, and the patient often goes to bed early and wakes late still feeling tired.

Management Patients benefit greatly from having their illness taken seriously, and if a precipitating infection can be positively identified this also helps. Sometimes the greatest problem is anxiety about the outcome or about whether there is an underlying disease that has not been diagnosed, so reassurance is important. When patients have restricted their activity it is worth trying to rehabilitate them by a programme of daily exercise, starting with something light such as walking or swimming, and gradually increasing it in a documented way. The patient is assured that exercise is not harmful and advised to sleep as required. Most patients recover spontaneously within weeks rather than months, but a few continue to suffer for over a year.

VIRUS INFECTIONS Viruses are usually classified according to their nucleic acid content and morphology. The classification of viruses used in this section is shown in Tables 9.17A and B. Arboviruses (arthropod-borne) are viruses transmitted by the bite of an insect vector. Most arboviruses are togaviruses, but bunyaviruses and orbiviruses (RNA viruses in the reovirus family) are also included.

ENVELOPED DOUBLE-STRANDED DNA VIRUSES HERPES VIRUSES Six herpes viruses cause disease in humans: • Varicella zoster virus (VZV)

• • • • •

Two herpes simplex viruses (HSV 1 and 2) Epstein-Barr virus (EBV) Cytomegalovirus (CMV) Human herpes virus 6 Herpes virus associated with Kaposi's sarcoma (HHV8).

Morphologically, herpes viruses are very similar and indistinguishable by electron microscopy. They are similar in some aspects of their behaviour, notably their ability to establish latent infection with the potential for reactivation. During latency in dorsal root ganglia VZV shows no demonstrable evidence of viral replication for many years. With HSV there is evidence of a low level of replication, and reactivation is more frequent. EBV infects B lymphocytes, and EBV and CMV replicate at a low level continuously. CMV can often be detected in the urine of asymptomatic carriers, and herpes simplex in saliva. The mechanisms by which virus replication is suppressed are

283

TABLE 9.17A Classification of viruses:DNA

TABLE S.176 Classification of viruses: RNA

Morphology

Virus

Morphology

Virus and disease

Enveloped double-stranded

Herpesviridae Varicella zoster (VZV) Herpes simplex (HSV I, II) Epstein-Barr (EBV) Cytomegalovirus (CMV) Human herpes virus 6 (HHV6) Human herpes virus 8 (HHV8) Poxviridae Vaccinia Variola (smallpox) Orf Parvoviridae Hepadnaviridae

Enveloped single-stranded

Orthomyxoviridae Influenza Paramyxoviridae Parainfluenza Respiratory syncytial disease Mumps Measles Togaviridae Rubella Alpha (equine encephalitis) Flaviviridae Dengue Yellow fever Encephalitis Hepatitis C Bunyaviridae Arenaviridae Lymphocytic choriomeningitis Lassa Filoviridae Marburg Ebola Retmviridae Lentiviruses (HIV 1 and 2) Oncoviruses - type C (HTLV1 and 2) Rhabdoviridae Rabies Vesiculostomatitis Coronaviridae Coronavirus Reoviridae Rotavirus Reovirus Picornaviridae Rhinovirus Enteroviruses Coxsackie A, B Echo Polio Enterovirus 70, 71 Hepatitis A Caliciviridae Norwalk Hepatitis E virus Astroviridae Astroviruses

Enveloped single-stranded

Hepatitis B Hepatitis D

Non-enveloped double-stranded

Adenoviridae Papovaviridae Papilloma Polyoma

incompletely understood, but T cell-mediated immune mechanisms probably have an important role, as all the herpes viruses are liable to reactivate in patients with depression of cell-mediated immunity, such as those with lymphoma or leukaemia.

Varicella zoster virus infection Chicken pox (varicella) This illness is easy to recognize, common and highly infectious. It is a febrile illness with a vesicular rash (Fig. 9.6). During the primary infection there is a viraemic phase when VZV becomes widely disseminated in white cells. In normal subjects it only produces significant lesions in the skin and mucous membranes, but infection of cutaneous sensory nerves leads to infection of dorsal root ganglia, and here the virus enters a latent state. Reactivation later causes herpes zoster (shingles). During primary infection VZV is present in nasopharyngeal secretions and is transmitted readily by coughing and sneezing. It is also in vesicle fluid in chickenpox and shingles, and this can be a source of transmission. The incubation period of chickenpox is 15-18 days and the patient is infectious from a few days before the rash erupts until new lesions cease to appear. The severity of chickenpox is determined by the age and immunological state of the patient. It is most severe in neonates if they have no maternal antibody to VZV, and in patients immunocompromised by disease such as leukaemia or by immunosuppressive therapy. It is usually a mild disease in children, but it becomes increasingly severe after puberty. 284

Clinical features In children chickenpox starts with the appearance of the

Non-enveloped double-stranded Non-enveloped single-stranded

rash, which is often sparse and not accompanied by constitutional symptoms. The most troublesome feature of the illness is itching. In adults there is a notable prodromal illness lasting 2-3 days, during which headache, backache, shivering and fever occur, sometimes with a sore throat and cough. The more severe the prodromal illness, the more profuse is the rash that follows. The rash starts on the trunk and spreads to the face and limbs (Fig. 9.6). Lesions appear in crops which progress rapidly through the sequence macule, papule, vesicle, pustule, scab. Often the first visible lesions are vesicles containing clear fluid with little surrounding erythema, but the pustular stage is reached within 24 hours. Vesicles on the

9

FIG. 9.6 Vesicles and pustules in a patient with chickenpox

palate, tonsils and pharynx rapidly burst, leaving superficial ulcers. Crops of skin lesions erupt over 2 or 3 days, but it takes several more days for them all to heal. Constitutional symptoms, if present, lessen with the onset of the rash and disappear within 2 days. Complications Secondary infection of the skin lesions, including the scalp, occurs because of scratching, but it usually responds to cleaning and bathing and rarely requires antibiotics. Chickenpox pneumonia occurs in one-fifth of adults, and more frequently in immunocompromised patients. It varies in severity from transient dyspnoea to a fulminant pneumonia. Postinfectious encephalitis is extremely rare and has a variable course, but it is often severe or fatal. Its onset is within a week of the rash and cerebellar involvement is characteristic. Bleeding into the skin, nose or intestine, also rare, is due either to steroid-responsive thrombocytopenia or to consumptive coagulopathy, when the prognosis is much worse. Neonatal chickenpox Most women are immune by the time they reach childbearing age, but if a mother develops chickenpox within a few days of having a baby the child is born without the protection of maternal IgG antibodies transferred across the placenta, and it is immediately exposed to the virus. If the child becomes infected during the first 5 days of life disease is severe, with multi-organ involvement. Disease can be prevented or made milder by the administration of varicella immune globulin 250 mgi.m. at birth. Aciclovir l0mg/kg is given intravenously 8-hourly for 7 days if any signs of infection (e.g. fever) appear. This does not seem to prevent the virus becoming latent. Congenital chickenpox Rarely, VZV can cross the placenta during the first 20 weeks of pregnancy, causing fetal damage or death. Diagnosis and management A clinical diagnosis of chickenpox suffices in most cases.

FIG. 9.7 Dermatomal distribution of varicella zoster lesions

Electron microscopy is the quickest way to establish that a herpes virus is present in vesicle fluid, but does not establish its type. The virus can be identified by tissue culture inoculation, and antibodies are detected by a complement fixation test. Treatment with aciclovir (p. 268) is required only in severe cases susceptible to complications such as chickenpox pneumonia, in immunocompromised patients (such as children with malignant disease), and in neonates (Table 9.5, p. 268). Intravenous treatment (l0mg/kgt.d.s in adults; 500mg/m2 for children aged 3 months to 12 years; l0mg/kg for children less than 3 months) is needed in order to achieve adequate serum concentrations. Many patients with chickenpox pneumonia need ventilation (see Ch. 14) and antibiotic treatment for secondary bacterial infection, the most common organism being Staphylococcus aureus. Shingles (zoster) When VZV reactivates within a dorsal root ganglion it migrates down the nerve into the skin and causes a crop of lesions identical to those of chickenpox but confined to the dermatome (Fig. 9.7). Pain within the dermatome may precede the eruption by a few days and is often severe. One or two adjacent dermatomes on the trunk are most commonly affected, but involvement of the ophthalmic division of the fifth cranial nerve is important because the eye can be damaged by the virus or by secondary bacterial infection.

285

In the elderly it is common for a mild viraemia to occur, causing a few scattered lesions to appear outside the involved dermatome. More serious dissemination with visceral involvement is seen in immunocompromised patients. Patients with Hodgkin's disease are particularly susceptible to shingles, especially during chemotherapy. The skin heals over 2-3 weeks and there is often some residual pain for a few weeks. Rarely, intractable pain continues for years (post-herpetic neuralgia). This is more common following ophthalmic zoster, and after severe infections in immunocompromised patients. 1 Management Most patients do not require admission to hospital and treatment of pain is the main concern. Continuous treatment with simple analgesics such as paracetamol is often effective and should be tried before drugs that have more side-effects. Oral treatment. A small increase in the rate of healing has been demonstrated when oral aciclovir is used in high dose, 800 mg five times daily for 7 days. This benefit may be enhanced with the use of newer alternatives such as valaciclovir, as higher serum levels of aciclovir are achieved. Less frequent administration is needed with famciclovir 250 mg thrice daily. Intravenous treatment. Intravenous aciclovir (l0mg/kg t.d.s for adults) should be given to immunocompromised patients for 1 week. It reduces pain, shortens the healing time and limits dissemination of the virus. Evidence for the benefit of prednisolone in reducing post-herpetic neuralgia is controversial and it is not recommended. Topical treatment • In ophthalmic zoster it is important to combine an antibiotic (e.g. chloramphenicol drops) with aciclovir ointment. • Pruritus treatment, e.g. calamine, is soothing for a few minutes. • Antivirals, e.g. idoxuridine and aciclovir, are impractical and of limited value. 0

Herpes simplex virus infection HSV 1 and 2 cause patterns of disease that differ clinically and epidemiologically but share some characteristics. Both establish permanent, latent infection within sensory nerve ganglia following primary infection, with reactivation causing episodes of disease in the area supplied by the nerve. These viruses are indistinguishable morphologically but can be separated by differences in antigen expression. Infection with HSV 1 is more common. Antibodies are detectable in 80-100% of adults in lower socioeconomic groups, and in 30-50% in higher socioeconomic groups. Both viruses are readily transmitted, mainly in vesicle fluid. HSV 1 is usually acquired during childhood, but HSV 2 infection is mainly transmitted sexually and is most common in sexually promiscuous adults. Clinical expression of infection with these viruses is strongly influenced by age, nutrition and the immunological state of the patient. An immunologically immature neonate is overwhelmed by disseminated HSV infection acquired during parturition. Infection in older children is often asymptomatic or causes localized disease which is most severe in the presence of malnutrition. 0 Primary infection with HSV 1 Acute herpetic gingivostomatitis in children 2-4 years old presents with fever, sore throat, and submental and anterior cervical lymphadenopathy. Clusters of vesicles appear around the lips and in the mouth from gums to pharynx (Fig. 9.8). Inside the mouth lesions rapidly ulcerate, causing pain and dysphagia with consequent drooling of saliva. The child is miserable and may become dehydrated if not encouraged to drink through a straw. The lesions heal over 1-2 weeks. Primary infection can occur anywhere on the skin and, on the finger, a herpetic whitlow can closely mimic a bacterial paronychia. Widespread involvement of the skin can occur in individuals with atopic dermatitis (eczema herpeticum, p. 399) and, particularly in infants, this can be associated

SUMMARY 4 Varicella zoster and herpes simplex infection

286

Primary

Reactivation

VZV HSV 1

Chickenpox Acute or asymptomatic gingivostomatitis Keratoconjunctivitis Encephalitis

Shingles Cold sores Keratoconjunctivitis

HSV 2

Genital

Genital

1

Fig. 9.1

2

MCQ 9.8

4

Fig. 9.3

5

Fig. 9.4

3

Fig. 9.2

FIG. 9.8 Gingivostomatitis caused by primary herpes simplex 1

with systemic symptoms. In the eye keratoconjunctivitis can cause scarring of the cornea and loss of vision. Herpetic encephalitis in adults presents a characteristic clinical picture owing to necrotic lesions in the temporal lobe. This rare manifestation of primary HSV infection is described on page 1422. Primary infection with HSV 2 Although HSV 1 and 2 can cause identical disease, HSV 2 is largely associated with genital infection (see p. 456). 4 Infection of neonates can occur during parturition when the mother has genital herpes, particularly if her infection is primary rather than recurrent. The infant develops scattered clusters of vesicles which recur in the same sites subsequently. Alternatively there may be rapid deterioration as the virus disseminates to all the viscera, causing encephalitis, pneumonia and hepatitis, with a high mortality. Usually this severe illness starts in the first few days after birth, but occasionally onset is delayed several weeks. Reactivation of HSV The mechanisms involved in maintaining latency are not fully understood. Reactivation may occur after exposure to ultraviolet light, an episode of infection or trauma to the sensory nerve. Many episodes have no identifiable precipitant. Reactivation is more common in irnmunocompromised patients, particularly those with leukaemia undergoing treatment. The characteristic lesion of reactivated HSV 1 is a cold sore. Patients often experience minor discomfort in the form of tingling or itching up to 24 hours before a lesion appears. Lesions can occur anywhere on the skin or mucous membranes, but are commonly adjacent to the lips. 5 A macule rapidly progresses to a papule and then to a cluster of vesicles which involute, ulcerate and scab. The whole process takes 1-2 weeks. See Chapter 11 for symptoms of reactivation of genital herpes (p. 456).

underdeveloped countries and in crowded living conditions, but infection is delayed in more prosperous environments. Infection in adolescents and adults usually causes infectious mononucleosis (glandular fever). Other diseases related to EBV infection are nasopharyngeal carcinoma, Burkitt's lymphoma, the very rare syndrome of X-linked recessive progressive combined variable immunodeficiency, and lymphoid interstitial pneumonitis in children with AIDS.

9

Infectious mononucleosis Infectious mononucleosis is an acute febrile illness caused by EBV and characterized by a severe sore throat and generalized lymphadenopathy. Pathology and pathogenesis EBV enters through the pharyngeal epithelium and replicates in B lymphocytes, causing them to proliferate. Fifteen per cent or more of blood mononuclear cells appear 'atypical' during acute infection: these are T lymphocytes reacting to the presence of infected B lymphocytes. EBV does not establish a true latent state but maintains a low level of replication throughout life. During infectious mononucleosis, the spleen and lymph nodes are intensely infiltrated with mononuclear cells and the spleen is unusually friable. Haemorrhage may occur into the subcapsular area, and rarely this leads to intraperitoneal haemorrhage. Clinical features Infectious mononucleosis is most common in adolescents and young adults, but children as young as 5 years old may be affected. The illness is more severe in older patients. The incubation period is uncertain but is probably 1-2 weeks. There is a gradual onset of malaise, general aching and fever. The throat becomes painful and engorged with enlarging lymphoid tissue, so that speech is nasal and swallowing is inhibited by pain. The tonsils are large and coated with thick, white exudate (Fig. 9.9). Petechiae may be seen on the soft palate. There is generalized lymphadenopathy

Management of herpes simplex HSVs are highly sensitive to aciclovir and other DNA polymerase-inhibiting drugs. Only serious primary infections with HSV 1, such as encephalitis in adults, severe eczema herpeticum or disseminated neonatal infection, warrant treatment with systemic antiviral drugs. The usual dose of aciclovir for intravenous treatment of herpes simplex is 5mg/kgt.d.s., but for encephalitis l0mg/kg is needed. Initial attacks of genital herpes are treated with 5 days of aciclovir, with an antibiotic if secondary infection has occurred. For recurrences attention to local hygiene is all that is required and 5% aciclovir ointment may be helpful for local lesions.

Epstein-Barr virus infection Epstein-Barr virus (EBV) is a herpes virus which is transmitted by close oral contact. Infection is frequently acquired asymptomatically during early childhood in

FIG. 9.9 Acute tonsillopharyngitis caused by primary Epstein-Barr virus infection

287

in most patients and splenomegaly is common. Uncommon features (occurring in less than 15% of patients) are rash, jaundice and tender hepatomegaly. Thrombocytopenia or autoimmune haemolytic anaemia occur rarely. The acute stage of the illness lasts 2-3 weeks and most patients are entirely well within a month. A small proportion of patients, particularly when older, make a slower recovery, continuing to feel lethargic for a variable period (see Chronic fatigue syndrome, p. 283). Differential diagnosis Other viral causes of sore throat and lymphadenopathy, such as adenovirus, influenza virus and rhinovirus, must be considered (see Table 9.16). Streptococcal infections can cause an identical appearance in the throat, but the lymphadenopathy is mainly in the anterior cervical region. Streptococci are found on throat swabs in 10% of patients with infectious mononucleosis and their role is unclear. Many virus infections are associated with atypical lymphocytes on the blood film, but these can be distinguished from glandular fever cells by an experienced observer. Atypical lymphocytes are also seen in acute CMV infection and occasionally in toxoplasmosis. Both of these conditions can cause lymphadenopathy but they rarely produce sore throat. Laboratory features Atypical mononuclear cells are usually present in the blood film at some stage of the disease. Tests for heterophil antibodies, i.e. antibodies to heterologous erythrocytes, are the Paul-Bunnell and Monospot tests. Heterophil antibodies are present in 90% of patients by the third week of illness, but in only 30% during the first week. Specific IgM antibody measurements are not reliable. Liver function tests are often abnormal. Management No specific treatment is available. Simple analgesic therapy is used to relieve discomfort in the throat, and frequently the patient can only take fluids in the early stages. Occasionally the degree of congestion is such as to threaten the airways; a short course of prednisolone, starting at 30 mg daily and tailing off over 2 weeks, can produce a rapid reduction in the size of the tonsils. If penicillin therapy is indicated for treatment of concurrent Streptococcal infection, ampicillin and amoxicillin should be avoided as they cause a severe maculopapular rash. X-linked recessive progressive combined variable immunodeficiency (Duncan's disease) This rare inherited disease is now recognized as an abnormal response to EBV infection. Patients either die of overwhelming EBV infection in the acute stage or develop

288

1

MCQ 9.9

4

MCQ 9.10

2

Fig. 9.5

3

Fig. 9.6

a progressive immunodeficiency following infection. Lymphoma may complicate the late stages. 1

Cytomegalovirus (CMV) infection CMV is a herpes virus which rarely causes significant disease in immunologically competent subjects. Infection is usually asymptomatic, but it may be associated with a mild febrile illness during which atypical mononuclear cells appear in the blood; these cells are similar to those seen in infectious mononucleosis. Primary infection in the neonate causes serious disease, and primary infection of pregnant women can cause congenital defects in the fetus. As with other herpes viruses, infection is lifelong. CMV replicates at a low level throughout life, but when immunity is suppressed the rate of replication increases. Patients with kidney or heart transplants are particularly at risk, and CMV sometimes causes severe disease in patients with AIDS (pp. 446-448). Transmission The virus is present in blood during primary infection and can cross the placenta. It can also be transmitted in blood used for transfusion. Infected organs transplanted from seropositive donors can cause primary infection in the recipient, but most CMV disease in renal transplant patients is caused by reactivation. The virus is shed from the cervix during active infection and is also present in semen and urine. Acute infection in normal adults There may be a mild febrile illness which is recognized as viral infection when atypical mononuclear cells are seen in a blood film. Sore throat, generalized lymphadenopathy and splenomegaly occur in less than one-third of patients. Abnormal liver function tests show that there is usually a mild hepatitis, but jaundice is rare. The average age for symptomatic infection is 30 years, about 10 years older than for infectious mononucleosis. Complications are rare in immunocompetent patients, and when they occur they are mild. Pneumonitis, myocarditis, meningoencephalitis, Guillain-Barre syndrome, thrombocytopenia and haemolytic anaemia have all been described. Infection in immunosuppressed patients Both primary infection and reactivation can cause disease in immunosuppressed patients, but in renal transplant recipients the more severe disease is seen during primary infection. Asymptomatic reactivation is common and is shown by culture of virus from urine. Fever, sometimes with atypical mononucleosis, is the most common presentation of disease, followed by interstitial pneumonitis. Interstitial pneumonitis presents with insidious onset of fever, a non-productive cough and dyspnoea. Chest radiography shows bilateral diffuse interstitial changes indistinguishable from those of many other opportunistic

infections. Diagnosis is usually made by transbronchial biopsy. Hepatitis is usually mild, but occasionally causes hepatic failure. CMV may be a harmless passenger in the gastrointestinal tract of patients. Its association with enteritis is strongly suggested by the coincidence of seroconversion and haemorrhagic lesions anywhere from the stomach to the rectum, causing a severe diarrhoeal disease and blood loss. A response to treatment with antiviral drugs supports the existence of this clinical entity. In the mouth and oesophagus of AIDS patients, CMV can cause large punched-out ulcers. 2 Severe choroidoretinitis causes progressive loss of sight, starting in one eye and then affecting the other. Large white exudates containing haemorrhagic areas occur. 3 Diagnosis Virus can be cultured from urine, throat washings or buffy coat during active replication, and IgM antibodies are detectable in serum by an immunofluorescence test. In histological sections 'owl's eye' cells are strongly suggestive of CMV infection. Management Intravenous ganciclovir, 5mg/kgb.d. for 14-21 days, is the treatment of choice for immunocompromised patients with CMV disease. AIDS patients with retinitis have a high risk of relapse with loss of vision, and maintenance treatment can be given with oral ganciclovir 1gt.d.s. Bone marrow toxicity is common and an alternative treatment is foscarnet 200mg/kg for 14-21 days, with monitoring of renal function and serum calcium. 4

Human herpes virus 6 infection Human herpes virus 6 (HHV6) has recently been identified in patients with AIDS and has subsequently been found to be widely distributed. It is the cause of roseola, a common febrile illness of young children characterized by the appearance of a rash when the fever and illness are resolving. The temperature is high but the child is relatively well. There is often cervical and postauricular lymphadenopathy. When the temperature starts to fall, or within 2 days, a maculopapular erythematous eruption appears on the trunk: this lasts for a few days. Blood count reveals neutropenia initially, which may be followed by mild leukocytosis. The illness resolves without complications.

Human herpes virus 8 (HHV8) or Kaposi's sarcoma (KS)-associated herpes virus HHV8 has been identified consistently in KS lesions of AIDS patients and it is now thought to be the cause of this spindle cell tumour. Epidemiological studies show that asymptomatic infection is common in Mediterranean and African countries where KS occurred independently of

HIV infection. It is likely that cofactors are required for the development of KS, but these have not yet been identified. Despite the relationship between KS and AIDS in homosexual men, transmission of HHV8 between individuals is likely to be similar to that of EBV, as infectious virions have been found in saliva. It is rarely detected in semen even though viral DNA is detectable in the prostate. Treatment of AIDS patients with antiretroviral drugs, particularly protease inhibitors, results in regression of KS lesions (see Chapter 11).

9

POX VIRUSES Variola and vaccinia Variola virus is the cause of smallpox, an infection of low infectivity transmitted between humans by close contact. Immunization introduced globally using the refined vaccine containing vaccinia virus led to the eradication of smallpox in 1979, and only accidental laboratory infections have occurred since then. Vaccination against smallpox is no longer justified.

Orf virus infection Orf virus causes a single nodular lesion at the site of an abrasion which has been in contact with an infected lamb.

ENVELOPED SINGLE-STRANDED DNA VIRUSES PARVOVIRUS The association between infection with human parvovirus and disease has only been recognized since the 1980s. The virus was first noticed in sera from normal subjects during screening by electron microscopy (1975), but in 1983 it was reported to be the cause of erythema infectiosum ('slapped cheek disease' or 'fifth disease'). A more serious consequence is that the virus can cause aplastic crises in patients with sickle cell disease.

Clinical features of erythema infectiosum Parvovirus infection is probably transmitted between humans by nasopharyngeal secretions and causes a range of illness ranging from asymptomatic to a febrile systemic illness with a rash. After a mild prodromal illness consisting of fever, headache, upper respiratory symptoms and myalgia lasting 2-A days, a rash appears on the arms and chest, spreading peripherally, even to the palms and soles. The classic 'slapped cheek' appearance of erythema infectiosum is seen in less than half of infected patients. Elsewhere the rash is maculopapular and fluctuates in intensity. It often develops a lacy reticular pattern on the limbs

289

and leaves a short-lived pale brown pigmentation as it fades over a week. The rash is frequently pruritic. Other features during the rash are generally mild, but include fever, malaise, headache and sometimes arthralgia, lymphadenopathy and splenomegaly.

Complications Complications are rare. Adult women may develop symmetrical peripheral polyarthritis, which can be the sole manifestation of infection. Although it usually resolves within 3 months it may continue for several years. Aplastic crises are not a feature of infection in patients who are otherwise normal haematologically, but when the red cell life is reduced, as in sickle cell disease or hereditary spherocytosis, arrest of erythropoiesis for 5-10 days causes a significant fall in haemoglobin level. The same effect has been observed in patients with advanced AIDS.

Diagnosis and management IgM serum antibodies detected by enzyme-linked immunosorbent assay (ELISA) or radioimmunoassay (RIA) are diagnostic of recent infection and persist for about 2 months. A change in litre of IgG antibodies also indicates recent infection. Virus isolation is not feasible and detection of viral protein is reserved for research purposes. There is no specific treatment at present.

NON-ENVELOPED DOUBLE-STRANDED DNA VIRUSES

affects adults and often follows minor trauma to the eye. Type 8 is the usual pathogen. Haemorrhagic cystitis. Caused by types 11 and 21, haemorrhagic cystitis is a childhood infection. Adenoviruses are also uncommonly associated with diarrhoeal disease, intussusception, and meningitis or encephalitis. Diagnosis of adenovirus infection is usually made on clinical grounds. It can be confirmed by virus isolation or by identification of viral antigen by immunofluorescence where these techniques are available, or by detecting a rise in serum antibody titre. There is no effective antiviral treatment, but most disease is mild and self-limited.

PAPOVAVIRUSES The important members of the papovavirus group are papilloma or wart viruses and polyoma viruses. Palmar and plantar warts are benign infections, common in childhood, which often regress spontaneously with increasing age. They are caused by viruses which are distinct from the anogenital warts viruses; the latter are sexually transmitted (see Ch. 11, p. 456, and Ch. 10, p. 393).

Genital warts See Chapter 11 (p. 456).

ENVELOPED SINGLE-STRANDED RNA VIRUSES ORTHOMYXOVIRUSES

290

ADENOVIRUSES

Influenza virus infection

Adenoviruses are a common cause of upper respiratory infection in children and cause a variety of other illnesses, depending on the serotype of virus and the age of the patient. Over 40 serotypes have been isolated from humans, and many of these are known to have pathogenic potential. Virus can frequently be isolated from the posterior pharyngeal lymphoid tissue of asymptomatic subjects, suggesting that it can remain latent for long periods. Upper respiratory infection. Common cold and pharyngitis occur in all age groups, and serotypes 1-6 are the usual cause. Lower respiratory infection. Tracheobronchitis and pneumonia are rare in children but young adults may be affected. This is one of the differential diagnoses of atypical pneumonia. Types 4 and 7 are the most common causes. Pharyngoconjunctival fever. An acute infection spreads among young children in close contact, often caused by adenovirus type 3. Epidemic keratoconjunctivitis. Epidemic keratoconjunctivitis is a more severe form of eye infection which

Influenza types A and B are the main viruses in this family, but there is also a type C virus. Influenza is an ancient disease characterized by regular winter epidemics, with occasional much larger outbreaks (pandemics) when there is a major shift in antigenic type against which there is little immunity in the population. Although the disease is normally mild and self-limiting, the elderly and those with chronic chest disease are susceptible to complications, resulting in a considerable mortality. Epidemiology Epidemics of influenza start with cases in children and spread to the adult population. As the epidemic progresses, vulnerable adults start to present in hospitals and the size of the epidemic becomes apparent when absenteeism causes problems at work and in schools. Epidemics usually peak within 3 weeks and last about 6 weeks. They occur in winter - December to April in the northern hemisphere and May to September in the southern hemisphere. In tropical countries virus isolates can be obtained year

round, but epidemics tend to occur after changes in weather, whereas in temperate climates it is rare to find the virus in the community between epidemics. Surprisingly, several variants of the virus are found to be circulating during an epidemic, but they are sufficiently similar to previously recognized viruses for some crossimmunity to exist. Variants that appear late in the course of an epidemic are sometimes the ones responsible for an outbreak during the following winter. Pandemics have occurred infrequently over the last century, in 1889, 1918, 1957, 1968 and 1977. They result from the emergence of new viruses in which there are major changes in the surface proteins, mainly neuraminidase and haemagglutinin, against which the populations of the world have no immunity. The mechanism for these changes is likely to be genetic reassortment and some evidence suggests that this occurs in animal reservoirs, such as birds and lower mammals. Two neuraminidases and three haemagglutinins have been detected at various times in the last century and these are recognized in the name given to a subtype of virus. For example, the virus responsible for the 1918 epidemic was H1N1, and another severe epidemic in 1957 was caused by H2N2. The severity of an outbreak is determined by the extent of change in antigenicity of the virus, so that if the neuraminidase and the haemagglutinin both undergo a major change, the severity of disease and the size of the epidemic are greater than if only one determinant changes. After a major antigenic shift causing a pandemic, smaller variations in antigenicity known as antigenic drift occur over a variable period (9-30 years in the past century) and the population gradually builds up immunity to the possible variants of the new virus subtype. The scene for the next major antigenic shift may be set by the level of herd immunity to the subtype variants, but other factors probably make a contribution also. Clinical features Influenza is transmitted by droplets of respiratory secretions and the incubation period is 1-3 days. Influenza A and B cause a similar illness, but influenza C only gives rise to sore throat. The onset of influenza is rapid, with symptoms of fever, including chills, sweats and sometimes rigors accompanied by malaise, anorexia, headache and marked myalgia. The severity of general symptoms varies greatly between individuals, from mild to prostrating. They are worst in the first 3 days and usually disappear within a week. Cough and sore throat are variable features which tend to increase as other symptoms subside and resolve over 2-3 weeks. On examination the patient is initially febrile, flushed and tachycardic, with minor nasal discharge. The pharynx may be inflamed and cervical lymph nodes palpable and tender. Complications The elderly are more vulnerable to all complications. The

most common is exacerbation of chronic bronchitis, and this is responsible for most adult admissions to hospital in influenza epidemics. Asthma is also exacerbated. Bacterial pneumonia, often caused by Strep, pneumoniae, Staph. aureus or H. influenzae, can occur in chronic bronchitics or patients with no previous history of chest disease, but it is more common in smokers. Viral pneumonia caused by influenza virus itself is a severe complication which usually occurs in patients with underlying disease of the chest or cardiovascular system, although in some epidemics it has affected previously healthy adults, suggesting that virus virulence is a factor. This type of pneumonia is bilateral, with opacities on the chest radiograph which do not correspond to clinical signs, low diffusing capacity and a progressive course. Other rarer complications include myositis, myocarditis, pericarditis, Guillain-Barre syndrome and encephalitis.

9

Diagnosis Influenza virus can be cultured from respiratory secretions, but the diagnosis is more often made by serum antibody measurement using paired sera. When an epidemic is in progress a clinical diagnosis suffices. Management and prevention Zanamivir (Relenza) has been the subject of controversy, as governments have been unconvinced that it should be licensed for general use. It is a competitive inhibitor of viral neuraminidase which is active against a range of influenza A and B subtypes, giving it a clear advantage over amantidine which is only active against influenza A. Zanamivir, given by inhalation and intranasally, has been shown to be effective in reducing virus shedding, shortening the duration of symptoms, and in prophylaxis. Trials have not been completed in groups at highest risk from influenza, including the elderly, and the cost/benefit ratio has been judged too high to justify licensing the drug at present. Amantadine and rimantadine, which interact with the M2 membrane protein of influenza A, have more sideeffects than zanamivir and are rarely used. Comparative trials have not been carried out. Symptomatic treatment with analgesics such as paracetamol and maintaining adequate hydration are important. Aspirin should be avoided in children because of the danger of precipitating Reye's syndrome. Secondary bacterial infection causing bronchitis or pneumonia is the most serious complication which requires antibiotic treatment with co-amoxiclav (250 mg amoxicillin t.d.s.) or a macrolide such as erythromycin (250 mg q.d.s.) in penicillin-allergic subjects. If staphylococcal pneumonia is suspected, flucloxacillin (250mg q.d.s.) should be given. Inactivated viral vaccine containing the currently prevalent strains of influenza A and B is recommended for persons at special risk, especially the elderly and those suffering from chronic pulmonary, heart or renal disease, diabetes, and those with some forms of immunosuppres-

291

sion. The vaccine is usually given in the autumn and confers about 70% protection for a year. O

PARAMYXOVIRUSES Para influenza virus infection There are four types of parainfluenza virus and the organisms are a common cause of mild upper respiratory infection in children under 2 years of age (Ch. 13). Types 1-3 occasionally cause more severe lower respiratory tract disease, such as croup and bronchiolitis. Although there is little antigenic variation in parainfluenza viruses, immunity is poor and reinfection causing mild disease is common.

Respiratory syncytial virus infection Respiratory syncytial virus (RSV) is the most important cause of lower respiratory tract infection in young children. Outbreaks occur annually, usually in late winter or spring, and about half of primary infections result in pneumonia, bronchiolitis or tracheobronchitis. Upper respiratory involvement is also common and otitis media is a frequent complication. Children over the age of 3 usually suffer milder disease, and adults are occasionally infected. Severe infection can occur in adults following bone marrow transplantation. RSV is highly infectious and is commonly transmitted to other patients in hospital. Respiratory isolation procedures should be applied and management is general support, with oxygen as required and maintenance of hydration. Successful therapy for RSV has been described in immunocompromised adults using ribavirin administered as an aerosol, but in practice this is difficult to apply.

Mumps Mumps is a systemic paramyxovirus infection which commonly causes inflammation of the salivary glands. It is transmitted by droplets from the mouth. Infection in childhood is common but infectivity is low, and many people reach adulthood without becoming affected. Clinical features The incubation period is about 18 days (Fig. 9.1, p. 261). Mumps is a mild disease in childhood and about one-third of children infected have no symptoms. In its mildest form there is fever with discomfort and swelling in the parotid glands, which lasts a week or less. In more severe forms there is considerable malaise, the glands are painful and there may be trismus. Other salivary glands are often inflamed and glands are sometimes involved sequentially.

1

292

MCQ 9.11

2

Fig. 9.7

Swollen parotid glands can be distinguished from lymphadenopathy because the angle of the mandible is obliterated by the former. The mouth is dry and there may be redness around the orifices of the parotid ducts, but there is no purulent discharge. About 20% of adult males develop orchitis in the course of mumps, but oophoritis is rare. Usually orchitis starts a few days after the onset of parotitis, but it may be earlier and may occur in the absence of parotitis. The swollen inflamed testicle is exquisitely tender and remains so for 3-4 days before gradually recovering. Orchitis is accompanied by fever and malaise, and the latter often persists for some weeks. Orchitis is bilateral in 15-30% of cases, but even in these patients sterility is extremely rare. Although some shrinkage of involved testicles may be detected shortly after the illness, this does not imply complete loss of function and it is important to reassure patients on this point. Mumps virus is highly neurotropic and involvement of the CNS is the most common complication of this infection. There are two distinct entities, meningitis and encephalitis. Lymphocytic meningitis occurs in about 5% of patients with mumps parotitis, but this figure underestimates its prevalence as it also occurs in the absence of parotitis. There is some seasonal variation in the clinical manifestations, meningitis with parotitis being more common in spring, whereas meningitis alone is seen more in summer. The time of onset of meningitis varies, but it usually becomes apparent a few days after parotitis. The CSF shows a lymphocytic pleocytosis and virus can be isolated from it. The course is invariably benign. Encephalitis is rare. It either occurs during parotitis as a result of direct viral invasion of the brain, or its onset is 7-10 days after that of parotitis, when it represents a postinfectious encephalitis. Most patients recover completely, but there is a mortality of 1-2%. Pancreatitis is rarely severe but may account for some of the abdominal symptoms in mumps. Some degree of abdominal pain is not infrequent and may be accompanied by vomiting. In severe cases there is marked abdominal tenderness and the serum amylase and lipase levels are elevated. Diagnosis A clinical diagnosis usually presents little difficulty, but if necessary the virus can be isolated from saliva and throat washings, or from CSF in cases of meningitis. Complementfixing antibody detection can be used to make a retrospective diagnosis. Management There is no specific treatment, but oral hygiene is important both for comfort and to prevent secondary infection. Orchitis requires testicular support and adequate analgesia. There is little evidence that corticosteroid treatment is beneficial in orchitis.

SUMMARY 5 Mumps

SUMMARY 6 Clinical features of measles

Fetal death possible if infection in first trimester

• • • • •

Gland involvement Parotid Submandibular Epididymo-orchitis Oophoritis Pancreatitis

65% 10% 25% (postpubertal) 5% (postpubertal) ?%

Meningitis Encephalitis

<10% 0.1%

Prevention Vaccination with a live attenuated vaccine was adopted as part of the routine programme in the UK from 1988. It is given in combination with measles and rubella in the second year of life.

Measles Measles is an easily recognized, highly infectious disease of childhood caused by measles virus, an RNA virus of the paramyxovirus group. Epidemiology Infection is spread by droplets coughed on to mucosal surfaces of the conjunctiva and nasopharynx, and because cough is a feature of the illness from the onset it is not surprising that infection spreads rapidly between non-immune subjects. One of the interesting features of the disease in older epidemics was the high mortality rate. It is not clear whether measles virus has changed to a more benign form or whether environmental factors are responsible, but the latter is more likely as the disease is still severe in Africa. The size of the pool of non-immune humans determines whether there is an epidemic of measles. In the past, epidemics have occurred in Britain every 2 years or so, but immunization has made this pattern less apparent. Pothogenesis Measles virus can infect many different cell types and during the incubation period, which is 8-11 days from infection to the onset of prodromal symptoms (Fig. 9.1, p. 261), the virus multiplies, becoming widely disseminated. It reaches most tissues, including skin, mucous membranes, lungs, gut, brain and lymphoreticular tissue. The lesions of measles, containing mononuclear and giant cells, are caused by a combination of viral cytopathic effect and non-specific and specific inflammatory reactions. The role of immune mechanisms is demonstrated by the modified or absent skin rash in immunodeficient subjects, who may nevertheless die from a giant cell pneumonia. Antibody is detectable in serum 24-48 hours after the onset of rash, and virus is rapidly cleared during this period.

9

Incubation period 8-11 days Initial symptoms of upper respiratory infection Koplik's spots on buccal mucosa Rash spreads from face to trunk to limbs Complications (pneumonia, otitis media, encephalitis) more common in malnourished children

Clinical features In Britain measles is most common at the age of 3-5 years but the age is lower in underdeveloped countries. In the 2-4 days before eruption of the rash the features are similar to those of many upper respiratory infections. The patient feels unwell and has a fever, red eyes, swollen eyelids, a congested streaming nose, diffuse redness in the mouth and a hacking cough. Koplik's spots, 1-2mm white spots like a grain of salt on a red background, may be seen on the buccal mucosa. Their number varies from a few to hundreds, and they often persist after the rash appears. The worst of the illness is during the first few days after eruption of the rash, when there is a spiking fever and the child is irritable and miserable. Spots appear first behind the ears and on the forehead and then spread to the rest of the face, on to the trunk and out to the limbs over 24-72 hours. © This pattern of spread is characteristic, but the rash of rubella (German measles) spreads in the same direction. Measles spots are red macules and maculopapules initially a few millimetres in diameter, which enlarge and coalesce to form irregular blotches, the most striking of which are on the cheeks. As the rash fades and the patient starts to feel better there is transient brown staining of the skin and sometimes fine desquamation. At the height of the eruptive stage the mucous membranes and eardrums are red, and there is generalized lymphadenopathy, a barking cough and widespread crackles in the chest. Complications Major complications are rare in Britain today. All complications are more common in malnourished children, and severe skin desquamation with secondary bacterial infection, diarrhoea, appendicitis, laryngitis, stomatitis and major weight loss all occur frequently in Africa. Measles sometimes precipitates the onset of kwashiorkor. Secondary bacterial infection usually manifests itself 3-4 days after the rash and prolongs the illness. The timing is an important clue, as renewed fever heralds a third phase of illness. Bronchopneumonia and otitis media are examples. Giant cell pneumonia is rare and usually affects immunocompromised children, such as those with leukaemias. Febrile convulsions may occur as the temperature rises in the incubation period or at the beginning of the eruptive phase. Encephalitis is rare and varies in severity. It occurs in the convalescent phase and in fatal cases virus is not recovered from the brain, so encephalitis may be

293

caused by an immune reaction to viral antigen in neural tissue. Subacute sclerosing panencephalitis (SSPE, p. 1426) is a progressive fatal form of encephalitis affecting older children who had measles many years before. High litres of measles antibodies and recovery of a measles-like virus from the brain suggest that it is related to measles. Diagnosis Clinical diagnosis of measles is usually straightforward. The rash of rubella has a similar distribution, but lesions are smaller and they tend to remain discrete. Rubella is a milder illness with a characteristic distribution of lymphadenopathy (occipital and posterior cervical). Laboratory confirmation of measles is rarely necessary, but measles virus can be cultured from nasopharyngeal specimens during the prodrome and for about 2 days into the eruptive phase. A blood count normally shows leukopenia, and neutrophilia suggests secondary bacterial infection. Antibody litres are measured by a complement fixation test. Management Most children with measles are looked after at home, but if hospital admission is required for social reasons or because of complications, the patient should be nursed in isolation. Antiviral treatment is not available at present and management is directed at providing comfort and maintaining hydration. Prophylactic antibiotics have no place, but appropriate antibiotics must be given for secondary bacterial infection. Important organisms in respiratory infection and otitis media are Strep, pneumoniae, H. influenzae and, less commonly, Staph. aureus. Therefore, the usual first choice of antibiotic is a broad-spectrum penicillin (e.g. amoxicillin) or erythromycin. Management of encephalitis is described on page 1423. Passive immunization with 500 mg normal human immunoglobin intramuscularly (250 mg for children less than 1 year old) attenuates or prevents clinical manifestations if it is given within 6 days of contact. This is useful in children under 18 months and in those made susceptible to severe forms of measles by immunosuppression. Protection lasts for about 3 weeks.

SUMMARY 7 Management of measles • • • •

General support including hydration Antibiotics for secondary bacterial infection Respiratory isolation Normal human immunoglobulin for children under 18 months if within 6 days of contact or immunosuppressed

1 294

MCQ 9.12

2

MCQ 9.13

FIG. 9.10 Notifications of, and deaths from, measles in England and Wales, 1940-2000 The impact of MMR vaccination is clearly visible.

Active immunization Live attenuated measles virus vaccine is given in the second year of life in Britain. In parts of the world where the disease is common in younger children immunization is recommended at an earlier age. A single subcutaneous or intramuscular dose gives prolonged protection. Contraindications are immune deficiency disease (primary or secondary), immunosuppressive therapy, and severe egg allergy. Atopy and febrile convulsions do not preclude use of the vaccine. The uptake rate for immunization has increased since the introduction of MMR vaccine in the UK and measles has become uncommon (Fig. 9.10), following the same pattern as in the USA. O

TOGAVIRUSES The togavirus group comprises the rubella virus, alphaviruses and flaviviruses. Rubella is the commonest togavirus infection in the western world. Alphaviruses are the cause of Eastern, Western and Venezuelan equine encephalitis. Flaviviruses cause dengue and yellow fever.

Rubella Postnatal rubella (German measles) is a mild illness occasionally complicated by arthritis or encephalitis. By contrast, infection in utero can have devastating effects on the fetus. Pathogenesis Rubella virus infects nasopharyngeal secretions, where it is found during subclinical as well as overt infection. During the viraemic phase it is disseminated widely and, as for

measles, the clinical illness corresponds in time with the development of immunity. The incubation period for rubella is 14-16 days. Clinical features Like measles, rubella is associated with fever, lymphadenopathy, rash and conjunctival suffusion, but it is a mild, short-lived illness which may be asymptomatic. The lymphadenopathy of rubella is characteristically suboccipital, postauricular and posterior cervical. It may persist long after the rash has faded. Although, as in measles, the rash starts on the face and behind the ears, spreading downwards and outwards over the trunk and limbs, the spots are paler red, circular macules which remain discrete and do not coalesce into brighter irregular blotches. The rash varies from hour to hour, fading and reappearing for up to 4 days. It is common to find petechiae on the soft palate (Forchheimer spots), but these are not diagnostic. Complications Arthritis is the only common complication of rubella. It usually affects young women, who develop polyarthritis of small joints, especially those of the fingers and wrists, although larger joints may be affected. It is easily confused with acute rheumatoid arthritis, especially in the absence of rash. It usually subsides within a few days but can last up to 3 months. Encephalitis and thrombocytopenic purpura are very rare complications. Congenital rubella Both the risk and the severity of damage to the fetus are closely related to the time of infection. Following infection during the first month of pregnancy over half of fetuses are affected, whereas by the fourth month only 5% have defects, although this is still twice the frequency in control groups. Minor damage can result from later infection. During the viraemic phase of the maternal illness rubella virus invades the placenta, causing villous placentitis, and then disseminates through the fetal circulation. The degree of damage is variable, even in twins infected at the same time. The common major abnormalities resulting from infection in the first trimester are cataracts, patent ductus arteriosus with or without pulmonary stenosis, and deafness. Rubella virus can be isolated from the baby's pharynx for several months, a much longer period than in acquired rubella. The virus is present in all other secretions, and these babies can readily infect nurses handling them. Diagnosis Serology is the mainstay of diagnosis in both congenital

SUMMARY 8 Clinical features of congenital rubella Cataracts Cardiac abnormalities - patent ductus, pulmonary stenosis. Deafness

and acquired rubella. In the haemagglutination inhibition (HAI) test, the virus agglutinates chick red cells. Serum antibodies inhibit this agglutination. A rise in total antibody titre or the presence of specific IgM antibodies indicates recent infection. In acquired infection antibody is detectable by HAI 14-16 days after infection and reaches a peak 6-12 days later. The fetus starts to produce IgM antibody at about the 20th week of life, and as only IgG antibody can cross the placenta the presence of IgM antibodies to rubella virus in the fetus or infant implies active infection. Rubella virus can be cultured from clinical specimens, but this is rarely used for diagnosis.

9

Immunization Live attenuated virus vaccine is given routinely as part of the measles/mumps/rubella (MMR) vaccination to boys and girls aged 15 months. This was introduced in 1988 and replaces the policy of immunizing girls aged between 11 and 13 years. Vaccine virus can cross the placenta so it should not be given during pregnancy, and pregnancy should be avoided for 2 months after immunization. Exposure to rubella during pregnancy Definite exposure to rubella of a non-immune pregnant woman during the first trimester is an indication for termination of pregnancy. If the woman does not know whether she is immune to rubella, the following course of action is recommended: • If possible, establish by serology the diagnosis of rubella in the patient with whom the pregnant woman was in contact. Find out whether contact was close enough to allow transmission of virus. • Test the pregnant patient for rubella antibodies: - // positive by HAI within 14 days of exposure to rubella, the patient was immune previously and she can be reassured. - // negative by HAI then test a further sample 2-3 weeks from contact. If the second serum is positive or if IgM antibodies are detected, discuss risks and offer termination. Some women prefer to continue with the pregnancy despite the risk to the fetus. In this case intramuscular rubella immunoglobulin should be given as soon as possible after exposure. 2

FLAVIVIRUSES Dengue Dengue is an arthropod-borne (arbovirus) flavivirus infection characterized by an acute febrile illness with severe bone pain, rash and lymphadenopathy. The vectors are Aedes mosquitoes. The dengue virus is a 50 nm RNAcontaining flavivirus. Four serotypes are recognized, and these can be differentiated in a range of standard serological tests. The serotypes are closely related antigenically,

295

FIG. 9.11 World distribution of dengue fever, 2000 (Source: Centers for Disease Control and Prevention)

but cross-protection between serotypes is lacking. The dengue shock syndrome and haemorrhagic fever occur in patients, particularly children, experiencing a second infection with a different Dengue serotype. Distribution and incidence Over the last three decades epidemics of dengue have occurred in all continents in the tropics, Africa, Asia, the Americas and the Pacific (Fig. 9.11), such that it is now the commonest mosquito-borne infection, and 50% of the world's population live in endemic areas. With the spread of dengue there has been a corresponding spread in the incidence of dengue haemorrhagic fever (DHF) which until the 1970s was confined to southeast Asia. The reasons for this spread are incompletely understood, but include the increase in travel worldwide distributing the serotypes to new areas, decline in mosquito control so that Aedes aegypti has proliferated, and the rise of densely populated periurban slums with no sanitary facilities except those that encourage the proliferation of A. aegypti. Transmission and epidemiology The most common vector is Aedes aegypti, a day-biting mosquito that breeds in small collections of water in tin cans, car tyres etc. peridomestically, but other species have more limited roles in transmission. Eight to 11 days after ingesting dengue virus in a human blood meal the mosquito is infective; it remains infective for the rest of its life. Peak biting times are the 2-hour periods after dawn and before dusk. Both sexes and all age groups are susceptible to dengue provided they have not been infected before with that serotype. Dengue haemorrhagic fever has been predominantly a disease of children, but with spread to new areas

1

296

Fig. 9.8

second infections, and hence DHF, are occurring in older age groups. Pathology and pathogenesis After inoculation the virus replicates in local lymph node cells. Viraemia follows and reticuloendothelial cells in skin and other tissues become infected. Local inflammatory changes occur around small vessels in the skin. Dengue haemorrhagic fever occurs in children who have previously been exposed to infection with a different serotype of the virus, or who have acquired antibody passively from their mother. Immune enhancement causes more cells to become more heavily infected through immune complex binding to Fc receptors. Cytotoxic T-cell memory is activated, causing enhanced clearance of virus but also resulting in excess cytokine release, with tissue-damaging consequences: increased vascular permeability, causing hypovolaemia and coagulation defects owing to the release of cytokines with procoagulant properties. Clinical features The incubation period is about 7 days prior to the onset of high fever, headache, eye pains, backache and chills. Limb pain is often severe in dengue and this gives rise to its common name, 'breakbone fever'. A blanching erythematous macular rash may appear on the third or fourth day of the illness. 1 Lymphadenopathy may be present. Encephalopathy, cardiomyopathy and liver damage also occur. Leukopenia is usual in the peripheral blood. In dengue haemorrhagic fever the patient is more ill, and blood pressure falls as a result of transudation of fluid from the vascular compartment. This fluid can accumulate in the abdominal cavity or in the pleural spaces. Monitoring blood pressure, haematocrit and platelet count in addition to urine output and conscious level give warning of its onset. There may be spontaneous bleeding into the skin and at other sites. The loss of circulating blood volume causes shock, with low blood pressure, rapid pulse, restlessness and abdominal pain (the dengue shock syndrome), which can have a mortality of 50% if untreated.

Diagnosis Virus can be isolated from blood using Aedes or mammalian cell lines. Virus can also be identified in blood using the polymerase chain reaction, and IgM antibody or a rising antibody titre in paired sera obtained 10-14 days apart are serological indicators of the diagnosis, although the serological response may not be so clear-cut in areas where populations are exposed to other flaviviruses. Management Patients with uncomplicated dengue require supportive measures, such as attention to nutrition and hydration and relief of pain. Paracetamol, not aspirin, should be used for pain relief and as an antipyretic. Intensive management of severe dengue haemorrhagic fever can reduce the mortality from 10% to 1% without highly sophisticated facilities. Intravascular volume is restored with Ringer's lactate, giving a volume equal to daily requirements plus 5% of body weight, and monitoring pulse, blood pressure and respiration to avoid overload. Plasma protein fraction and dextran 40 may also be given to replace colloid losses. Haematocrit should be monitored to assess progress. The need for fluid replacement may last only 1-2 days. No specific antiviral treatment is beneficial. Control Live attenuated vaccines are under development. It is important that they contain all four serotypes of the virus to avoid the risk of dengue haemorrhagic fever. Vector control comprises the removal of mosquito breeding sites, which are often collections of water in tins, tyres, tubs and water storage vessels around homes, and larviciding.

Yellow fever Yellow fever is an acute arboviral (arthropod-borne) infection caused by the yellow fever virus, a flavivirus. The illness is characterized by high fever, jaundice and encephalopathy in its severe form. The disease occurs in focal outbreaks in Africa, the Caribbean and Central and South America. This condition is prevented by the 17D attenuated live vaccine strain, which should be given to travellers to endemic areas and to indigenous populations in outbreaks. There are infrequent occurrences of importation of yellow fever to Europe, which can present as PUO or viral haemorrhagic fever. Distribution and incidence Sub-Saharan tropical Africa (especially West Africa), the Caribbean islands (most recently Trinidad) and South America (Brazil, Peru, Bolivia, Ecuador, Venezuela and Colombia) are the areas affected by yellow fever. Relatively small numbers of cases occur every year in these areas, although cases in isolated areas may go unreported. Transmission and epidemiology Transmission is shown in Figure 9.12. In South America, and Central and East Africa, a jungle cycle involves

9

FIG. 9.12 Cycle of transmission of yellow fever virus Infected humans bring the virus from the jungle. Mosquitoes can breed peridomestically in small collections of water such as puddles and water pots.

monkey-mosquito-monkey and maintains the virus in the mosquito reservoir. Haemagogus mosquitoes are the vector. The epidemiology in savannah and forest-savannah areas involves other Aedes species with humans and monkeys. The insect vector passes infection to the next generation of mosquitoes by transovarial transmission. Non-immune persons of all races, all ages and both sexes are susceptible to infection. Prior exposure to other flaviviruses produces some degree of cross-protection. Men are at particular risk of infection because of occupations that take them into forests. Pathology and pathogenesis The liver is the main organ involved and shows mid-zonal necrosis of hepatocytes. Councilman bodies result from the degeneration of hepatocytes. The kidneys show acute tubular necrosis which may relate to shock and hypovolaemia. There may be haemorrhage into mucous membranes and the skin, associated with the bleeding tendency that is common in yellow fever. The underlying pathogenic mechanisms are poorly understood. Clinical features There is a range of severity of disease, from mild to severe life-threatening illness. The latter constitutes a minority of all those infected. The incubation period is about 6 days before the onset of headache and fever. More severe illness is associated with marked limb pains. Proteinuria is usual. High fever, headache, severe limb and back pain, chills associated with fetor, haemorrhages in the gums, and nosebleeds are early features in severe cases. There may then be a short period of remission of symptoms for up to 24 hours before the recurrence of fever with vomiting and jaundice. Bleeding into the gut, skin and other sites is usual in severe cases. Occasional cases are seen with organ damage limited to the heart or kidneys, causing cardiac or renal failure. Bleeding and renal failure are the main causes of the high mortality (up to 50%) in severe cases. Resolution in severe cases can take from 3 days to 6 weeks.

297

Diagnosis The clinical features in severe cases suggest the diagnosis, but in milder cases without evidence of organ dysfunction the diagnosis may not be made. Virus can be isolated most often from blood taken in the first 4 days of clinical illness. Rising antibody titres in paired sera may also give the diagnosis, but serological tests may not be easy to interpret in people exposed to related viruses. Differential diagnosis In mild cases the range of febrile illnesses to be considered is extensive and includes malaria, typhoid, the prodromal phase of viral hepatitis, leptospirosis and rickettsial diseases. The presence of jaundice with fever prompts consideration of leptospirosis, malaria, East African trypanosomiasis, typhoid, biliary tract sepsis and Marburg virus diseases. Ebola virus fever and Lassa fever are other causes of haemorrhagic fever, although marked jaundice is not usual. Laboratory features Anaemia, leukopenia and thrombocytopenia are usual features in more severe cases. Conjugated bilirubin levels and transaminases are high in jaundiced cases. Coagulation abnormalities comprise prolonged prothrombin time, reduced fibrinogen levels and detectable fibrin degradation products. Renal failure with proteinuria, oliguria and raised creatinine and urea may occur. Prevention and control All travellers to endemic areas - apart from pregnant women, infants under 1 year and immunosuppressed patients - should receive the attenuated 17D yellow fever vaccine. Patients with yellow fever should be nursed under mosquito nets to prevent mosquitoes becoming infected. Vaccination is used to help control epidemics. Other arthropod-borne (arboviral) infections There are a considerable number of arboviruses that affect humans. In these infections the main clinical manifestations are fever, chills, joint pain and a maculopapular rash. Mild forms of the diseases are frequent. Vaccines are not available. Infections may emerge in cities if stagnant water is infected by mosquito larvae. The primary virus hosts are birds. A recent outbreak of acute viral encephalitis in New York is an example. 1

VIRAL HAEMORRHAGIC FEVERS (VHF) In the patient who has acquired a febrile illness in Africa, consideration must be given to the possibility of infection

1

298

MCQ 9.14

TABLE 9.18 Risk assessment for VHF A Strong suspicion Patient left a known endemic area in the previous 3 weeks, particularly healthcare personnel from such an area. Contact of confirmed case Laboratory worker who handled VHF viruses B Moderate suspicion Patient left tropical Africa in the previous 3 weeks, but from a rural area or small town not considered endemic Where the onset and clinical course of the fever are consistent with VHF C Low suspicion Patient left a major city in tropical Africa in previous 3 weeks where risk of VHF is negligible

with one of the agents of VHF, Lassa fever, Ebola or Marburg. It is necessary to make an assessment of risk based on information about the area visited, contact with sick people etc. (Table 9.18). In an endemic area (Table 9.19) isolation and observation of the patient will be necessary, depending on the likelihood of some other clinical diagnosis. Public health authorities should be informed to arrange serological and virological testing, epidemiological investigation, and initiate additional measures. For a patient seen in a non-endemic area the risk assessment outcome will determine where the case is nursed; for A and B (Table 9.18) that will probably be a high security isolation unit, whereas for C side-room accommodation with strict barrier nursing precautions will be adequate. The viruses causing VHF, a group of clinically similar infections, are from the bunyaviridae, arenaviridae and filoviridae (Table 9.20). They are characterized in the most severe forms by a bleeding tendency with purpura, ecchymoses in the skin, bleeding from mucosal surfaces, and often bleeding into lungs, gut and other internal organs; haemoglobin levels fall rapidly. Disease onset is abrupt, with severe headache, rigors, myalgia, and an erythematous flush of the face and chest. Often patients have been treated for infections over the first 10-12 days with antimalarials initially, followed by an antibiotic for typhoid. Progressive renal failure can occur. Not all patients with these infections develop severe clinical disease: some have a febrile illness which settles without these severe features. Treatment in severe cases is supportive, with restoration of blood volume by transfusion with platelet support if possible, and dialysis if there is renal failure. Lassa fever is an exception in which ribavirin is known to be effective. In managing such patients it is essential at the outset to diagnose more common, treatable, causes of fever, particularly falciparum malaria.

Lassa fever This acute febrile illness was first described in patients who came from the village of Lassa in eastern Nigeria near the border with Cameroon. It is an arenavirus (RNA) infection characterized by high sustained fever, ulcerative pharyngitis and swelling of the neck in typical cases.

TABLE 9.19 The distribution of viral haemorrhagic fever viruses Lassa fever virus Ebola virus

Marburg virus Arenaviral haemorrhagic fevers in South America Hantaviral haemorrhagic fever with renal syndrome

Guinea, Liberia, Sierra Leone, Nigeria Congo, Gabon, Sudan, Cote d'lvoire, Uganda Reston strain in monkeys from Philippines (infectious but avirulent in humans) Uganda, western Kenya, ?Zimbabwe Brazil, Argentina, Bolivia Asia, the Americas, Europe

TABLE 9.20 Causes of viral haemorrhagic fever Type of virus

Bunyaviridae Hanta Crimean-Congo HF Rift Valley California encephalitis Arenaviridae Lymphocytic choriomeningitis Lassa Junin Machupo Guanarito Sabia Filoviridae Marburg Ebola

Disease HF with renal failure HF Fever (+retinitis in 10%); HF in 1% Encephalitis/meningoencephalitis Aseptic meningitis Lassa fever Argentine HF Bolivian HF Venezuelan HF Brazilian HF

HF HF

HF = haemorrhagic fever

Transmission and epidemiology Mastomys nataliensis, the multimammate rat, is the reservoir of infection and persistently excretes virus in its urine, which contaminates food eaten by humans. Person-toperson spread occurs through contact with body fluids. Infection can occur through cuts and needle-stick injury, or contact of body fluids with mucosal surfaces. In endemic areas, which include Nigeria, Sierra Leone, Liberia and other west African countries, there is a range in severity of disease with appreciable numbers of mild infections; severe infection occurs in 5-10% of cases with a hospital mortality of 15%. Children develop less severe disease than adults. Clinical features After an incubation period of up to 17 days there is a gradual onset of fever with anorexia, lethargy and headache. In the second week of illness ulcerative pharyngitis and swelling of the neck and face may occur. Vomiting, diarrhoea, cough and abdominal and chest pain also occur. A maculopapular rash may appear on the face, trunk

and upper limbs in the second week. Bleeding occurs in the third week and affects gums, nasal mucosae, gut and other sites. Fatal cases show declining renal function and oliguria, declining conscious level and secondary bacterial infections. Leukopenia, low albumin, raised urea and elevated transaminases with normal platelet counts are also evident. Neutrophil leukocytosis can be a feature later in the course. Platelet function is abnormal. An aspartate transaminase of greater than twice normal is associated with rising mortality.

9

Diagnosis This may have to depend on clinical features in endemic areas, but virus can be identified in body fluids by electron microscopy, in cell culture and by PCR. Serological testing is also used. Management Supportive measures are used, with maintenance of fluid balance, analgesia and antipyretics, avoiding aspirin because of impaired platelet function. Ribavirin given intravenously reduces mortality if given in the first week of the illness: 30mg/kg as a loading dose, followed by 15mg/kg 6-hourly for 4 days, then 7.5mg/kg 8-hourly for a further 6 days. Close contacts should be treated immediately they develop a fever. Diagnosed cases in Europe and North America are managed in secure isolation facilities to give maximum protection to staff caring for the patient, but in endemic areas standard barrier nursing precautions are adequate. Oral ribavirin treatment may be given to staff and others exposed to the virus by needle-stick or other injury.

Marburg and Ebola virus disease These viruses also cause VHF. They are members of the filoviridae. Marburg was first recognized in 1967 as causing an outbreak among staff at a laboratory in Marburg, Germany, as a result of handling a recently arrived shipment of African green monkeys from Uganda. There were also cases in former Yugoslavia as a result of exposure to monkeys from the same shipment. The next outbreaks were of Ebola virus infection. These occurred in southern Sudan and in what was then Zaire in 1976; despite the similar temporal occurrence these outbreaks were caused by different strains of the same virus. Subsequently there have been outbreaks in Sudan, Gabon, Cote d'lvoire and Zaire. There were also outbreaks of disease in non-human primates. Epidemiology and transmission No reservoir has been identified for filovirus infection, though it is likely to be an animal. Chimpanzees and a range of monkey species die of the disease and so are not the reservoir. The average incubation period in one outbreak was 7 days. During the 7 months of the outbreak 80% died, with a decline in mortality over the 7 months. The mean duration from onset of disease to death was 10

299

days. A quarter of all the cases were healthcare workers and this was largely due to contact with blood and body fluids. Reuse of contaminated needles and instruments and not wearing gloves for procedures were also factors in nosocomial transmission. Direct contact with body fluids from patients and with the bodies of dead patients were additional factors in transmission. Aerosol transmission was not thought to be a major factor. Subclinical infection is very uncommon. Seropositivity rates were only 2% in the nearby town and 9% in surrounding villages. Clinical features Three manifestations present relatively early in the course are conjunctival injection, a maculopapular rash, and sore throat with marked pain on swallowing. Severe muscle pain is often present. The rash appeared towards the end of the first week on flanks and in inguinal and axillary regions, then rapidly spread to cover the whole body apart from the face over hours. This is evident on white skins but not on black skins. Bleeding from mucosae and puncture sites, anuria, hiccup and tachypnoea were features indicating that death was likely in days, with defervescence over the last 2 days before death. Fever, asthenia, nausea, vomiting and diarrhoea - often dysenteric - and headache were additional common symptoms. Diagnosis Intense viraemia is the rule in cases, and so antigen detection using ELISA-based technology is the most valuable diagnostic technique, particularly when processing large numbers of samples. Serology, PCR and virus isolation are other methods for diagnosis. Management This is essentially supportive. Paracetamol and not aspirin should be used as antipyretic. Fluid intake and nutrition should be maintained and blood transfusion given where there is anaemia. In an endemic situation there may be very little that can be offered, and the major efforts need to go into protecting healthcare workers and family members from infection by caring for cases in hospital with strict barrier nursing precautions and attention to disinfection. No antiviral agents are effective in this infection. 1

RHABDOVIRUSES

Rabies Rabies is an acute encephalomyelitis caused by an RNAcontaining rhabdovirus. It is characterized by paroxysms of muscle spasm that particularly affect muscle groups associated with swallowing and inspiration, in response to attempts to drink water, as well as the sight, sound or mention of water. For practical purposes it is invariably fatal. Aetiology The rabies virus is an RNA-containing rhabdovirus which, with five related viruses, comprise the Lyssavirus genus. The virus can be grown in cell culture. Distribution and incidence Antarctica and Australasia are free of rabies, otherwise all continents are endemic; the UK, Scandinavia, Japan and Taiwan are free of the disease. The incidence of human rabies is difficult to estimate. Occasional cases occur in Europe and North America. It is a considerable health problem in Asia, Africa, and Central and South America. Transmission and epidemiology Rabies virus is inoculated in saliva by the bite of a rabid animal, most often a dog, less often a cat (Fig. 9.13). Licking abraded or cut skin can transmit infection. Virus can penetrate mucous membranes of the mouth and eyes. Endemic canine rabies is the major determinant of risk for human infection. All mammals can become infected, but transmission to humans is mainly related to biting animals, commonly stray dogs in the tropics. Wolves, jackals, mongooses, bats (insectivorous, vampire and fruit eating), monkeys and ungulates (sheep, cattle and deer) can all become infected and pose a greater or lesser degree of risk to humans. Apart from bats all species die of this infection. In Europe the red fox is the reservoir of infection. Vampire bats infect cattle in

FURTHER READING Peters C J, LeDuc J W, eds. 1999 Ebola: the virus and the disease. J Infect Dis 179; Suppl 1:288.

1

300

Case 9.2

FIG. 9.13 Cycle of transmission of rabies The commonest routes of spread are shown in solid lines. Humans are usually infected by dog bites.

the countries of South America, causing considerable economic losses through cattle rabies. Human infection has been acquired directly from vampire bats and following exposure to aerosols containing virus excreted by bats in bat caves. Pathology Virus first penetrates skeletal muscle cells and then enters nerves through the fibres supplying muscle spindles and travels centripetally to infect nerve cells. Virus replication occurs and then ruptures cells, releasing virions to infect other neurons. Neuropathological changes are not gross. Collections of inflammatory cells are seen around small vessels; dead neurons and phagocytosis of degenerating neurons are seen especially in the medulla and the midbrain. Viraemia is not thought to contribute to spread of the virus; rather, it occurs through centrifugal spread via fibres of the autonomic nervous system to cornea, skin and salivary glands. Clinical features The incubation period is 20-60 days on average, ranging up to 1 year. Initial symptoms are headache, fever, insomnia, anorexia and paraesthesiae or altered sensation at the site of the bite or some peripheral site. After 2 or 3 days the patient may become markedly agitated, hyperexcitable, anxious, confused and lucid by turns. The earliest hint of hydrophobia may be evident at this stage, with rapid progression to overt hydrophobia, with involuntary contraction of the main and accessory inspiratory muscles in a sustained spasm brought on by attempts to drink water or at the sound, sight or mention of water. Variations of mood persist. Saliva may be continuously spat out as it cannot be swallowed. The spasms can be provoked by fanning air across the face. Cranial nerve palsies, paralysis and bizarre patterns of breathing follow. Ascending paralysis without hydrophobia occurs in a few patients, and this can pose problems in diagnosis. Death occurs in 7-14 days from respiratory failure or cardiac arrhythmias. Differential diagnosis Tetanus, other causes of ascending paralysis, and encephalitis and rabies hysteria are the main things to consider. Laboratory diagnosis Virus may be grown from saliva, nasal washings, CSF and urine. Viral antigen can be identified in skin biopsies, preferably taken from the hairy skin of the neck, where hair follicles contain peripheral nerves, or less easily in corneal impression smears using fluorescent antibody labelling techniques. The sensitivity of this method for antemortem diagnosis is not ideal. Corneal impression smears and skin biopsies for antigen detection should be refrigerated immediately after collection. In specialist centres PCR is available using reverse transcriptase PCR technology, and this method can detect virus in saliva. Blood and CSF should also be tested for rabies antibodies, and in the unvaccinated patient a positive test supports the diagnosis.

Prevention, control and treatment Correct management at the time of biting prevents rabies.

9

First aid The wound should be washed under running water for 5 minutes with soap or detergent. Where the wound is washed with soap all traces of this must be washed away if a quaternary ammonium compound is to be subsequently applied, as soap inactivates cetrimide. Primary suture and scrubbing of the wound should be avoided if at all possible. Alcohol 40-70% or an aqueous solution of iodine or a quaternary ammonium compound with known lethal effect on rabies virus, such as 0.1% cetrimide solution, should then be applied and the wound covered with a dressing. Tetanus prophylaxis should be given and antibiotics as indicated. The need for rabies vaccine must then be considered. • Low risk: following licks of the skin, abrasions or scratches, minor bites on covered areas of arms, trunk and legs, start vaccine (schedules given below) if animal is rabid or suspected rabid. If the animal is rabid at the time of exposure, or is wild or not available for observation, then vaccine plus rabies imimmoglobulin (RIG) should be started together. If the biting animal is a dog or cat and is healthy after 15 days of observation, then further vaccination can be stopped. • High risk: following licks of the mucosa, major bites, i.e. multiple bites, or bites to the head, face, fingers or neck by a suspect or rabid domestic or wild animal, or by an animal that is not available for observation, give vaccine plus rabies immunoglobulin. Again treatment can be stopped if the biting cat or dog is healthy at 15 days. Rabies vaccination The vaccine used now is human diploid cell vaccine (HDCV). • For pre-exposure vaccination (e.g. prior to travel etc.) the recommended schedule is three doses of 1.0 mL of HDCV given by deep subcutaneous or intramuscular injection (deltoid region, not gluteal region) on days 0, 7 and 28. • For low-risk cases HDCV should be given; for those who have not received full pre-exposure immunization give five doses of 1.0 mL of vaccine intramuscularly, starting on the day of biting, day 0, and thereafter on days 3, 7, 14 and 30. The vaccine is given by deep subcutaneous or intramuscular injection. If intramuscular, the injection is given into the lateral aspect of the thigh or the deltoid but not the gluteal region. For those who have had full pre-exposure vaccination, i.e. three doses of HDCV, two doses of vaccine are given on days 0 and 3-7. When RIG is needed it is given as detailed below. • For high-risk cases human RIG is given, 20iu/kg body weight, up to half infiltrated in and around the wound after cleansing and half given intramuscularly. In some parts of the world human RIG will not be available and equine RIG will be used. The dose is worked out on the

301

basis of 40iu/kg body weight. As reactions to equine RIG are not uncommon a test dose should be given first. HDCV is given as five doses of 1.0mL of vaccine on days 0, 3, 7, 14 and 30 to those with no or incomplete preexposure vaccination for rabies. For those with full preexposure vaccination two additional doses of 1.0 mL of HDCV are given on days 0 and 3-7, RIG is not usually needed. While patients with established rabies can be kept alive by intensive care, the fatal outcome is unaltered. Postmortems carried out in patients managed in this way have shown extensive destruction of nervous tissues. Practical management in an endemic area comprises sedation and pain relief. Close contacts of cases should be vaccinated. The problem of controlling reservoirs of infection remains. The capturing and killing of stray dogs in endemic areas has reduced an important source of human rabies. Vaccination of household dogs and cats is a helpful measure but does not exclude the possibility of the animal contracting rabies. Good results have been obtained in the control of fox rabies in Europe by putting down baits containing an oral antirabies vaccine. The incidence of rabies in sentinel species has declined markedly. Advice about rabies vaccination etc. can be obtained in the UK from Virus Reference Division, Central Public Health Laboratory, Colindale, London NW9 5HT (Tel: 020 8200 4400).

NON-ENVELOPED SINGLE-STRANDED RNA VIRUSES PICORNAVIRUSES Rhinovirus infection Rhinoviruses are the most common cause of colds (p. 631). There are over 90 serotypes and reinfection with the same serotype can occur. Infection is usually confined to the nasal mucosa, but complications such as sinusitis and otitis media can occur as a result of mucosal swelling and

obstruction to drainage of secretions. Rhinoviruses can also cause exacerbations of chronic bronchitis but the mechanism for this is not clear.

Enterovirus infection Coxsackie A, Coxsackie B and Echoviruses, of which there are 24, 6 and 34 types, respectively, cause a wide variety of clinical infections, but overall most are asymptomatic. A febrile illness with upper respiratory symptoms is another common manifestation of infection with many different serotypes. More specific disease associations are shown in Table 9.21. Polio viruses types 1, 2 and 3 cause poliomyelitis, which is described on page 1426. Acute meningitis Enteroviruses are a common cause of lymphocytic meningitis (more than 70% of cases), which is a benign disease. Enteroviral meningitis is common in young children and rare over the age of 40. The clinical features are described on page 1422. Other neurological disease Encephalitis is a rare manifestation of enteroviral infection which may occur with or without meningitis. Flaccid motor paralysis is only seen with any frequency in poliomyelitis. Fever and exanthem A variety of skin rashes occur. Maculopapular eruptions associated with fever are more common with Echo than with Coxsackie viruses. Petechial or purpuric eruptions are uncommon. The most distinctive exanthem is the vesicular eruption of hand, foot and mouth disease caused by Coxsackie virus A16 (and occasionally others). Lesions are found in the mouth consistently, and on the palms, soles and extensor surfaces of the hands and feet. Vesicles are sometimes surrounded by a ring of erythema and lesions may be slightly tender. The child feels unwell for 1 or 2 days and has a fever, but recovers completely. Virus can be isolated from vesicles, throat washings or stools. Respiratory tract infection Most enteroviruses can cause non-specific upper respiratory infection, but group A Coxsackie viruses are the

TABLE 9.21 Enterovirus types and their disease associations Virus types most commonly associated

302

Disease

Coxsackie A types

Coxsackie B types

Echo types

Acute meningitis Encephalitis Herpangina (pharyngitis) Hand, foot and mouth disease Pleurodynia (Bornholm's disease) Pericarditis/myocarditis Chronic meningoencephalitis in hypogammaglobulinaemics

Most One-fifth 2-6, 8, 10, 22 5,7,9,10,16 None None None

All Most None None 1-5 1-5 None

Most One-third None None None None One-third

common cause of herpangina, an acute painful vesicular eruption on the fauces and soft palate associated with fever, malaise and myalgia and lasting up to 10 days. Pleurodynia (Bornholm disease) Pleurodynia (sometimes known as epidemic myalgia) is an acute febrile illness that usually affects young adults and is characterized by 'pleuritic' chest pain; it is caused by infection with group B Coxsackie viruses which occurs mainly in late summer. The pain is due to infection of muscle and there is no pleural inflammation. About half of patients have pain in an area of the thorax but upper abdominal pain occurs in others, especially children, and this may mimic an abdominal emergency. Tenderness over involved muscles is the characteristic sign. Chest X-ray is normal. The illness is usually over in 5 days, but it may last up to 3 weeks and sometimes recurs. Non-steroidal antiinflammatory drugs such as ibuprofen 400 mg t.d.s. are helpful in treating the pain. Myopericarditis Enteroviruses, mainly group B Coxsackie, are an important cause of acute myocarditis and pericarditis in young adults. Usually the two conditions occur simultaneously, but one predominates in the clinical presentation. The clinical features are described on page 570.

SUMMARY 9 Management of rotavirus diarrhoea Clinical assessment No dehydration

Management Encourage ingestion of fluids orally Encourage normal feeding

Some dehydration (5-10% body weight lost)

ORS 80 ml/kg body weight over 4 hours + 400 ml ORS after each diarrhoeal stool + water to provide for insensible and renal losses

Severe diarrhea (>10% body weight lost)

Intravenous fluids (e. g. Ringer's lactate) 30 ml/kg over 30 minutes and then 70mL/mg over 2 hours

(Note: dehydration is not usually severe in adults with viral gastroenteritis)

Epidemic haemorrhagic conjunctivitis This epidemic form of acute conjunctivitis, which is associated with pain, subconjunctival bleeding and oedema of the eyelids, lasts about 1 week and is usually caused by enterovirus 70. The virus is highly infectious and is spread by contact and fomites rather than by droplet.

9

NON-ENVELOPED DOUBLE-STRANDED RNA VIRUSES REOVIRUSES Rotavirus diarrhoea Rotavirus diarrhoea is an acute dehydrating diarrhoeal disease caused by a 70 nm RNA-containing virus. This organism is the commonest cause of diarrhoea in children up to 2 years old in the tropics. It is spread by the faecal-oral route and possibly by droplet transmission. The infection occurs in both endemic and epidemic forms worldwide. Clinical features After an incubation period of up to 5 days (most often 2 days) anorexia, vomiting, diarrhoea and fever begin suddenly. There may also be a blocked nose and sore throat. The stools are watery and non-dysenteric. The illness is self-limiting, terminating in about 5 days in most cases. Moderate dehydration is the usual consequence of the illness, but a number of fatalities from severe dehydration have occurred. The organism may be found in faeces by immune electron microscopy and viral antigen detection using an enzyme-linked immunosorbent assay. Management Management comprises assessment of the degree of dehydration (see Table 9.14, p. 280) and replacement of that volume of fluid in the first 4-6 hours after presentation. Continuing fluid losses must also be replaced until diarrhoea ceases (see p. 280).

CHLAMYDIAL, MYCOPLASMA AND RICKETTSIAL INFECTIONS CHLAMYDIA Chlamydia are bacteria which contain both DNA and RNA (unlike viruses) but which are unable to replicate extracellularly (Table 9.22). They are Gram-negative and have a lipopolysaccharide endotoxin. Chlamydia psittaci, a common pathogen of birds and animals, can infect humans, causing psittacosis (ornithosis). Chi. trachomatis is transmitted between humans by sexual contact or from a mother's infected genital tract to her

newborn infant, causing conjunctivitis or occasionally pneumonia. Different serotypes of Chi. trachomatis cause lymphogranuloma venereum (LGV; see p. 457) and trachoma. The role of chlamydial infection in non-specific genital infection (NSGI) is discussed in Chapter 11 (p. 457).

Psittacosis Although ornithosis is a commonly used name for infection with this organism, this is a misnomer because a wide range of animals other than birds are infected by Chi. psittaci.

303

TABLE 9,22 Chlamydial infections Type

Transmission

Disease

Clinical features

Chlamydia psittaci Chlamydia trachomatis Types A, B, C Types D-K

Zoonosis

Psittacosis

Pneumonia

Fingers, flies, fomites Sexual Mother-to-infant Sexual

Trachoma Non-gonococcal urethritis, salpingitis

Keratoconjunctivitis Urethritis, pelvic inflammatory disease, Reiter's syndrome, perihepatitis (rare) Conjunctivitis, pneumonia (rare) Suppurative inguinal lymphadenitis

Types L1, L2, L3

Lymphogranuloma venereum

Epidemiology Psittacosis occurs sporadically worldwide, but it is most common in people closely associated with birds. Infected birds are usually ill, but continue to shed organisms in excreta and feathers for months after recovery. Humans are infected by inhalation. Clinical features After an incubation period of about 2 weeks there is fever, malaise, headaches, myalgia and sometimes arthralgia. Dry cough and variable dyspnoea occur, with confusion and clouding of consciousness if severe hypoxia develops. There are few abnormal signs in the chest, although there may be fine crackles in the lower zones. The chest radiograph shows unilateral or bilateral patchy infiltrates, especially in the lower zones, but occasionally there is the appearance of lobar consolidation. Gastrointestinal symptoms occur uncommonly. Rarely the disease is complicated by aortic or mitral valve endocarditis. Abortion in pregnant women may occur. Diagnosis and management Diagnosis is suspected from the history and evidence for atypical pneumonia, and confirmed by testing for serum antibodies by immunofluorescent or complement fixation tests. Tetracyclines given for at least 10 days are the most effective treatment.

Trachoma Trachoma is a chronic infection of the conjunctiva of the globe and eyelids caused by Chl. trachomatis serotypes A, B, Ba and C. It is widespread throughout the tropics and subtropics, and is primarily a disease related to poor personal and public hygiene. It is one of the commonest causes of blindness in the world, with 500 million affected, 2 million blind and 100 million with severe visual impairment. Transmission and epidemiology The organisms present in ocular secretions infect other people by direct contagion on contaminated fingers or

304

1

Fig. 9.9

4

Fig. 9.11

2

Fig. 9.10

3

MCQ 9.15

clothes. Lack of water for washing means that infected secretions are wiped on to hands and clothes. Flies may also carry the organisms from one person to another. Infection occurs early in life, and subsequently there are episodes of reinfection and secondary bacterial infection. Pathology and pathogenesis Acute inflammation is present in the conjunctiva. This progresses to a chronic state, with lymphoid aggregates around infected conjunctival cells. It is likely that damage due to trachoma is exacerbated by repeated bacterial infections. The tarsal plates fold inwards due to fibrosis (entropion) and the cornea is continually damaged by the inverted eyelashes, resulting in corneal opacity from scarring. There is evidence that cell-mediated immune responses to this intracellular pathogen are important in pathogenesis, with evidence of genetic factors determining the expression of reactivity. Clinical features The earlier features are hypertrophic follicles on the upper tarsal conjunctiva and at the corneoscleral junction superiorly. 1 With time and repeated reinfection this inflammatory response leads to fibrosis. The scarring disorganizes the arrangement of Meibomian glands, and their orifices are pulled away from the lid margin. The eyelid is inverted and sites of eyelash trauma on the cornea are seen. New vessels grow into the cornea (pannus formation) and corneal scarring follows. 2 Diagnosis Diagnosis is made on the appearances described above. Giemsa staining of scrapings from the affected conjunctiva shows intracytoplasmic inclusion bodies. Antigen can be detected by enzyme-linked immunoassay. Management and prevention Topical 1 % tetracycline eye ointment, applied two to four times daily for 3 weeks and repeated at 1-month intervals up to six times, is effective but it has been shown that a single dose of oral azithromycin, 20mg/kg, was as effective as 6 weeks of topical 1 % tetracycline eye ointment, though neither gave permanent cure and relapse is frequent. Continued trauma to the cornea must be prevented. Removal of the offending eyelashes helps. Surgical reconstruction of the eyelid may be needed.

Education concerning the importance of washing with soap and water is an important measure in preventing transmission. This depends on the availability of an adequate water supply sufficient for all domestic needs. The use of antibiotics in infected persons reduces the inflammation in the eyes and prevents progression of the disease while eradicating sources of infection for others. 0

MYCOPLASMAS Mycoplasmas are the smallest organisms able to grow in cell-free media, wherein they differ from viruses. They differ from bacteria in lacking a cell wall. Mycoplasma pneumoniae is an important cause of atypical pneumonia and a number of other clinical syndromes described in Chapter 13. M. hominis and Ureaplasma urealyticum can cause non-specific genital infection (see p. 457), prostatitis, pelvic inflammatory disease and postpartum fever. They are sensitive to tetracycline, the treatment of choice (p. 304).

RICKETTSIA Rickettsial infections comprise a group of acute infectious diseases caused by small bacteria which are transmitted by arthropods that in some cases act as reservoir and amplifying hosts. Fourteen rickettsial infections are currently recognized, eight have been described in the last decade, and there may well be more to be discovered. Some rickettsiae are currently regarded as non-pathogenic, but it may well be that all are pathogenic to some degree. The main groupings recognized are the typhus group of infections and the spotted fever group. The agent of scrub typhus has now been assigned to a new species, Orientalis tsutsugamushi. Current approaches to taxonomy have placed much reliance on homology between ribosomal 16S RNA sequences. The rickettsiae produce diseases with a range of severity from mild to severe and life-threatening. All are obligate intracellular parasites living free in host cell cytoplasm, not confined within a parasitophorus vacuole, and can only be grown in cells in vitro. They are Gram-negative organisms and, like Gram-negative bacteria, produce endotoxin which may contribute to the pathogenesis.

Typhus Distribution and incidence (Table 9.23) Typhus group Louse-borne typhus (Rickettsia prowazekii) occurs in louse-infested populations. These are usually relatively poor people living where night-time temperatures are low even though it may be hot in the day; people wear as much clothing as they can to keep warm, and do not change or

TABLE 9.23 Geographic distribution of the main rickettsial infections Epidemic typhus (R. prowazekii) Murine typhus (R. typhi) Rocky Mountain spotted fever (R. rickettsii) Mediterranean spotted fever (R. conori)

9

Highland regions: South America, Africa and Asia Texas (USA), Africa, Asia, Europe USA, Canada, Mexico, Panama, Costa Rica, Brazil, Colombia Countries of the Mediterranean littoral, sub-Saharan Africa, India, around the Black Sea, Vladivostok

wash their clothing. Body lice are frequent biters and feeders and are very sensitive to dehydration. If clothes are changed and washed then transmission of louse-borne typhus is prevented. Mountainous areas of South America, the Himalayas and highland regions of Ethiopia are endemic areas. Louse-borne typhus had become less common, but there was a large outbreak of 30000 cases in Burundi in relation to the civil war there, with large numbers of displaced persons crowded together in unsanitary conditions. In louse-borne typhus humans are the reservoir of infection, with latent infections recrudescing under the stress of displacement and war (Brill-Zinsser disease), infecting body lice which spread from person to person carrying the infection. Murine typhus (R. typhi) is widely distributed in warmer climates in Africa, Asia, Europe and southern parts of the USA. The rat flea spreads the infection. Spotted fever group Rocky Mountain spotted fever (R. rickettsii) occurs most often in the hills and mountains of the eastern seaboard and southern states of the USA. South American countries are also affected. African tick typhus affects the countries of North Africa, and eastern, central and southern Africa. The incidence is unknown. African tick typhus, caused by either R. conori (Mediterranean spotted fever) or R. africae, is the commonest imported rickettsial infection in the UK. Rickettsial pox (R. akari) infections occur in Asia, Asian regions of the CIS (former USSR) and Korea. With increased interest in rickettsiae, pathogens such as R. helvetica, first described from Switzerland, with a widespread distribution through Europe, R. sibirica in Siberia and China, and R. slovaca from the Balkan region, have been recognized. Transmission and epidemiology Typhus group Louse-borne typhus is transmitted by Pediculus humanus, the body louse (Fig. 9.14). The head louse does not transmit this infection. After taking a blood meal the louse defecates, and R. prowazekii in the faeces of an infected louse is scratched into the bite site or abrasions in the skin. 4 Humans are the reservoir for R. prowazekii, and infection

305

Reservoir Humans

Reservoir Small mammals

Reservoir Larval mites

Reservoir Ticks

Body louse

Fleas

Larval mite

Tick

Louse-bome typhus

Murine typhus

Scrub typhus

Tick typhus

FIG. 9.14 Transmission of rickettsial infections

passes from one person to another as the lice pass between people. The louse dies by 14 days after infection. Epidemic disease occurs when infection is introduced into a large group of people not previously exposed. Refugee camps present an ideal situation for such epidemics. Spotted fever group This is transmitted by the bite of infected ticks, which are also the reservoir of infection. Occupational and leisure activities, such as walking, camping, farming, lumber work and hunting, take people into tick-infested areas in North America. Tourism, tending animals and gathering wood take people into similar areas in Africa. Pathology and pathogenesis These organisms replicate within vascular endothelial cells of small vessels in many tissues, and vasculitic responses and blood vessel damage underlie the pathogenesis. R. rickettsii infection is more invasive than other organisms, and causes more severe illness owing to full-thickness vessel wall damage. The effect of blood vessel damage is a rash that may become purpuric and haemorrhagic, even ecchymotic in louse-borne typhus and R. rickettsii infection. Gangrene with loss of fingers and toes can occur in Rocky Mountain spotted fever. Leakage of protein-rich fluid from the vascular compartment into the interstitium causes hypovolaemia, hypotension and reduced tissue perfusion. Thrombotic lesions occur in a variety of organs, including the brain. Renal and liver damage from vasculitis occurs in severe cases. Clinical features Typhus group The incubation period is usually about 12 days, with a range of 10-14 days. Headache and mild fever are the usual initial symptoms, followed after 48 hours by a sharp rise in temperature to 40°C, which is sustained. Conjunctival suf-

1

306

Fig. 9.12

2

Case 9.3

3

Fig. 9.13

fusion is seen. Meningism and photophobia also occur. The rash occurs on about the 6th day and begins as red macular lesions on the axillary folds and trunk, spreading on to the limbs, with the palms, soles and face affected in the most severe cases. The rash may evolve through purpuric stages to form ecchymoses. Initial bradycardia progresses to tachycardia in hypotensive patients. Haemorrhagic manifestations and hypotension lead to renal involvement, with rising blood urea and falling urine output. Mild cases get better in the second week of the disease, but untreated severe cases may go into the third week before recovery begins. Death from shock, cerebral involvement or renal failure occurs in the second or third week. Mortality rates of 10-40% without treatment are recorded, and increasing age is associated with higher fatality rates. Brill-Zinsser disease is a mild form of louse-borne typhus caused by the recrudescence of latent infection, and in this condition recovery is the rule. Spotted fever group The incubation period is about 7 days for Rocky Mountain spotted fever (RMSF) and African tick typhus, and rather longer (around 13 days) in rickettsial pox. Fever and headache with generalized limb pains are usual early in the course. The clinical features are similar to those of typhus. There is early deterioration in those with severe RMSF. Death may occur between 3 and 6 days of onset, occurring in up to 25% of patients not receiving specific treatment in RMSF. African tick typhus is generally mild and selflimiting. Fever, systemic upset and headache are the usual presenting features. Rash may occur, causing macular lesions which have a widespread distribution on the trunk and limbs. The palms and soles may be affected. A black ulcerated eschar is usually found marking the site of the tick bite. Diagnosis The clinical features usually lead to suspicion of a rickettsial infection. This can be confirmed by serological tests using group-specific rickettsial antigens. The main differential diagnosis in RMSF is meningococcal septicaemia, which also requires urgent treatment. Laboratory features Severely affected patients with all forms of rickettsial disease will show generalized derangement of coagulation and evidence of liver and renal dysfunction. Platelet counts are low, in association with prolongation of the prothrombin time, reduced fibrmogen levels and elevated levels of fibrin degradation products. Anaemia is usual. Plasma albumin levels fall. Declining urine output indicates renal failure. Management Doxycycline (100 mg 12-hourly orally or i.v. for 7 days, or for 2 days after defervescence) is effective against all forms of rickettsial infection. Doxycycline 200 mg as a single dose is effective in epidemic typhus. Pregnant women and children under 8 years should receive chloramphenicol

(50mg/kg/day in divided doses for 7 days or for 2 days after defervescence; 25mg/kg/day in divided doses for infants under 1 month). Supportive measures Prompt administration of antibiotics is essential in RMSF and louse-borne-typhus. Additional management is essentially supportive, depending on the facilities available, and comprises restoring blood volume with infusions of plasma and fluids, and restoring coagulation mechanisms by fresh blood transfusions or fresh frozen plasma. Prevention and control Delousing is the main way to control louse-borne typhus. Thorough and regular washing of clothes in hot water with detergent kills lice and their eggs. DDT 10% or malathion 1% can be applied to clothed people with good effect. Appropriate clothing should be worn to prevent ticks reaching the skin. In tick-infested areas the skin should be carefully inspected at the end of the day. Any ticks that are found should be removed by gentle pulling to keep them intact. The duration of tick attachment, which can last 5 days, is important in relation to development of disease and strength of attachment, with 20 hours being the length of time needed to give a greater chance of infection.

Scrub typhus Scrub typhus (Orientalis tsutsugamushi) occurs throughout Asia and is probably one of the commonest causes of fever there. Studies in Malaysia showed that it accounted for 23% of patients with febrile illnesses presenting at rural hospitals. In most cases it is a self-limiting infection, with resolution in up to 14 days. Mortality has been reported in up to 60% of severe cases. It is transmitted by chiggers, the larvae of leptotrombiculid mites. Occupational exposure in rural areas (farming, animal tending, rubber plantation work etc.) and military personnel deployed in rural areas are associated with increased risk of infection. The incubation period is about 10 days (range 6-18). Fever, headache, chills and general body pains are the usual symptoms. Conjunctival suffusion, macular skin rash 1- pneumonitis and hepatosplenomegaly may occur, and cough with pulmonary infiltrate on chest radiographs is seen. An eschar at the site

of biting is often seen, with enlarged, tender regional adenitis. 2 Treatment with doxycycline is effective, with chlo- 9 ramphenicol as an alternative (see above). Resistance to these agents has been documented in northern Thailand. 3

Q fever Q fever is an acute illness caused by Coxiella burnetii; it is transmitted between animals by ticks and to humans from domestic animals by close contact, particularly with aborted placentas or with infected carcases in abattoirs or, more often, by inhalation of the endospores, which are resistant to desiccation and persist in the environment. Q fever has a range of clinical manifestations: atypical pneumonia, prolonged fever of undetermined cause, granulomatous hepatitis and culture-negative endocarditis. The main clinical features of acute infection are high fever with severe headache and myalgia, after an incubation period of about 20 days. Hepatosplenomegaly is common, with biochemical evidence of hepatitis and histologically granulomatous liver disease with very well denned granulomas that can suggest the diagnosis. Pneumonia is less common but may be severe and bilateral. Q fever may present as PUO (p. 274) with a prolonged course. Epididymo-orchitis can occur and so must be differentiated from brucellosis, which also produces prolonged fever. The natural history of the disease is to resolve spontaneously within 2 weeks in most patients, but a slower recovery with progression to chronic disease may occur. The main feature of chronic Q fever is endocarditis, which is usually fatal if left untreated (p. 574). Diagnosis Diagnosis is usually made by serology with positive responses to the phase II antigen in acute disease, IgM and IgG, and positive serology to the phase I antigen in chronic disease, with raised IgA litres suggestive of endocarditis. Laboratory handling of the organisms is dangerous. Treatment Doxycycline 100 mg twice daily for 14 days is the treatment of choice, but it is not known whether it prevents progression. Endocarditis is very difficult to eradicate by antibiotic therapy alone, and surgery is often required.

BACTERIAL INFECTIONS Despite the proliferation of antibiotics since the discovery of penicillin in 1941, bacterial infections remain a challenging problem. The continuing emergence of resistance to antibiotics, particularly but not exclusively in the hospital environment, is one aspect of the problem which is discussed earlier in this chapter (p. 264). Another is the rapid development of serious illness leading to multiorgan failure when antibiotics administered too late are relatively

ineffective. Gram-negative septicaemia is a classic example of this, and much research has been focused on the pathogenetic mechanisms involved, with recent advances defining the influence of cytokine release and highlighting the central role of nitric oxide in the vascular microenvironment. The disastrous consequences of failure to treat very early draw attention to the value of prophylactic measures, which are widely used in the form of antibiotics for surgery.

307

TABLE 9.24 Classification of bacteria

TABLE 9.25 Streptococci

Gram-positive cocci Streptococci Strep, pneumoniae Strep, pyogenes Strep, viridans group Anaerobic Staphylococci Staph. aureus Staph. epidermidis

Gram-positive bacilli Bacilli B. anthracis B. cereus Corynebacterium diphtheriae Listeria monocytogenes Clostridia (anaerobic) Cl. tetani Cl. botulinum Cl. perfringens Cl. difficile Cl. septicum

Gram-negative cocci Neisseria N. meningitidis (meningococcus) N. gonorrhoeae (gonococcus)

Gram-negative bacilli Enterobacteria Escherichia coli Klebsiella Proteus Salmonella Shigella Parvobacteria Haemophilus Bordetella Brucella Yersinia Pseudomonas P. aeruginosa Burkholderi pseudomallei Francisella tularensis Bartonella bacilliformis Cat-scratch bacillus Vibrios V. cholerae Campylobacter jejuni Anaerobic bacteria Bacteroides Fusobacteria

Mycobacteria Spirochaetes Treponema T. pallidum T. pertenue T. carateum Borrelia Leptospira icterohaemorrhagiae

Vaccines have been developed against some bacteria and this will be an area of rapid development during the lifetime of this edition of the textbook. A classification of bacteria is given in Table 9.24.

GRAM-POSITIVE COCCI

STREPTOCOCCI

308

B-Haemolytic streptococci are classified by the Lancefield grouping system, which detects polysaccharide antigens on their surface by a panel of antisera. Table 9.25 shows these

Group and name Group A S. pyogenes

Group B S. agalactiae Group D Enterococci, e.g. S. faecalis S. pneumoniae (pneumococcus) S. viridans group, e.g. S. milleri, mitior, mutans, salivarius, sanguis

Characteristics

Disease/syndromes

p-haemolytic Secrete enzymes, e.g. hyaluronidase, and toxins e.g. erythrogenic

Pharyngitis Scarlet fever Otitis media Sinusitis Erysipelas/impetigo Lymphangitis Puerperal sepsis

p-haemolytic Common vaginal commensal

Neonatal infection (puerperal sepsis)

B- or non-haemolytic Anaerobic Normal bowel flora Antibiotic resistance oc-haemolytic Common throat commensal More than 80 serotypes a- or non-haemolytic Dental commensals

Urinary tract infection Septicaemia Endocarditis Pneumonia Acute meningitis Otitis media/mastoiditis Endocarditis Cerebral abscess and empyema (S. milleri)

and other streptococci which are a-haemolytic or nonhaemolytic. Not all group D streptococci are P-haemolytic. Cutaneous streptococcal infections are described in Chapter 10.

Scarlet fever Scarlet fever is a febrile illness with a characteristic rash caused by infection with an erythrogenic toxin-producing strain of Strep, pyogenes. Infection is usually in the throat, but the skin or genital tract may be the primary site. It is now a much milder disease than it was 50 years ago. Its severity is probably determined by variations in the microorganism, but host factors also play a part. The same bacterium may be a harmless passenger in one subject, or produce a sore throat or scarlet fever in another. Host antibodies are largely responsible for these differences. Antibodies to the erythrogenic toxin prevent the rash of scarlet fever but not tonsillitis. Clinical features Infection is spread by droplets. After an incubation period of 2-4 days there is an abrupt onset of sore throat, pain on swallowing and fever. There is tender cervical lymphadenopathy. The tonsils are enlarged, congested, and often covered with exudate. The tongue is coated and initially red papillae are visible, but later the fur is lost, leaving the classic 'strawberry tongue' appearance. The rash (Fig. 9.15) appears on the face within 2 days. It is a diffuse erythema, often with perioral pallor (not specific for this con-

TABLE 9.26 Invasive disease due to Staph. aureus Site of infection

Disease

Skin

Impetigo Boils and carbuncles Surgical wound infections Septicaemia Pneumonia with abscesses Abscess Acute endocarditis

Blood Lungs Liver, bone and other Heart

9

Diagnosis and management Strep, pyogenes can readily be cultured from a throat swab, and the presence of neutrophilia helps to distinguish the condition from infectious mononucleosis, which can cause an identical appearance. Retrospective diagnosis can be made by measuring antistreptolysin O antibodies. Treatment with oral penicillin is commonly used, but may not eliminate streptococci unless given in high dose for 10 days. Eradicating the organism is less important since complications have become so uncommon. Erythromycin is an alternative agent for penicillin-allergic patients. FIG. 9.15 Scarlet fever rash

dition). On the trunk it is a punctate erythema which is less brightly coloured. Superficial peeling of the skin starts on about the 4th day. The illness resolves within a week, although the patient may continue to feel unwell for several more days. Complications Complications are uncommon in the temperate world, perhaps because of prompt treatment with antibiotics, but post-streptococcal glomerulonephritis is well recognized in the tropics and sub-tropics where infected insect bites, infected scabies etc. are common. Soft tissue infection due to this organism can be serious and life-threatening in the context of necrotizing fasciitis. This may occur without any obvious site of entry for the organism, but the process begins with localized pain in the muscle of the affected area. There is spreading cellulitis accompanied by severe haemodynamic disturbance with shock. Nausea, vomiting and diarrhoea may be present early in the course. Progressive increase in the severity of pain is a clinical indicator of this condition. Bullae can develop over affected skin. Outcomes differ but include tissue destruction, severe systemic illness, loss of the affected limb as a result of gangrene, and sometimes death. Prompt treatment with intravenous antibiotics should include clindamycin. Septic complications include otitis media. Late complications are rheumatic fever (p. 1069) or acute glomerulonephritis, which may develop 2-3 weeks after the infection.

STAPHYLOCOCCI

Staph. aureus and Staph. epidermidis are the clinically important staphylococci. Both are common skin commensals, but Staph. aureus is more virulent and is commonly associated with disease.

Staphylococcus aureus infection In most conditions where host susceptibility to infection is increased, Staph. aureus is an important pathogen. This applies particularly to the skin after surgical or accidental trauma and burns, and after skin infection such as chickenpox. More general predisposition is seen in diabetes mellitus, uraemia, alcohol excess, malnutrition and malignancy. When a surgical foreign body is present, for example an intravenous catheter, a heart valve or a prosthetic joint, both Staph. aureus and epidermidis can take root, with potentially disastrous consequences. There are two distinct mechanisms by which Staph. aureus can cause disease. Most often it is by direct invasion of tissues (Table 9.26), when some of the enzymes produced by Staph. aureus may be important in pathogenesis, e.g. catalase,coagulase,hyaluronidase. Less often, the major disease manifestations are the result of toxin secretion (Table 9.27). Management of deep Staph. aureus infections When Staph. aureus localizes in deep tissues it quickly causes acute inflammation and necrosis, leading to abscess formation. If an abscess can be seen on scanning (e.g. ultrasound, isotope or CT) it may be possible to aspirate pus as

309

TABLE 9.27 Staph. aureus toxin-mediated diseases Disease

Manifestations

Organism

Pathological effect

Site of infection

Staphylococcal scalded skin syndrome

Skin erythema; may progress to bullae, exudate, crusting, positive Nikolsky sign

S. aureus phage group II, often phage type 71 (produce epidermolytic exotoxins)

Split skin at stratum granulosum

Skin

Toxic shock syndrome

Diarrhoea and vomiting Hypovolaemia Erythema of mucosae and skin leading to peripheral desquamation Fever and myalgia

5. aureus phage group I, often phage types 29 and 52 (produce enterotoxin F, identical to exotoxin C)

Uncertain

Vagina, skin or other (occurs mainly in young women using tampons)

Food poisoning

Vomiting and diarrhoea

S. aureus phage group III (produce enteroxins A, B, C, D, E)

Uncertain

Food (toxin ingested)

both a diagnostic and a therapeutic measure. Antibiotic treatment is more likely to be successful if this can be achieved, and it may substitute for open surgery. In some sites, such as a heart valve, early surgery is essential because the valve is destroyed and a new one cannot be inserted if the surrounding valve root tissue has been destroyed. The decision whether to operate or aspirate and give antibiotics, or to give antibiotics alone, for a deep Staphylococcal infection is determined by the amount of pus formation, the accessibility of the site and the delicacy of adjacent tissue. For example, drainage of some CNS abscesses may lead to unacceptable damage, in which case the neurosurgeon would opt to try to sterilize the lesion with antibiotics in the first instance. Staph. aureus often produces pMactamase, particularly in hospital-acquired infections, and so treatment is usually started with a B-lactamase-resistant antibiotic such as flucloxacillin. Methicillin-resistant Staph. aureus (MRSA) is less pathogenic but it can cause deep-seated infection in vulnerable patients. The cutaneous syndromes of Staphylococcal infections are described on page 384. The management of septicaemia is discussed on page 274. Lung abscess is described on page 647.

Staphylococcus epidermidis infection Staph. epidermidis is a skin commensal. It is now clear that it causes serious infections in patients who are immunocompromised and in patients with intravenous cannulae, prosthetic heart valves, pacemaker wires, CSF shunts and other foreign materials in place for long periods. Staph. epidermidis produces B-lactamase and is usually resistant to a wide range of antibiotics, including methicillin. Resistance to methicillin is mediated by alteration to penicillin-binding proteins. The mainstay of treatment is vancomycin.

1

310

Fig. 9.14

2

MCQ 9.16

FIG. 9.16 Necrotic haemorrhagic lesions in meningococcal septicaemia, also showing the unusual features of gangrenous terminal phalanges

GRAM-NEGATIVE COCCI NEISSERIA Neisseria gonorrhoeae (gonococcus) infection is described in Chapter 11 (p. 458).

Neisseria meningitidis (meningococcus) infection There are three types of meningococcus - groups A, B and C - which are the most common cause of acute bacterial meningitis (p. 1413). They also cause septicaemia with or without meningitis (Fig. 9.16). Disease occurs sporadically or in epidemics, and the epidemiology is poorly understood.

FIG. 9.17 Incidence of meningococcal disease, England and Wales, 1989-2001 (source PHLS Meningococcal Reference Unit)

Meningococcal disease affects predominantly young children, although it can occur at any age. Its epidemiology varies between different countries, and particularly between Africa and European countries or the USA. In the meningitis belt of Africa it is a disease of young schoolaged children caused by serogroup A, which occurs in major epidemics at intervals of 8-10 years during the dry season, affecting 500/100000 people. In the UK, disease caused by serogroups B and, increasingly, C is endemic (Fig. 9.17), affecting 1-3/100000, with sporadic increases in incidence (hyperendemicity) in small areas from time to time, mainly during the winter months. About 50% of cases are in the under-5 age group. Clinical features Meningitis, which is the most common form of disease, is described on page 1415. It can occur with or without meningococcal bacteraemia, but it is important to recognize that meningococcaemia also occurs without meningitis in nearly 10% of cases. This can cause a range of illness, from almost asymptomatic bacteraemia to fulminant septicaemia with disseminated intravascular coagulation. The central feature is rash on any part of the body or mucous membranes which varies from maculopapular to petechial to ecchymotic. 1 In chronic meningococcal septicaemia there is persistent bacteraemia, together with rash, arthritis and low-grade fever. This may last for some weeks before progressing to more severe disease. Other syndromes rarely associated with meningococcal infection are pneumonia, pharyngitis, pericarditis, osteomyelitis and endophthalmitis. Diagnosis is made by blood cultures or Gram stain and culture of aspirates from skin lesions. Cultures from other sites may be indicated in particular situations, e.g. joint aspirate in arthritis.

Management and prevention Benzyl penicillin is highly bactericidal for meningococci and it should be the first choice of antibiotic. It is given intravenously at a dose of 2.4g 4-hourly initially in very sick patients, reducing to 1.2g after about 24 hours and then to 6-hourly (adults). If there is good evidence that the patient has penicillin sensitivity, cefotaxime or ceftriaxone can be used. Treatment is continued for 5-7 days. Nasopharyngeal carriage is not terminated by penicillin treatment, which should therefore be followed by rifampicin 600 mg b.d. for 2 days. This may need to change as resistance to rifampicin emerges, and an alternative would be a single dose of ciprofloxacin 500 mg. Polysaccharide vaccines are available against serogroups A and C meningococci but not against group B and there is a conjungate vaccine against serogroup C. The polysaccharide vaccines provide only short-lived protection but they are useful in curtailing the kind of limited outbreaks seen in European countries and the USA. Mass vaccination of villagers in Africa ahead of the spread of epidemics has been attempted, but it is more difficult to achieve success. 0

9

GRAM-POSITIVE BACILLI

BACILLI

Bacillus anthracis: anthrax Few cases of anthrax are seen in western countries but about 100000 cases occur in the world each year. The Grampositive rods of B. anthracis give rise to endospores in conditions unfavourable for replication, and these can survive for many years in the soil. Making use of this property they have been used in warfare in both world wars during the 20th century, and as a terrorist weapon on other occasions. The spores can contaminate the hides, hooves and bones of herded animals, and contact with these can cause cutaneous anthrax both in the endemic areas and in industrialized countries to which these animal materials are taken for processing. Bacilli are found in the tissues of animals dying of systemic infections. Transmission to humans is by skin contact with infected material such as hides or meat, by inhalation of spores or by ingestion of infected meat, causing skin, respiratory and gastrointestinal disease, respectively. Disease onset is within 5 days in all forms, and is usually about 2 days. Pathogenesis depends on the expression of virulence factors coded by two plasmids. Oedema toxin, lethal factor and protective antigen are the products of these plasmids. Endospores ingested by macrophages become vegetative and rupture macrophages, releasing organisms that replicate further in lymphatics and in the bloodstream, producing massive septicaemia, up to 108/mL. Expression of virulence factors is increased by elevated host temperature.

311

Cutaneous anthrax Cutaneous anthrax is the most common form of the disease and the most benign. A red macule or papule develops at the contact site, e.g. the neck of a meat carrier in a slaughterhouse. A pruritic vesicle develops, satellite lesions appear, and the whole evolves into an ulcer, then a black eschar. There is usually some surrounding oedema, but occasionally there is a severe local reaction with gross oedema, and general toxaemia ensures. Despite the size and appearances of the lesion it is painless unless secondarily infected. Inhalation anthrax At first there are mild general symptoms and a nonproductive cough. Improvement over a few days is followed by abrupt decline. The patient rapidly becomes very ill, with cough, fever, tachypnoea, cyanosis, tachycardia, vomiting and chills. Diffuse crackles are heard over the affected lung fields. The spleen may enlarge. There is massive enlargement and oedema of the neck and chest wall. Untreated, the patient often dies within 24 hours. There is an 80% mortality in this form. Gastrointestinal anthrax This may occur in clusters if a family or tribal group eat contaminated meat. Following ingestion, lesions can occur anywhere in the gastrointestinal tract, including the throat. There is diffuse abdominal pain, with rebound tenderness. Ascites can develop, and with this the abdominal pain declines. There are varying degrees of gastroenteritis with haemorrhage, and mortality is high. Symptoms resolve after about 14 days in those surviving. Haematogenous dissemination of organisms can sometimes cause meningitis with bloody CSF. Diagnosis Many cases probably go undiagnosed in the tropics and subtropics. The cutaneous form is probably most often diagnosed on the clinical appearances. Occupational exposure may be suspected from the nature of the patient's job. The severity of a pneumonic illness may prompt consideration of pulmonary anthrax. The organisms are readily found in Gram-stained smears from skin lesions and grow well in blood culture samples.

312

Management The cutaneous form may resolve without treatment. Penicillin V, 2 g/day in divided doses, is very effective. For more severe infections benzyl penicillin, 4.8-7.2 g in divided doses intravenously, is used. When the patient is allergic to penicillin, ciprofloxacin 400 mg twice daily intravenously or 500 mg twice daily by mouth, tetracycline 2 g/day or chloramphenicol lOOmg/kg/day in divided doses orally or intravenously can be used. The latter is given to children and pregnant women. Pulmonary or intestinal cases will require supportive care, with maintenance of oxygenation, giving normal saline, plasma or blood transfusions as necessary to maintain blood volume and hydration. Prevention is by elimination of the infection from

animals. Immunization is available when there is danger of occupational exposure. Vaccines are available for those who may be exposed to anthrax.

Bacillus cereus infection B. cereus is one of the causes of food poisoning (see Gastroenteritis, p. 278). Two illnesses are associated with different incubation periods. A short incubation period (1-6 hours), vomiting and abdominal pain caused by heatstable enterotoxin are often related to the ingestion of reheated rice. Longer incubation (8-16 hours), diarrhoea and abdominal pain are caused by a heat-stable toxin formed in the infected intestine. The treatment is directed at maintaining the patient's hydration. There is no specific therapy.

CORYNEBACTERIUM DIPHTHERIAE INFECTION

Corynebacterium diphtheriae is the cause of diphtheria, a toxin-mediated disease which is rare in countries with an effective immunization programme. Transmission is mainly person to person by nasopharyngeal droplets, but skin infection is also important in tropical countries. Pathogenesis C. diphtheriae (korynee - club, refers to the shape of the bacteria; diphtheria = leather, refers to the pharyngeal membrane) infects the nose, throat, larynx and trachea or the skin. It causes local inflammation, but its major effects result from the production of a protein toxin of MW 62000 which inhibits ribosomal polypeptide chain production by inhibiting elongation factor 2. The toxin is elaborated only by C. diphtheriae strains infected by a bacteriophage carrying the toxin production gene. Its main effects are on the myocardium and nervous tissue. Clinical features After an incubation period of 2-6 days there is a slow onset of malaise and fever, sometimes with mild sore throat. Immunized subjects develop disease at the site of infection but no toxin-mediated effects. In the unimmunized, at an early stage the predominant findings are profound weakness, restlessness and irritability, with fever, pallor and a rapid thready pulse. The pharynx shows the characteristic membrane which is adherent; attempts to scrape it off result in bleeding. In laryngeal diphtheria the membrane involves the larynx and trachea, causing respiratory obstruction which may require tracheostomy. The primary infection may also involve the nose, larynx or bronchi. ECG abnormalities suggesting myocarditis (flattening or inversion of T waves) or first-degree heart block appear after the first week and can progress to major dysrhythmias and circulatory collapse. Neurological complications start later, with paralysis of the palate, followed by the ocular muscles, then the pharynx, larynx and respiratory muscles, and lastly the limbs.

Diagnosis C. diphtheriae can be isolated from sites of infection but demonstration of toxin production takes several more days - too long to delay management decisions. Polymerase chain reaction for the tox+ gene has revolutionized management, as it can demonstrate the presence of the bacteriophage on the same day that a positive culture is obtained. Management and outcome Diphtheria antitoxin is administered without waiting for bacteriological confirmation in suspected cases, but it is no longer used for prophylaxis because of the danger of hypersensitivity reactions. The treatment dose is 1000030000 units i.m. for mild disease, increasing to as much as 100000 units for severe cases of more than 3 days' duration. For doses greater than 30000 units the remainder is given intravenously after an interval of 0.5-2 hours. Penicillin or erythromycin will eradicate the organism from the primary infection site but the patient must be isolated until the risk of cross-infection has been removed. Intensive nursing is required to protect the airway, and tracheostomy may be needed. ECG monitoring is used to detect dysrhythmias, and physiotherapy to preserve the range of movement in paralysed limbs. Contacts must be traced, treated if infected, and immunized. If the patient survives the acute illness, recovery from complications such as paralysis is complete. LISTERIA MONOCYTOGENES INFECTION

Listeria monocytogenes is a Gram-positive bacillus which is commonly dismissed as a diphtheroid contaminant on examination of CSF specimens. Microbiologists are now aware of its significance as a cause of serious disease, particularly in immunocompromised patients, pregnant women and neonates. The bacillus is a common contaminant of soft cheeses, especially if these are unpasteurized. It causes septicaemia, often associated with meningoencephalitis. Ampicillin or penicillin is the drug of first choice, with erythromycin in reserve for penicillin-allergic patients.

Epidemiology and pathogenesis Cl. tetani is found in human and animal faeces and its spores abound in soil. It is anaerobic and proliferates in ischaemic or necrotic tissue, usually in a poorly tended wound. In rural areas farm workers or gardeners are infected, but in cities the disease has been observed in drug addicts. Because immunization against tetanus toxin is standard in western countries, tetanus is rare. In poorer countries immunization is variable. Tetanospasmin is a neurotoxin which acts in several different ways. It inhibits the release of acetylcholine from nerve endings in muscle, but more importantly it interferes with synaptic reflexes in the spinal cord, causing disinhibition which results in muscle spasms. Clinical features The incubation period can be up to 3 weeks, depending on inoculation dose and site. Disease is most severe in the completely non-immune, in drug addicts and at extremes of age. The outcome is worst when the incubation period is short and there is rapid onset of the first spasm. The first (and sometimes the sole) sign of tetanus is often muscle spasm local to the wound; in most the CNS is involved, and trismus (lockjaw) heralds the onset of generalized tetanus. The patient notices difficulty in opening the mouth and swallowing, followed by stiffness in the neck, and back. Examination reveals the initial site of infection and 'risus sardonicus', the facial expression caused by trismus and contraction of facial muscles. The muscles of the back and abdomen become stiff, and painful spasms, provoked by sudden movement or noise, occur in the back muscles, causing opisthotonos. Autonomic involvement causes sweating and cardiovascular instability. Death is from exhaustion, aspiration or secondary infection.

Tetanus

Management Further toxin production is prevented by local debridement of wounds and treatment with metronidazole 500 mg intravenously 6-hourly. Human immunoglobulin containing antibodies to tetanus toxin in high titre is given to neutralize free toxin; 500 units are administered i.m. If surgery is indicated, it should be delayed for an hour after the administration of antibody. The patient is nursed in a quiet room to avoid stimuli likely to provoke spasms. Benzodiazepines such as diazepam are used to control spasms, as well as for their anxiolytic and amnesic effects. Neuromuscular block may be required, and then the patient is ventilated (see p. 737). Because prolonged ventilation is usual, tracheostomy is performed early. Nutrition is maintained enterally or parenterally. An attack of tetanus does not render the patient immune and a full course of immunization should be given after recovery.

Tetanus is a potentially fatal disease dominated by muscle spasm caused by an exotoxin, tetanospasmin, produced by Clostridium tetani.

Prevention Immunization with tetanus toxoid is the only way to

CLOSTRIDIA Clostridia are a group of obligate anaerobic bacteria that cause illness largely through toxins which act locally or enter the circulation. Clostridia are frequent wound contaminants without necessarily causing disease; when disease occurs, however, it is severe.

9

313

prevent the disease altogether. A full course of toxoid confers immunity for at least 5 years. Neonatal tetanus can be prevented by immunizing the mother during pregnancy. Patients with wounds at risk from Cl. tetani infection are managed according to their state of immunity. All patients require local wound toilet. Fully immunized patients receive a booster dose of toxoid. Unimmunized patients with a wound considered susceptible to Cl. tetani receive human tetanus immunoglobulin, and active immunization with adsorbed tetanus toxoid is started simultaneously. Injection of long-acting penicillin, followed by a course of oral penicillin, is also indicated in these cases.

Botulism Botulism is a virulent form of food poisoning caused by heat-labile neurotoxins of Cl. botulinum. Disease follows ingestion of food containing toxin (usually tins of meat), ingestion of bacterial spores which germinate and produce toxin in the intestine, or contamination of a wound by spores which germinate there. The highly potent toxins interrupt cholinergic neurotransmission. There are seven antigenic types, A-G, but only A, B, E and F cause human disease. They consist of two polypeptide chains linked by a disulphide bond; the heavy chain binds to presynaptic receptors and the toxin is endocytosed. It inhibits acetylcholine release and the cell cannot release transmitters, so that there is motor paralysis and autonomic dysfunction. Damage to the synapse is permanent and repair requires the formation of a new synapse. These long-term effects of botulinum toxin are being used in the treatment of conditions in which there is sustained, inappropriate muscle contraction, such as torticollis. In the case of food-borne botulism the initial illness is an afebrile gastroenteritis that develops within 12-36 hours of ingestion in one-third of patients, but the remainder have only malaise and dizziness (with postural hypotension), progressing to extreme dryness caused by autonomic paralysis. Other neurological symptoms appear during the next 3 days: • Diplopia followed by bulbar signs • Symmetrical weakness of the limbs • Bulbar and respiratory muscle weakness may lead to inhalation pneumonia. Treatment is essentially supportive, with maintenance of the airway, including incubation or tracheostomy as required. Autonomic dysfunction does not usually cause severe problems (cf. tetanus). Antitoxin therapy is normally given although its value has not been established in controlled trials. Hypersensitivity reactions are common,

1 314

MCQ 9.17

particularly in those with a history of atopy, and skin testing is recommended. Trivalent horse antitoxin active against types A, B and E is given intravenously, 20 mL by slow infusion, followed by lOmL 2-4 hours later and further doses at 12-24-hour intervals if necessary. Infants who have ingested spores do not have circulating toxin, and antitoxin is of no benefit. Antibiotic therapy with penicillin or metronidazole is also of uncertain value, but wound debridement should be carried out in wound botulism. Recovery of muscle strength is extremely slow, occurring mainly in the first 3 months but continuing sometimes for up to a year. Prophylaxis with antitoxin may be needed in subjects thought to have ingested toxin. 1

Other clostridial syndromes Gas gangrene Cl. perfringens (welchii) is one of a number of gasproducing clostridia that contaminate wounds containing necrotic, anoxic material. Cl. perfringens produces 12 toxins, alphatoxin being the one associated with gas gangrene. It proliferates rapidly, producing myonecrosis. Gas infiltrates neighbouring tissues, causing crepitus. Toxins entering the circulation cause profound toxaemia, with a high mortality. Skin and soft tissue infections Local infections, often involving muscle, occur in amputated stumps, diabetic ulcers of the foot, decubitus ulcers, perirectal abscesses and abscesses at injection sites in drug addicts. Such infections frequently contain multiple organisms, including clostridia. Gut infections Pseudomembranous colitis is a diarrhoeal disease caused by toxins from strains of Cl. difficile which colonize the large intestine after antibiotic therapy. Cl. difficile produces two toxins, an enterotoxin (A) and a cytotoxin (B), both of which probably contribute to the pathogenesis of disease. Although the association with clindamycin therapy is well recognized, most antibiotics have caused this disease. It is particularly likely to occur in debilitated and elderly patients, but can affect any age group. It can be transmitted nosocomially, so patients should be isolated. The onset of watery diarrhoea (without blood), abdominal pain and fever is usually within a few days of exposure to an antibiotic, although it is delayed by 4-6 weeks in some cases. There may be severe systemic toxicity. In the majority of cases sigmoidoscopy reveals an inflamed mucosa covered by a white membrane, but this is absent in the 20% where only the ascending colon is involved. The diagnosis is confirmed by culture of the organism and identification of toxin in stool specimens, but as about 5% of healthy people carry the organism some laboratories just look for toxin. If symptoms continue after the offending antibiotic has been stopped, treatment is with oral metronidazole 400 mg

9

TABLE 9.28 Syndromes of enteric infection Organism

Toxins

Mechanism of action

Syndromes

ETEC

Heat labile Heat stable

Cholera-like; cyclic-AMP mediated Activates guanylate cyclase; cyclic GMP increases

EHEC (0157:H)

Shigella-like

?; toxic for Vero cells

Cholera-like Travellers' diarrhoea Infant gastroenteritis Haemorrhagic colitis Haemolytic-uraemic

t.d.s. for 7 days. Oral vancomycin has a slightly higher cure rate but is much more expensive than metronidazole. For severely ill patients who cannot take oral antibiotics, intravenous metronidazole 500 mg is given every 8 hours initially. Suppurative intra-abdominal infection caused by clostridia is particularly associated with carcinoma of the colon or pancreas. Neutropenic patients may develop severe enterocolitis caused by Cl. septicum.

GRAM-NEGATIVE BACILLI ENTEROBACTERIA The enterobacteria group comprises Escherichia coli (E. coli), Klebsiella, Proteus, Salmonella and Shigella. E. coli causes a wide variety of diseases, usually as a result of autoinfection. Urinary tract infection is the most common, but systemic infection is a major problem in hospitals. Septicaemia arises from the bowel when there is a bowel disease or when surgery has been carried out, or from the urinary tract as a complication of infection or surgery there. In neonates, E. coli is a common cause of septicaemia and meningitis. Several strains of E. coli cause gastroenteritis by a number of mechanisms; they are referred to as enteropathogenic (EPEC) (or alternatively enteroadherent - EAEC), enteroinvasive (EIEC), enterohaemorrhagic (EHEC) and enterotoxigenic (ETEC). EHEC and ETEC produce toxins (see Table 9.28). Klebsiella causes urinary tract infection, septicaemia and pneumonia. Kl. pneumoniae causes an acute severe cavitating pneumonia, most frequently seen in alcoholics (p. 638). Proteus is mainly associated with urinary tract infection. Salmonella is found in several hundred strains. Salmonellae other than S. typhi and S. paratyphi are amongst the most common causes of gastroenteritis related to food poisoning (see Gastroenteritis, p. 278). They cause disease by direct invasion of the intestinal mucosa, and some strains, e.g. S. typhimurium, cause septicaemia, but usually disease remains localized. Typhoid (enteric fever) is caused by S. typhi, and is described below. Four strains of Shigella are the cause of bacillary dysen-

tery, a form of acute diarrhoeal disease characterized by severe, bloody, purulent diarrhoea and abdominal pain.

Shigella dysentery Shigella sonnei and S. flexneri are the common organisms in the UK, the incidence of S. flexneri infection having increased in recent years; the other species are S. dysenteriae and S. boydii. Shigella infections occur worldwide, and epidemics are common where conditions are crowded and hygiene poor. Faecal-oral transmission results from lack of handwashing and the organism is highly infectious, ingestion of fewer than 200 organisms causing disease in healthy adults. The infection localizes in the colon, where the epithelium is invaded and becomes inflamed. There is an acute inflammatory exudate which produces diarrhoea containing large numbers of leukocytes. Enterotoxin production contributes to the intestinal damage, and the organisms may also produce a neurotoxin. Clinical features After an incubation period of 2-3 days there is a rapid onset of vomiting, abdominal colic and diarrhoea, which is usually bloody and purulent. Severity ranges from mild to fulminant. The most severe illness is caused by S. dysenteriae. Although the main illness is usually over within 1-2 weeks, diarrhoea and abdominal discomfort may persist for several weeks. Fever is common in the first few days and dehydration can occur rapidly, especially in children. Complications include haemolytic-uraemic syndrome (HUS), seen in children and the elderly, with a leukaemoid reaction in the blood frequently preceding the onset. Post-infective aseptic arthritis and Reiter's syndrome (conjunctivitis, urethritis and arthritis) are complications that vary in frequency in different epidemics but which are more common in HLA-B27 positive individuals. The diagnosis is made when organisms are cultured from faeces. Management Patients are managed as for any severe form of gastroenteritis, rehydration being the key factor (see Gastroenteritis, p. 278). This can usually be achieved using oral fluids, but if the patient is severely ill, and particularly if vomiting cannot be controlled, intravenous rehydration

315

may be required. Antibiotics are given as they shorten the duration of and ameliorate symptoms. They also reduce the risk of transmission, making a case for their use in milder disease, but plasmid-mediated resistance is common and transfers rapidly, so it is important to know the local antibiotic sensitivity pattern. Currently the first choice of antibiotic is ciprofloxacin 500 mg b.d. orally for 5 days. Other possibilities are co-trimoxazole, ampicillin or tetracycline. In the acute stage of invasive infection inhibitors of intestinal motility such as codeine phosphate and loperamide may exacerbate symptoms, and even precipitate toxic dilatation of the colon. Pain management can therefore be difficult, but regular paracetamol may help. Codeine is much more effective but it should be used only with great caution.

Typhoid Typhoid (enteric fever) is a serious, prolonged systemic infection with the Gram-negative bacillus S. typhi. Epidemiology Typhoid is endemic in many poor countries, and although uncommon in Britain (about 200 cases reported per year) it is an important cause of fever in returning travellers. Humans are the reservoir of infection and spread of infection is by contamination of water or food by human faeces. Epidemics usually occur when a common water supply is contaminated. Small outbreaks sometimes results when a chronic excreter of the organism works as a food handler. Pathogenesis Ingested organisms invade the small intestinal mucosa. In the absence of specific immunity, organisms phagocytosed by macrophages survive and proliferate. After an incubation period varying from 1 to 3 weeks (usually 10-14 days), determined by the infecting dose, septicaemia develops and foci of infection may be established in any organ. The gallbladder is an important site, as it is the source of chronic excretion when this occurs. Necrotic inflammation of Peyer's patches may be so vigorous as to cause intestinal perforation or haemorrhage. Clinical features In Britain typhoid usually presents as a pyrexia of insidious onset in a patient returning from abroad, often from the Indian subcontinent, although there have been small outbreaks in the United Kingdom related to take-away food outlets with a chronic carrier as the source. The classic stepwise rise in temperature is rarely seen but the untreated illness has a remorseless progression. Headache and dry cough are common in the first week, and

1

316

MCQ 9.18

the patient is often vague and withdrawn. Constipation is usual. In the second week the spleen is often palpable, and the elusive rash consisting of rose spots is seen in less than half of patients. The rash is sparse - fewer than a dozen spots in all - and consists of 1-2 mm pale red macules, which are hard to discern on a pigmented skin. By the third week diarrhoea may develop and the patient, now extremely ill, may lapse into coma. The overall mortality in untreated typhoid is about 10%, but with appropriate antibiotic therapy it is less than 5%. Spontaneous remission is often followed by relapse within 2 weeks. An unusual form of typhoid is encountered in patients infected with schistosomiasis, caused by S. mansoni or haematobium, in which fever is less dramatic but the illness more prolonged. Typhoid in children Typhoid is even more featureless in children than in adults, and presents as an abrupt severe septicaemic illness. The mental state is striking, the child initially being irritable but later frankly delirious. Febrile convulsions may be the presenting feature, and some degree of meningism is not uncommon. Diarrhoea is perhaps more frequent than in adults with typhoid, although intestinal infection with other organisms complicates the issue. Complications The most serious complications in the acute phase are intestinal haemorrhage or perforation. In severe cases there may be myocarditis or bone marrow suppression. Later complications are relapse, even after correct antibiotic treatment, in 15% of cases, and local sepsis in any organ as a result of septicaemia. Osteomyelitis is slow to cause symptoms. Chronic carriage Faecal excretion of S. typhi often continues several weeks after resolution of the illness and, rarely, patients excrete the organism for years. Chronic carriage is defined as excretion of the organism for more than 6 months after an acute episode. The organism is not detected in faeces during antibiotic treatment but reappears after therapy is stopped. The source of organisms in long-term carriers is probably the gallbladder, and gallstones readily become infected. Chronic excretion has obvious epidemiological consequences. Diagnosis The diagnosis is usually made by blood culture. The organism is not present in stools initially but can be isolated in

SUMMARY 10 Complications of typhoid • Relapse • Local sepsis, e.g. lobar pneumonia, meningitis, acute cholecystitis, urinary tract infection, osteomyelitis • Intestinal perforation and haemorrhage (2-3% with antibiotic therapy)

the second to third weeks. Bone marrow culture is particularly useful for identifying S. typhi in patients who have received antibiotics. It may also be cultured from urine in 30% of patients. Serology is rarely helpful in diagnosis. The blood count is helpful as neutropenia is common. Managemenf Patients must be nursed in isolation, preferably in an isolation unit, even though person-to-person transmission is uncommon. Intravenous fluids are given to correct dehydration. Currently ciprofloxacin 500 mg b.d. for 10 days is the drug of first choice. Treatment is started when there is a strong suspicion of the diagnosis and appropriate samples have been taken for culture without waiting for the results. Although duration and severity of symptoms, as well as mortality, are reduced by treatment, the response to treatment is not immediate. Controlled trials have shown the efficacy of dexamethasone 6 mg 6-hourly given to very toxic patients with typhoid. Relapse can be treated with the original antibiotic provided the organism was sensitive. Stools are collected for culture twice weekly after treatment to detect prolonged carriage. This is difficult to treat, but success has been reported with high-dose amoxicillin and ciprofloxacin for 4-6 weeks. Cholecystectomy may be necessary. Three negative stool specimens are considered to indicate that bacteriological clearance has been achieved. Perforation of the intestine is dangerous and carries a high mortality. Good results have been achieved by resuscitation followed by laparotomy, oversewing of single perforations or resection when a length of intestine is involved, and lavage of the peritoneum. When typhoid occurs in a patient with concurrent schistosomiasis the latter must also be treated to allow the antibiotic to be effective. Prevention A monovalent whole-cell heat-killed vaccine gives 75% protection against S. typhi for 1 year but causes local pain, fever and malaise in a high proportion of recipients after both primary vaccination and booster doses. This vaccine is no longer available in the UK. Another parenteral preparation containing Vi polysaccharide antigen is effective for 3 years. A live oral vaccine which uses the mutant strain of S. typhi,Ty 21a, gives similar protection after three doses but may not be as long-lasting. The protection may be overcome if the infecting dose of organisms is high enough, emphasizing the importance of clean water and thorough cooking in prevention.

Haemophilus infection

Haemophilus organisms are normal colonizers of the upper respiratory tract. H. influenzae is a major human pathogen; other Haemophilus species rarely cause disease, although H. parainfluenzae can cause illness similar to H. influenzae. A small proportion of H. influenzae isolated from humans are encapsulated and six antigenic types are recognized. The most important is type b, which can cause septicaemia and several clinical syndromes in children under 5 years old (Table 9.29). Older children acquire immunity to the capsular antigens. Untypable (nonencapsulated) H. influenzae are potential pathogens in the lower respiratory tract. Strains of H. influenzae have become increasingly resistant to antibiotics in recent years: 5-15% of strains are resistant to amoxicillin. This is particularly relevant to children with life-threatening diseases such as acute meningitis, in which a delay in treatment with the right antibiotic could be fatal. Resistance is mediated by pMactamases. Vaccine containing the polysaccharide capsular antigens has now been introduced in the UK for children below the age of 4 years, and this has greatly reduced the incidence of H. influenzae disease. Whooping cough (pertussis)

Whooping cough is a prolonged lower respiratory tract infection caused by the Gram-negative coccobacillus Bordetella pertussis. It is highly infectious and is transmitted by droplet. Epidemiology Spread of infection is entirely human-to-human and the disease is endemic, cases occurring continuously. Epidemics in susceptible children occur every 4 years in the UK (Fig. 9.18). Since reports of a possible association between pertussis immunization and encephalopathy in

TABLE 9.29 Diseases caused by H. influenzae Type

Disease

Type b

Acute meningitis Epiglottitis Septic arthritis Orbital cellulitis Pneumonia Exacerbations of chronic bronchitis Otitis media and sinusitis Conjunctivitis Rarely pathogenic

Salmonella paratyphi infection

S. paratyphi B is the most common paratyphoid infection in Britain. Other strains are called S. paratyphi A and C. Paratyphoid has a shorter incubation period than typhoid and is a milder disease with fewer complications. Diarrhoea tends to occur early, in contrast to typhoid. Treatment is with ciprofloxacin. 1

9

PARVOBACTERIA

Non-encapsulated

Other encapsulated types

Age

Up to 5 years

Any age

317

prolonged coughing spasms may cause convulsions. Minor complications are subconjunctival haemorrhages, periorbital bruising and frenal ulcer, all of which result from prolonged spasms of coughing.

FIG. 9.18 Incidence of whooping cough in England and Wales, 1940-2000 Note the effect of immunization on incidence. The death rate was falling before immunization was introduced (Communicable Diseases Surveillance Centre, Colindale).

1973, the rate of immunization has fallen off and epidemics have been larger. The disease may affect any age group but is most common in children under 5 years and most severe in those below 1 year. Pathogenesis The infection is concentrated in the bronchial walls, producing damage to the epithelium and accumulation of viscid mucus. Toxins produced by B. pertussis probably have a role in the tissue damage. Collapse of segments of lung distal to obstructed bronchioles is common. Clinical features After an incubation period of 7-10 days there is sometimes a prodromal illness that resembles a cold and lasts 2-3 days (catarrhal stage). This is followed by onset of the characteristic cough, with long spasms of coughing interrupted by rapid inspiration, which sometimes produces a whooping sound. This is often followed by vomiting. During a spasm of coughing the child becomes pink and then blue, and small babies may become apnoeic after a spasm. Between spasms the child looks and feels deceptively well. Whooping cough lasts about 2 months on average, and in the later stages the cough is mainly at night. It is distressing to watch small children rendered helpless by the spasmodic cough, and anxiety and disruption of sleep cause a major disturbance to family life. Complications Secondary bacterial infection causing pneumonia, or mucus plugging with segmental collapse, are the most common serious complications. Cerebral hypoxia during

1 318

MCQ9.19

Diagnosis Clinical diagnosis rests on recognition of the pattern of cough and is fairly simple when whooping is heard. However, in the early stages in babies and in adults the whoop may be absent. Laboratory support comes from the lymphocytosis and culture of B. pertussis from a pernasal swab; the latter is taken by passing a special swab through the nostril to the nasopharynx. Management Attentive nursing is the most important aspect of treatment, and parents need help and reassurance to learn to manage coughing spasms. Babies frequently feed poorly and lose weight, necessitating feeding through a nasogastric tube. The organism is sensitive to erythromycin, and early treatment with 125mg every 6 hours (up to 2 years) or 250 mg every 6 hours (over 2 years old) may be beneficial. Erythromycin can be used for secondary bacterial pneumonia if it has not already been started, or co-amoxiclav can be added. Prevention Pertussis vaccine is a crude preparation of killed organisms, but it is 80% effective in preventing whooping cough and often ameliorates the disease if not wholly preventing it. There is a small risk of convulsions after immunization (1 in 6000 doses) and a minute incidence (1 in 100000 doses) of encephalitis with lasting consequences. As there is severe morbidity and a small mortality in whooping cough epidemics the vaccine is strongly recommended. Prophylactic erythromycin treatment of close contacts has proved ineffective. 1

Brucellosis Brucellosis (undulant fever) is a septicaemic infection caused by Brucella abortus, Br. melitensis or Br. suis. Br. abortus is acquired from cattle, Br. melitensis from camels, llamas, sheep or goats (Malta fever), and Br. suis from swine. Most northern European infection is by Br. abortus acquired during occupational exposure in vets, farm workers and slaughterhouse employees, by ingestion of raw milk or by accidental infection in the laboratory. Most herds in the UK are brucella free, and pasteurization is an additional insurance against infection of milk. As a result of control measures the disease is now very rare in the UK but it remains a major problem in parts of Europe, the Middle East, and Central and South America. Brucellae are small Gram-negative coccobacilli which proliferate intracellularly. The incubation period of the disease is about 3 weeks.

Clinical features Some people have asymptomatic infections but the common presentation is an acute illness with persistent, swinging fever and profuse sweats. Back pain is common. There are no specific physical signs but 40% of patients have hepatosplenomegaly, 20% have splenomegaly only, and 20% have arthritis in one large joint. Less common manifestations are epididymo-orchitis, spondylitis, an erythematous papular eruption on the extremities, endocarditis, and neurological involvement with a range of manifestations, including meningitis, encephalitis or radiculitis. If no treatment is given the fever may persist for weeks or months, although it gradually declines, only to relapse again in many cases. A prolonged debilitating illness can be associated with depression. Depressed patients with moderate brucella antibody litres rarely have chronic brucellosis in the absence of a preceding history suggestive of acute brucellosis. Diagnosis There is neutropenia and lymphocytosis in the acute phase, and brucellae can be cultured from blood in 50-60% of cases and also from bone marrow in over 90% of cases. Complement fixing and agglutinating antibodies can be detected in rising titre following acute infection, and an ELISA-based serological test is also being used. Managemenf Inadequate treatment is associated with a high rate of relapse within 3 months, and some relapses occur despite the best known therapy. Doxycycline 200 mg daily combined with rifampicin 600 mg daily for 6 weeks, or doxycycline 200 mg daily for 6 weeks with streptomycin Ig intramuscularly daily for the first 3 weeks, gives the lowest relapse rates. Rifampicin is an enzyme inducer and may induce enzymatic breakdown of doxycycline reducing blood levels and hence its efficacy. Treatment is continued for 6 weeks in uncomplicated cases or 12 weeks in those with joint or spinal involvement. Endocarditis requires at least 12 weeks of therapy and probably surgical removal of the valve. In pregnancy rifampicin is given alone, and children less than 8 years old should receive co-trimoxazole plus rifampicin or streptomycin. There have been indications that rifampicin plus doxycycline has a greater relapse rate than streptomycin plus doxycycline. The explanation for this may be enzyme induction of doxycycline metabolism by rifampicin and so this effect could be overcome by increasing the doxycycline dose to ensure blood levels of the latter; this has not been tried. Because of susceptibility of brucella to other aminoglycosides such as gentamicin and netilmicin, these have been used for 5 to 7 days with good results although there are no controlled trials comparing these aminoglycosides with streptomycin. Some authors have suggested that for abscess, spondylitis, endocarditis and neurological disease treatment with three drugs, adding rifampicin to a doxycycline plus aminogly-

SUMMARY 11 Treatment of Brucellosis in adults Doxycycline 200 mg daily + rifampicin 600-900 mg daily

9

6 weeks uncomplicated 12 weeks complicated

or Doxycycline 200 mg daily + streptomycin 1g i.m. daily for first 3 weeks

6 weeks uncomplicated 12 weeks complicated

In patients with spondylitis Streptomycin 15mg/kg up to 1 g/day maximum i.m. (reduced dose in those under 50kg, over 40 years old or with renal impairment; drug level monitoring indicated in renal impairment) Children less than 8 years old Rifampicin 10-20 mg/kg/day for 8 weeks, together with cotrimoxazole for 6 weeks - every 12 hours, 6 weeks to 5 months, 120mg; 6 months to 5 years, 240 mg; 6-12 years, 480 mg.

coside regimen. For children, rifampicin plus co-trimoxazole can be used, adding an aminoglycoside when a third drug is needed. Preventive measures concentrate on elimination of brucellae from domestic animals, ideally by testing and slaughter of positive reactors, with compensation for the financial losses. Pasteurization of milk also prevents transmission, although in countries where the traditional custom is to drink milk and make other products using unpasteurized milk, changing these practices is immensely difficult. Vaccination of animals: the 19S strain of B. abortus in cattle or the Rev 1 strain of B. melitensis in sheep or goats is effective, but it should be noted that although these are attenuated strains for animals they are pathogenic for humans if accidentally injected.

Yersinia infection Yersinial infections are zoonoses. The important human pathogens are Y. enterocolitica and Y. pseudotuberculosis, which are enteric pathogens, and Y. pestis, the cause of plague. Y. pseudotuberculosis and Y. enterocolitica Infection with Y. pseudotuberculosis and Y. enterocolitica (see also p. 816) is less common in the UK than in some other European countries, notably Scandinavia and Belgium. The organism is a common commensal in the pharynx and tonsillar tissue of pigs, which are included in ground pig meat. Other foods, including milk and seafoods, have been incriminated in human outbreaks. Y. enterocolitica causes febrile enterocolitis in young children (less than 5 years old), but in adolescents an illness which is difficult to distinguish from acute appendicitis is caused by mesenteric adenitis. The latter is also the most common manifestation of Y. pseudotuberculosis infection. The presence of exudative pharyngitis may give a clue in

319

Y. enterocolitica infection. Many laparotomies have been carried out for these self-limited infections, which may be partially responsive to gentamicin therapy (often given as blind treatment or as prophylaxis for surgery). Reactive polyarthritis and erythema nodosum are well-recognized complications which are more common with Y. enterocolitica.

Plague An acute severe infection caused by Y. pestis manifests by either lymphadenitis with suppuration (bubonic plague) or by necrotizing pneumonitis (pneumonic plague). It is a zoonosis. Aetiology The organism is a Gram-negative aerobic bacillus of the parvobacteria family. It grows well at 28°C. At 37°C it produces an antigen which inhibits phagocytosis. It has a lipopolysaccharide capsular endotoxin. Distribution and incidence Asia, Africa, the Middle East and the Americas (South America and the southern states of the USA) are endemic areas with comparatively small numbers of reported cases each year. The disease is probably under-reported. Plague is considered a re-emerging disease because the number of cases being reported to the WHO is increasing, and since 1994 it has reappeared in several countries such as Malawi, Madagascar, Mozambique and India, where it had not been experienced for 15-30 years. Also, the number of foci have increased in some countries, including the United States. The last pandemic started in 1894, and by the middle of the twentieth century public health measures had led to it being purely sporadic. Transmission The infection is transmitted to man by the bite of the rat flea (Fig. 9.19). The natural hosts are rodents and the vector is the rat flea, Xenopsylla cheopis, which takes blood meals from its rodent hosts. Infected rats die and the fleas leave the dead rats for other hosts. The organisms produce a coagulant which prevents the flea taking a full blood meal. The flea bites the host repeatedly, encouraging transmission. The coagulant is inactive above 28°C and transmission diminishes. Pathology

Necrosis and haemorrhage are the main features, together with marked inflammatory changes. An eschar is occasionally found at the bite site. Buboes are grossly enlarged, inflamed lymph nodes draining the site of the flea bite. The normal lymph node architecture is destroyed and necrotic, with large numbers of bacilli. The pathological changes in pneumonic plague are very similar, with haemorrhagic necrosis of lung tissue. 320

Clinical features The incubation period is 2-8 days before the onset of

FIG. 9.19 Cycle of transmission of plague

systemic upset, fever, chills and headache. Within 24 hours of this the patient finds a bubo, which is exquisitely tender. Any group of lymph glands can be involved. Tachycardia and high fever are usual, with hypotension and shock in severely affected patients. Purpura may occur. Pneumonic plague results from haematogenous spread of organisms to the lungs. Cough, chest pain and haemoptysis in a severely ill patient are usual features. Patchy shadowing is seen on the chest X-ray. Aerosol spread can occur from patients with pneumonic plague. Septicaemic plague has a high mortality. There is a neutrophil leukocytosis. Some degree of disseminated intravascular coagulation is common, although spontaneous bleeding is uncommon. Diagnosis The organism may be found in material aspirated from the necrotic centre of buboes and in blood films. Smears made on slides can be stained with Giemsa stain, which shows the Gram-negative coccobacilli. The organism grows readily on standard media. Antigen (F1) can be detected in sputum by the second day of symptoms using a dipstick test which is simple and highly practical. Management The high mortality in untreated plague (over 50%) has been altered by antibiotics. Streptomycin 15mg/kg twice daily i.m. for 10 days is the treatment of choice. Defervescence occurs by the third or fourth day. Tetracycline and chloramphenicol are alternative choices but there are reports of multiple resistance to antibiotics mediated by a plasmid. Supportive treatment is needed in those with low blood pressure and shock, with volume replacement using saline or plasma. Buboes should be aspirated.

Prevention and control A vaccine is available for those working with plague in the laboratory or in the field. Proven cases are not infectious unless there is pneumonic involvement, when isolation of cases, with respiratory precautions to avoid aerosol spread, is necessary. Antibiotics rapidly reduce infectivity. Vector control with insecticides and rat control will minimize transmission in towns and cities.

followed by chloramphenicol 10mg/kg four times daily, doxycycline 2mg/kg twice daily, and co-trimoxazole (trimethoprim 10mg/kg/day + sulfamethoxazole 50 mg/ kg/day) given for up to 20 weeks. The efficacy of maintenance therapy depends on full compliance over this long period. There is evidence to suggest that co-amoxiclav plus extra amoxicillin gives equally good results, as good as the conventional regimen, with fewer side-effects, though full compliance is again necessary over 20 weeks.

9

PSEUDOMONAS TULARAEMIA Pseudomonas aeruginosa infection

Pseudomonas aeruginosa is an opportunist organism which rarely causes disease in immunocompetent subjects. It causes serious infection in patients with deficient immunity and in those who have had surgery. Septicaemia, osteomyelitis, infection of burns and pneumonia are examples. Skin changes due to Ps. aeruginosa are described in Chapter 10. Patients in intensive care units may become colonized superficially and treatment with antibiotics is often counterproductive, as resistance develops rapidly.

MELIOIDOSIS Burkholderi (Pseudomonas) pseudomallei is a Gramnegative aerobic bacillus with bipolar staining. It is found contaminating surface water, particularly in southeast Asia and northern Australia, although cases have been reported from a much wider geographical range. It has a very low host specificity and will infect a very wide range of animals, including humans. Infection occurs by ingestion or inhalation of contaminated water, or by contact of infected water with abraded skin. If health is good there may be subclinical infection, shown by seroconversion without systemic upset. Debilitating illness, for example diabetes mellitus and cirrhosis, and immunosuppression enhance susceptibility, and the development of clinical infection, which can be acute or chronic, may follow. After a 2-3 day incubation period there is fever, myalgia, vomiting, confusion or delirium. Diarrhoea may also occur, as may shock with features of endotoxic shock. Pustular skin lesions may occur. The lungs may be involved, with consolidation, cavitation and pus formation. Suppurative parotitis is seen in children. Chronic and relapsing sepsis may occur, with abscess formation affecting lymph glands, lungs, skin, liver, genitourinary tract and bones. Multiple small abscesses tend to form, producing multiple small cavities on liver ultrasound scanning. The organism is a dangerous pathogen which may be grown from pus, lymph gland biopsies, liver biopsy material and bone marrow aspirate. Serological responses using indirect haemagglutination (IHA) or complement-fixing antibody techniques are available. Antibiotic treatment comprises initial treatment with ceftazidime 2g t.d.s. intravenously for at least 2 weeks,

Francisella tularensis is a Gram-negative bacillus that infects a wide range of wild animals; humans are occasionally infected in rural areas of the northern hemisphere (North America, Sweden, Norway, northern Europe, Russia and Japan). Infection occurs by the bite of an infected deerfly or tick, by mucosal or skin contact with an infected animal, by inhalation, and by eating inadequately cooked meat. The effects in humans may be localized cutaneous pustular disease with regional adenopathy, a pneumonic illness, or a systemic illness sometimes with diarrhoea. Inapparent (subclinical) infections occur. The North American rabbit strain produces severe disease, with a mortality of up to 10% without treatment. Mortality is negligible (except for the American rabbit strain) even without treatment. The diagnosis is made by finding the organisms in material from infected tissues. Direct immunofluorescence may be used. Serological tests may also be used to detect antibodies to the organism by ELISA and agglutination techniques. Great caution is needed in handling specimens to avoid the risk of laboratory infection. Streptomycin (lOmg/kg twice daily for 14 days) is very effective. Tetracycline and chloramphenicol are effective. Patients need to be isolated until they have been on treatment for 48 hours to minimize the risk of cross-infection.

BARTONELLOSIS

Bartonella bacilliformis is a Gram-negative aerobic bacillus transmitted by the bite of the female sandfly Lutzomyia verrucarum. The disease is limited to the valleys of the western Andes, mainly in Peru, Colombia and Ecuador, between 2500 and 8000 feet above sea level. Humans are the reservoir of infection. There are two manifestations of the disease: haematological and cutaneous. After a 21-day incubation period there is sudden onset of systemic upset. The onset of fever is often followed by acute severe intravascular haemolysis, which may last up to 4 weeks before resolution. Several weeks to several months later the eruptive phase of the disease appears, with nodular lesions of various sizes over the body in the miliary form, or fewer nodular, deep lesions particularly

321

concentrated on the extensor aspects of the limbs. Mucous membranes of the mouth and gut, and serous cavities can be affected. The organisms can be grown in blood cultures taken in the first week of the illness and can be readily found in Giemsa-stained thick and thin blood films, or in material obtained from skin lesions. Treatment with antibiotics kills the organisms and produces rapid defervescence and clearance of organisms from the blood. Penicillin, streptomycin and chloramphenicol are effective; the latter is recommended in view of the reported association between systemic salmonellosis and bartonellosis. Treatment is for at least 7 days. Bartonella quintana This organism, formerly known as Rochalimaea quintana, was recognized as the cause of trench fever among troops in the First World War. The vector is the body louse, and so infections can occur when levels of sanitation are low. The organisms are present in louse faeces and infection occurs by introducing B. quintana through skin abrasions by scratching. There is no identified animal host. The organism causes a range of febrile conditions, including a febrile illness lasting 4 days or so, a relapsing febrile illness with three to five episodes of fever, and a continuous febrile illness lasting up to 21 days. Despite the association with the First World War the condition occurs worldwide among communities where sanitation is poor. Doxycycline or erythromycin are the treatments of first choice. This agent and the related B. henselae have been recognized as the cause of cutaneous bacillary angiomatosis among patients with HIV infection (see Ch. 11). Infection of deep organs such as the liver produces similar lesions.

itself appear ill. The cat is probably infectious for about 3 weeks. Clinical features Most cases occur in children or young adults. There is gross enlargement of a single group of lymph nodes, which are usually tender. Lymphadenopathy develops about 2 weeks after the scratch and usually lasts for 2-4 months, but it may persist for up to a year. A papule may occur at the site of the scratch; it appears about a week after the scratch and persists for 1-3 weeks. Although most patients remain well, one-third have a low-grade fever for a few days and may experience malaise, headache and sore throat. Rare manifestations are conjunctivitis and preauricular lymphadenopathy (oculoglandular syndrome), encephalopathy, thrombocytopenic purpura, osteomyelitis and pneumonia. Diagnosis The diagnosis is usually based on the histology of excised lymph node with negative cultures for bacteria and mycobacteria. The Warthin-Starry silver impregnation stain may reveal the CSD bacilli but they are present in small numbers. There may be a mild neutrophil leukocytosis, but other tests are unhelpful. Antibody tests are being developed. Management Because the disease resolves spontaneously the only treatment usually required is analgesia if the lymph nodes are painful, and reassurance. Antibiotics do not help.

VIBRIOS

322

CAT-SCRATCH DISEASE

Cholera

Cat-scratch disease (CSD) is a benign lymphadenopathy in a site determined by an area scratched by a cat, most often in the neck or axilla.

Cholera is due to infection with the Gram-negative bacillus Vibrio cholerae, which causes acute secretory diarrhoea with death due to dehydration.

Aetiology and pathogenesis Histology of affected lymph nodes reveals scanty small pleomorphic rod-shaped Gram-negative bacilli which are normally seen using the Warthin-Starry stain. These have now been identified as Rochalimaea henselae (possibly to be renamed Bartonella henselae) in most cases, although a small number may be a related organism called Afipia felis. The organism has been identified in lymph nodes from patients by culture, polymerase chain reaction for DNA, and by immunocytochemistry; serum antibodies are detectable in 84-96% of those affected. This organism is also one of the causes of the skin condition called bacillary angiomatosis which affects immunocompromised subjects such as AIDS patients (see Ch. 11). The infection is transmitted by scratching or licking by a young cat with chronic bacteraemia which does not

Aetiology V. cholerae is a curved Gram-negative motile bacillus. Two biotypes are recognized, the classic biotype and the El Tor biotype. The organism is present in vast numbers in the faeces of infected patients. The El Tor biotype tends to cause more asymptomatic and mild infections, which resolve spontaneously and leave the patient excreting vibrios to infect others. The organism is sensitive to desiccation but survives in saline waters. Vibrio cholerae 0.139 has been recently recognized in India and Bangladesh. Until that time the 0.1 serotype was the sole pathogenic organism and 0.2 to 0.138 were non-pathogenic. Distribution and incidence Distribution and incidence are shown in Figure 9.20. Cholera has been imported into European countries but

9

FIG. 9.20 World distribution of cholera 1994 (WHO)

no outbreaks have resulted, apart from that in Naples in 1974. Transmission and epidemiology Organisms are ingested in contaminated food and water. In India, Bangladesh and Pakistan, where cholera is endemic, it is predominantly a paediatric disease. In epidemic cholera all ages are affected. Individual susceptibility is governed by the infecting dose and by the gastric acid barrier. The 0.139 serotype affected adults and children in India and Bangladesh, indicating that there was no cross-protection between the usual serotype and the new one. Pathogenesis V. cholerae secretes an exotoxin which binds to the GM1 ganglioside, a glycolipid on the surface membrane of jejunal enterocytes (Fig. 9.21). The effects of cholera toxin are mediated by several mechanisms, one of which may be accumulation of cyclic AMP in the cell. Local neurohumoral stimuli may also contribute to the diarrhoea. The outcome is that the upper small intestine becomes a site of net secretion, with the accumulation of isotonic fluid containing Na+, Cl-, HCO3- and water. This fluid passes down the intestine and in the colon K+ exchanges for Na+ under the influence of aldosterone. Stooling rates of 500mL/h are common, and up to 1 L/h may be produced. The effect of cholera toxin lasts for the duration of the life of the enterocyte, which is 3-4 days; during this time the cell migrates from the crypt to villous tip, where it is desquamated into the gut lumen. There are no histological changes in the affected mucosa. Cholera is a secretory diarrhoea with no mucosal inflammation. Secondary changes in a range of tissues, e.g. kidney, occur as a result of hypovolaemia and shock. Clinical features The incubation period is about 2 days. Diarrhoea begins

FIG. 9.21 Mechanism of action of cholera toxin on intestinal cells A Attachment of cholera toxin to receptor on enterocyle. B Activation of cAMP - note different effects on villous and crypt cells.

suddenly, with vomiting during the first 24 hours of illness. Faecal residues disappear, and opalescent rice-water stools are passed. There are no pus cells in faecal smears. The abdomen may be distended and bowel sounds are increased. Stool losses are maximal within the first 48 hours of the illness and decline spontaneously thereafter, but death can occur from dehydration within 24 hours in the most severe cases. The clinical features are due to isotonic volume depletion (Table 9.14, p. 280). Haematocrit and plasma specific gravity rise with haemoconcentration. Hypoglycaemia may occur in young children and may be associated with convulsions. Loss of bicarbonate causes acidosis. There is prerenal uraemia. Colicky abdominal pain and slight fever are features of infection with the serotype 0.139. There is a considerable range in the severity of diarrhoea in cholera, particularly the El Tor biotype, which is the cause of many mild cases.

323

Diagnosis Severe diarrhoea causing rapid progression to dehydration suggests that cholera is the cause in severe cases. Milder cases will not be diagnosed unless a stool is cultured or immediate dark-ground microscopy of a fresh faecal smear shows the characteristic morphology of V. cholerae. Differential diagnosis All the acute infective causes of diarrhoea need to be considered, including ingestion of staphylococcal enterotoxin, invasive bacteria such as Salmonella, Shigella and Campylobacter, which are usually readily distinguished because of the presence of systemic upset and colicky abdominal pain, and enterotoxigenic strains of E. coli. Rotavirus infection may also be difficult to distinguish from cholera.

the duration of both diarrhoea and faecal excretion of Vibrio cholerae. In pregnancy and childhood ampicillin or furazolidone may be given. Co-trimoxazole is also effective. Prevention and control Adequate clean water supplies and the safe disposal of faecal waste would do much to prevent the occurrence of cholera. Concurrent education of the population about the relationship between infective diarrhoeal disease and personal and general sanitation practices is essential. The widespread use of ORS early in the course of all diarrhoeal illness reduces mortality. Vaccination with the current killed vaccine is of no benefit, either for prophylaxis in the traveller or for the control of epidemics.

Management Mild to moderate dehydration can be managed with oral rehydration solution (ORS) (p. 280), but patients with severe dehydration need intravenous infusions, e.g. Ringer's lactate, to restore blood volume. The aims of rehydration are: • To replace estimated fluid losses in the first 4 hours after presentation; • To replace continuing intestinal fluid losses with equal volumes until diarrhoea stops. Frequent small amounts of ORS are better tolerated than less frequent large volumes, especially if the patient is vomiting. Additional fluid, such as water, fruit juice, tea or breast milk, but not ORS, should be given to allow for insensible losses and obligatory urinary losses. The glucose in ORS is not sufficient to counter hypoglycaemia and so extra sugar should be given orally. A solution prepared using rice flour as the source of glucose is capable of reducing faecal fluid losses by producing net absorption of sodium, glucose and water from the gut. Bowel sedatives merely paralyse smooth muscle, leaving fluid containing organisms and the toxin in the intestinal lumen, and should not be used in management. Tetracycline, 250 mg four times daily for 3-5 days, will reduce

SUMMARY 12 Managing dehydration in cholera

324

Clinical assessment No dehydration

Management Encourage fluid intake Encourage normal diet

Some dehydration (5-10% body weight lost)

ORS 80mL/kg body weight over 4 hours + 400 mL ORS after each diarrhoeal stool + water to provide for insensible and renal losses

Severe dehydration (>10% body weight lost)

Ringer's lactate 30 ml/kg body weight over SOmin, then 70 ml/kg over 2 hours

CAMPYLOBACTER Campylobacter are curved rods morphologically similar to Vibrios but regarded as a separate genus, as they have different biochemical characteristics. There are six species, of which C. jejuni and C. fetus are important human pathogens. C. jejuni is a major cause of prolonged severe enterocolitis, frequently associated with bloody stools and abdominal pain and tenderness suggestive of peritonitis. It is now the pathogen most frequently isolated from patients with gastroenteritis and the incidence of infection peaks in late summer to autumn. Infection may be acquired from food or drinks, but there is often a relationship with domestic animals, a wide variety of which are normally colonized (e.g. dogs, fowl and cattle). In the tropics. C. jejuni can often be isolated from the faeces of asymptomatic subjects, but it frequently causes diarrhoea in children less than 5 years old, implying that asymptomatic infection occurs in those with some degree of immunity.

Pathogenesis The incubation period is up to a week, this and the severity of the illness being determined by the infecting dose of organisms. The infecting dose required to cause illness is about 104, but this may be reduced by conditions of low acidity in the stomach. Human bile encourages growth of the organism, and the small and large intestine are colonized and the mucosa is invaded. Although toxins are produced, they are not thought to contribute to pathogenesis. C. fetus sometimes causes bacteraemia but rarely C. jejuni, usually in the very young or elderly. Excretion in the stools continues for 2-3 weeks after infection in the absence of immunity, which causes more rapid clearance.

Clinical features Fever, headache and malaise begin 2-5 days after infection, shortly followed by diarrhoea, often with blood, cramping

abdominal pain and tenderness. The severity of pain and tenderness before the onset of diarrhoea sometimes suggests an abdominal emergency, such as appendicitis. The symptoms usually resolve within a week, but in a significant minority the duration is longer and there may be relapse in the absence of antibiotic treatment. Guillain-Barre syndrome is an uncommon complication of C. jejuni infection (about 1 in 2000 cases), although because the incidence of this infection is high it is linked to nearly 50% of cases of GBS.

Diagnosis The diagnosis is established normally by microscopy and culture of stools under microaerobic conditions, but occasionally blood culture is positive.

Management As with other forms of gastroenteritis, the first priority is rehydration and correction of electrolyte disturbances (see p. 280). Appropriate analgesia is required when abdominal pain is a prominent symptom. Drugs such as codeine phosphate which reduce gut motility should be used with caution in the acute phase. Simple analgesics (paracetamol) may be adequate. Antibiotics are indicated for severe cases when there is fever together with systemic symptoms. Erythromycin 40mg/kg/day in four doses may alleviate symptoms in children, and it terminates the excretion of organisms in the faeces. If treatment is clinically indicated before the organism has been identified, ciprofloxacin, 500 mg b.d. for adults, is more likely to improve symptoms and reduce the duration of illness, but there is increasing resistance to quinolones which necessitates local surveillance.

ANAEROBIC BACTERIA

Bacteroides fragilis is the most commonly encountered pathogenic anaerobe but many others are found among the normal flora of the gastrointestinal and genital tracts; these include other Bacteroides, Fusobacteria and some anaerobic cocci, such as peptostreptococci and microaerophilic streptococci. Anaerobic organisms often cause wound infection, abscesses and septicaemia in patients who have had abdominal surgery or intestinal perforation. Antibiotic sensitivity is restricted; metronidazole is very useful, as it is well absorbed orally or rectally and B. fragilis is usually sensitive to it. Clindamycin is also useful.

Mixed infections Infection may involve several different bacteria, including anaerobes. Abscesses associated with the bowel, especially the colon, or the vagina are examples. Some tissue infections, especially in debilitated patients, may also contain

multiple pathogens. Severe stomatitis, one form of which is known as Vincent's angina, commonly occurs in malnourished patients with poor dental hygiene. The pathogens responsible are members of the normal mouth flora, including anaerobes and spirochaetes which are sensitive to penicillin. Metronidazole is an alternative treatment. Necrotizing fasciitis is a rapidly spreading subcutaneous infection; it results in extensive necrosis, which may involve any part of the body. Type I is caused by mixed infection and type II by group A streptococci (see p. 386). Type I occurs particularly in diabetics, common sites being on the feet, spreading upwards, and on the perineum. Scrotal infection is called Fournier's gangrene. The organisms involved are mixed anaerobic bacteria, mainly Bacteroides or peptostreptococci. Treatment is by a combination of aggressive surgery and antibiotics aimed mainly at anaerobes, for example clindamycin plus ampicillin.

9

MYCOBACTERIA Mycobacteria are classified as a group because they contain mycolic acids. They are bacilli which are mostly slow-growing and able to survive within macrophages. On staining by the Ziehl-Neelsen method they are acid and alcohol fast. The most important pathogens within the group are Mycobacterium tuberculosis and M. leprae. Tuberculosis is still relatively uncommon in the western world but a major problem in less developed countries. The disease is discussed fully on page 640. M. bovis is rarely incriminated now but can cause similar disease. Other nontuberculous mycobacteria, sometimes called atypical, such as M. kansasii, are less pathogenic than M. tuberculosis but can cause disease. M. avium intracellulare has become more familiar in recent years because it is one of the infections that afflict patients with AIDS. Mycobacteria other than M. tuberculosis are often resistant to standard antituberculous drugs, so their treatment presents difficult problems in immunocompromised patients.

LEPROSY Leprosy (Hansen's disease) is a chronic infection of the skin and cutaneous nerves caused by M. leprae. The manifestations in the patient may be localized or generalized, and are determined largely by the host response to the infecting organism. Aetiology Mycobacterium leprae is an acid-fast bacillus that cannot be cultured in vitro. Ziehl-Neelsen staining of smears of the organism differentiates viable (solid staining) from non-viable (granular) organisms. In humans the cooler superficial regions of the body are the preferred sites for

325

Among those developing clinical disease there appears to be a racial variation in the type of leprosy occurring: up to 50% of cases in Asia have lepromatous disease, compared with 10% in Africa. The male-to-female ratio is 2:1 in adults and unity in children. Pathogenesis and pathology The host's cellular immune response determines the outcome of infection. If macrophages are able to kill the infecting organisms there is either no clinical disease or the lesions of indeterminate leprosy at most (Fig. 9.23). Those in whom the infection progresses further develop one of the forms of disease described in Table 9.30 and classified as one of the following: FIG. 9.22 Geographical distribution of leprosy

bacterial proliferation. The generation time is about 13 days. M. leprae was one of the first pathogenic bacteria to be identified, and yet in vitro culture remains elusive. Our knowledge of the organism's sensitivity to antimicrobial agents is gained from the response of mouse foot-pad lesions induced by injection of suspensions of M. leprae to antileprosy drugs given in water. It is known that M. leprae is a very hardy organism capable of surviving prolonged drying in the shade in south India, and capable of prolonged carriage in the noses of 5% of the general population and 16% of leprosy workers. Distribution and incidence Leprosy is mainly a disease of the tropics and subtropics (Fig. 9.22). It is relatively common in Africa and Asia, with a prevalence of 2-9 per 1000 in most areas and higher rates in a few countries. There are about 600000 new cases per year, including a considerable number of childhood cases, which indicates persistence of transmission. Lepromin skin test positivity rates in endemic areas are much higher than rates of clinical cases, suggesting that host factors do much to influence the outcome of the disease. The disease has almost completely disappeared from Europe, although cases do occur in Europeans who have had prolonged exposure in an endemic area. Transmission and epidemiology There is little evidence to support skin-skin contact as the route of infection. Aerosol spread from the upper respiratory tract of patients with multibacillary (lepromatous) leprosy is the likely source of infection. There is considerable variation in the course of infection from person to person, and the majority of those infected develop no clinical disease.

326

1

Fig. 9.15

2

Fig. 9.16

4

Fig. 9.18

5

Fig. 9.19

3

Fig. 9.17

• • • • •

Tuberculoid (TT) Borderline tuberculoid (BT) Borderline (BB) Borderline lepromatous (BL) Lepromatous (LL).

It should be noted that the clinical disease is a biological continuum between the two polar extremes of pure tuberculoid and pure lepromatous. Few bacilli are found in the tuberculoid forms, whereas large numbers are found in patients with lepromatous leprosy. Host cellular responses to mycobacterial antigens determine the form of disease that develops, with delayed-type hyper sensitivity at the tuberculoid end of the spectrum and anergy at the lepromatous end. The borderline forms are immunologically unstable, and without treatment will deteriorate to the lepromatous end of the spectrum. The pathology of leprosy is a granulomatous response. At the lepromatous end of the spectrum the granuloma comprises macrophages laden with solid staining bacilli and globules of fatty products of bacillary metabolism, including phenolic glycolipid 1. This appearance of the macrophages is called foamy macrophages. Although lymphocytes are conspicuous by their absence from lepromatous lesions they are well represented in tuberculoid lesions. The earliest lesions in leprosy begin in cutaneous nerves in the Schwann cells which have phagocytic functions and ingest bacilli. The most marked damage to nerve trunks is seen in TT leprosy. At the lepromatous end of the spectrum the involved nerves are not so large. The auriculotemporal, ulnar, median, radial cutaneous, lateral popliteal and sural nerves are preferentially involved at sites where the nerve trunks are close to the skin and therefore cooler. The bacterial load and the morphology of the organisms need to be assessed, initially for classification purposes and subsequently to determine the response to treatment. Slit skin smears are obtained from the edges of lesions. Numbers of bacilli and their viability can be graded. Lepromin is a skin test antigen obtained from M. leprae. The antigen is injected intradermally and the test is read at 3 weeks; a nodule at the site of injection indicates a positive response, meaning delayed hypersensitivity and supporting a classification at the tuberculoid end of the spectrum.

9

FIG. 9.23 Evolution of leprosy lesions with and without treatment (TT = tuberculoid; BT = borderline tuberculoid; BB = borderline; BL = borderline lepromatous; LL = lepromatous) (After Ridley D S 1977 Skin biopsy in leprosy. Ciba-Geigy, Basle.)

TABLE 9.30 Major features in leprosy Feature

Tuberculoid leprosy (TT)

Borderline leprosy (BB)

Lepromatous leprosy (LL)

Lesions Nerve involvement Numbers of bacilli Lepromin test Reactions Reversal Erythema nodosum leprosum (ENL)

Few (1-3) Local Fragmented only +++

More numerous Present in nerves of predilection Viable bacilli present in lesions -

Large numbers Diffuse, symmetrical, late in disease Vast numbers of viable bacilli -

-

+++

Two types of acute reaction are recognized in leprosy: • Erythema nodosum leprosum (ENL) is an immune complex-mediated vasculitis that occurs in LL patients and, less often, in BL, either during treatment or in the untreated patient. There is inflammation with oedema and a prominent infiltrate of neutrophils. • A reversal reaction indicates a change in cellular responsiveness to the infecting organism, either increasing (after starting treatment) or decreasing responsiveness. Skin lesions and affected nerves become more inflamed. The swollen lesion resolves completely in time, becoming smaller in 'upgrading' and bigger with 'downgrading'. Clinical features Indeterminate leprosy Indeterminate leprosy is manifest as a hypopigmented macule on a dark skin, and as a slightly reddened macule on a light skin, with either normal or mildly reduced sensation. The lesions may regress spontaneously or progress to one of the stages described below. Tuberculoid leprosy Tuberculoid leprosy (TT) is characterized by a hairless, hypopigmented anaesthetic lesion with a flat centre and a well-defined raised edge with minimal inflammation. 1 The skin of the lesion does not sweat. Single lesions are usual, but up to three may be found. Common sites are the extensor aspect of the arms, around the elbow and the

+++

knee. A thickened cutaneous nerve supplying the affected area may be felt. 2 To feel an enlarged nerve the examining fingertip should be moved to and fro over the nerve trunk at right-angles to the direction of the nerve. Neural TT leprosy occurs with a single thickened nerve as the only physical sign. Borderline tuberculoid leprosy Borderline tuberculoid (BT) lesions are similar to those of TT but more numerous and asymmetrically distributed. The edges of the lesions are inflamed and irregular. 3 The nerves preferentially involved in leprosy are usually thickened and damaged, causing anaesthesia, loss of function and deformity to a variable extent in hands and feet. Penetrating ulcers in the soles of the feet, burns and skin sepsis are often present. There may be osteomyelitis of the bones of the feet caused by an infected penetrating ulcer. Borderline leprosy Borderline leprosy (BB) produces multiple lesions that vary in form, size and shape. Papules, plaques and lesions with elevated wide outer rims and depressed, well-defined anaesthetic centres are seen. The lesions may be hypopigmented or hyperpigmented. 4 There is diffuse involvement of the peripheral nerves. Borderline lepromatous leprosy Borderline lepromatous leprosy (BL) causes numerous asymmetrically distributed inflamed or hyperpigmented lesions, which may be papules, nodules or plaques. 5 Sen-

327

sation over the lesions may be normal. Vague macules that vary in size are present in some patients. Peripheral nerves may only show slight enlargement. Lepromatous leprosy Lepromatous leprosy (LL) is associated with numerous symmetrically distributed skin lesions. 1 At the earliest stage, numerous slightly reddened hypopigmented macules and nodules occur. Without treatment these evolve into papules and plaques. The eyebrows are lost and the skin of the face and ears becomes diffusely thickened (leonine facies). Involvement of the nasal mucosae may block the nose and cause a bloodstained nasal discharge. Destruction of the nasal cartilage may occur later. Iritis and keratitis are often present. The voice may alter because of inflammation and thickening of the laryngeal mucosa. Testicular atrophy and gynaecomastia occur because of testicular damage. Bilateral oedema of the legs is often present. Nerve involvement occurs late in the course of the disease. Renal failure due to glomerulonephritis and amyloid disease are late complications in LL. Reactional states In reversal reactions the skin lesions and involved nerves become more inflamed and swollen. These reactions may also be accompanied by swelling of the face, hands and feet. Involved nerves become very tender, often with further deterioration in function. This type of reaction may persist for 2-3 months before subsiding. ENL produces acute inflammation of existing skin lesions in BL. The lesions become swollen and painful and slowly evolve through a livid appearance to residual hyperpigmentation of the affected skin. Bullae form over some lesions; these ulcerate and are slow to heal. New papules appear in LL cases, typically on the extensor aspects of the limbs. They may evolve over 2-3 days, through a bullous stage, to indolent ulcers which are slow to heal. Uveitis, lymphadenitis, orchitis and neuritis are also features of ENL. Diagnosis Diagnosis is based on the clinical features, characteristic histology, a positive lepromin test (at the tuberculoid end of the spectrum) and acid-fast bacilli in slit skin smears taken from the edge of lesions and other sites, the earlobes, elbows and knees (BB, BL and LL). In pure neural leprosy the diagnosis can be made from a biopsy of a few fascicles of an affected distal cutaneous nerve with no motor function, such as the radial cutaneous nerve at the wrist, or the sural nerve. Management

Multidrug therapy is now the rule and the current WHO

328

1

Fig. 9.20

2

4

Figs 9.23, 9.26

MCQ 9.20 5

3

Fig. 9.27

Figs 9.21, 9.22

TABLE 9.31 Treatment of leprosy Type Paucibacillary Tuberculoid (TT) and borderline tuberculoid (BT) Multibacillary Borderline (BB), borderline lepromatous (BL), lepromatous (LL)

Drugs Rifampicin, 600 mg monthly, supervised + dapsone, 100mg daily, unsupervised both for 6 months Rifampicin, 600 mg monthly, supervised + dapsone, 100mg daily, unsupervised + clofazimine, 300 mg monthly, supervised + clofazimine, 50 mg daily, unsupervised All drugs are continued for 2 years from time of diagnosis and thereafter until slit skin smears contain no bacilli

recommendations are set out in Table 9.31. These drugs are generally well tolerated. Dapsone occasionally causes haemolysis. Clofazimine may stain the skin a rather reddish hue. Mild upgrading reactions may be controlled with aspirin or non-steroidal anti-inflammatory drugs. More severe reactions that threaten nerve function require oral steroids. ENL may be controlled by steroids, but sometimes thalidomide is needed. The patient needs careful instruction regarding the care of hands and feet, prevention of cracking and infection of the skin of the feet and appropriate footwear. Where the cornea is exposed as a result of facial nerve palsy, early lateral tarsorrhaphy is essential to prevent corneal damage. Tendon transplants may help in managing foot and hand deformities. Prevention and control There is still a considerable stigma attached to leprosy; this is based on ignorance. Education of the public regarding the disease and the efficacy of treatment is being attempted. BCG vaccination reduces the incidence of leprosy. Currently there are efforts to define protective antigens which may result in a genetically engineered vaccine. 2

MYCOBACTERIUM ULCERANS

M. ulcerans causes chronic ulceration, usually on an exposed area of the body, often the lower limbs. The condition has a wide distribution in the tropics and is often called Buruli ulcer. The pathogenesis of M. ulcerans disease is related to its unique ability to produce a tissuedamaging toxin, mycolactone, recently characterized as a macrolide molecule which causes cell death by apoptosis. Histology shows numerous acid-fast bacilli within extensive areas of necrotic fatty tissue with little inflammatory reaction. The organisms can be grown in Lowenstein-Jensen medium at 30-33°C. Clinical lesions begin as a subcutaneous nodule which ulcerates through the skin.

© Subcutaneous extension produces an ulcer with markedly undermined edges. Patients (most commonly children aged 5-15) are systemically well unless there is secondary bacterial infection. Long-term complications are determined by the site of the lesion; facial ulcers can destroy an eye; limb lesions may require amputation; scars around a joint lead to contractures. The disease is a major cause of morbidity in West Africa. Standard treatment is surgical and excision of early lesions is usually curative. Established ulcers require extensive debridement to remove all infected tissue, followed by grafting. Antibiotics have not been shown to be effective despite in vitro sensitivities, but there have been few formal trials. Local heat may encourage healing, as the bacteria are killed by temperatures over 33°C. Many topical treatments have been tried but there is little evidence about their benefit. Recent evidence suggests that topical nitric oxide treatment may be beneficial. When the lesion affects a region near to a joint, active and passive movement of that joint must be encouraged to prevent contractures.

SPIROCHAETES Syphilis, caused by infection with Treponema pallidum, is primarily transmitted sexually and is discussed in Chapter 11 (p. 458).

YAWS Yaws is a chronic infection due to Treponema pertenue; it is characterized by a primary site of skin infection, with later dissemination to other cutaneous sites and periosteum. Unlike syphilis, cardiovascular and nervous systems are not affected and congenital infection does not occur. Distribution, incidence and transmission Following mass eradication campaigns yaws became very uncommon, but there has been a resurgence of cases in South America, west and central Africa, Indonesia and the Pacific region. The organism must be introduced into the skin by contact between an infective lesion and an abrasion in the skin of the recipient. Clinical features 4 The primary papillomatous lesion or 'mother yaw' develops at the site of infection. This lesion is highly infectious. Other forms are a localized area of dry papules or a localized maculopapular rash. Regional lymph nodes are often enlarged. Secondary infectious lesions appear within a few weeks, either before or after the primary lesion has healed. Rashes continue to appear over the next 5 years. Condylomata form at moist sites, and are highly infectious. Later periostitis may occur, affecting fingers, tibia (causing sabre

tibia) and nose (a destructive lesion called gangosa). Painful, deep fissures in the soles of the feet also occur.

9

Diagnosis and management The organisms are readily found in serous material from primary and secondary lesions examined by dark-ground microscopy. Serological tests for syphilis are positive. Penicillin aluminium monostearate is very effective given once only in a single dose of 1.5 g (0.75 g for children under 10 years). Tetracycline (2.0g/day in divided doses for 14 days) can be used in those allergic to penicillin, apart from pregnant women and children under 8 years, who should receive erythromycin for 2 weeks.

ENDEMIC SYPHILIS (BEJEL) Endemic syphilis is due to non-venereal infection with an organism that is intermediate between T. pallidum and T. pertenue. It is present in the desert regions of North Africa and the Middle East. Infection probably occurs by personto-person spread among children and older women who look after children. The first lesions are mucous patches in the mouth. Later, moist raised lesions of various sizes in axillae, groins, perineum and between the buttocks are seen. Hyperkeratotic lesions, like those of yaws, affect palms and soles. Bone involvement (periostitis) is associated with pain in the knees and shins. Destructive lesions of the nose, similar to gangosa in yaws, may occur. Congenital transmission does not occur, and late disease affecting the cardiovascular and nervous systems is very rare. Treatment with penicillin aluminium monostearate, as in yaws, is effective.

PINTA Pinta is a treponemal infection caused by T. carateum. Geographically it is localized to the drier regions of Central and South America. Transmission occurs by direct contagion from infectious skin lesions, and breaks in the skin may be a route of entry for the organism. The primary skin lesion is a papule sited on the extremities, around which smaller, similar lesions appear. 5 Spirochaetes may be found in serous material expressed from the erythematous and violaceous lesion. Serological tests for syphilis are positive. Treatment is with a single dose of benzathine penicillin (1.5g intramuscularly, or 0.75 g for children under 10 years).

RELAPSING FEVER Relapsing fever is an acute infectious illness caused by Spirochaetes of the genus Borrelia. Borrelia recurrentis is the cause of louse-borne relapsing fever, and a range of other species cause the tick-borne variety.

329

Aetiology and transmission Borrelia are spiral organisms which can be identified in peripheral blood smears using Giemsa and Leishman stains; they can also be readily identified using dark-ground microscopy and cultured on artifical media. Humans are the reservoir of infection of louse-borne relapsing fever. The human body louse, Pediculus humanus, is the vector. Sudden mass migrations of people grouped together in adverse conditions encourage the spread of lice and louse-borne diseases, relapsing fever and typhus. Tick-borne relapsing fever is a sporadic zoonosis in humans in areas where there is a cycle of transmission between ticks and wild rodents. The disease has a worldwide distribution, except for Australia and the Pacific region. Pathogenesis and pathology During the incubation period spirochaetes divide intravenously. Platelets are sequestered, intravascular coagulation occurs, and a bleeding tendency results. After 4-5 days of symptoms there is a crisis associated with massive phagocytosis of Borrelia, helped by opsonizing antibody. Following the crisis there is an afebrile period during which numbers of organisms expressing different surface antigens build up and symptoms recur. This can occur up to five times. The spleen, liver, heart and brain are the main sites of pathological lesions, which consist of macrophages surrounding and ingesting large numbers of organisms. Clinical features After an incubation period that varies from 4 to 18 days the illness begins suddenly with rigors and a fever which rises rapidly to 40°C. Headache, joint pains, anorexia, malaise, nausea and vomiting are usual features, with mental confusion frequently present. Hepatosplenomegaly is common, with jaundice present in some patients. A petechial rash, most prominent on the trunk, is seen. Bleeding from the nose, the respiratory tract or the gut may occur, as may focal neurological signs. The first febrile period lasts about 5 days before spontaneous defervescence. This follows a severe febrile paroxysm with cold extremities, a rising pulse rate and blood pressure, and tachypnoea. An afebrile period lasting about 7 days recurs before the first relapse. Tick-borne disease is milder than louse-borne disease. In both, death can occur during the initial rigor or during defervescence. Diagnosis Organisms can be found in Giemsa-stained peripheral blood smears. 1 Dark-ground microscopy can be used to demonstrate the motile spirochaetes in a fresh blood smear.

1

330

Fig. 9.28

2

MCQ 9.21 3 Fig. 9.29

Management Antibiotics kill spirochaetes readily. Tetracycline can be used in patients over 8 years old. Erythromycin can be used in pregnant women and young children. Louse-borne relapsing fever requires a single dose of 500 mg of either drug. Tick-borne relapsing fever requires treatment for 10 days. Supportive care is needed, particularly in patients with louse-borne disease, as they can develop severe reactions, including Jarisch-Herxheimer, after being treated. Central venous pressure monitoring is essential if facilities are available. Hypoxia is corrected by giving oxygen continuously. Vitamin K is given if the prothrombin time is prolonged. Prevention Delousing by washing with soap and water, followed by dusting with 10% DDT, is effective. Clothes should be washed with soap in water at 55°C to kill lice and nits (lice eggs), which are found in the seams of clothes. Lice cannot survive on a person who has two sets of clothes worn on alternate weeks.

LYME DISEASE Lyme disease is caused by infection with Borrelia burgdorferi, transmitted by the tick Ixodes dammini. It is described on page 1160.

LEPTOSPIRAL INFECTION Infection with Leptospira icterohaemorrhagiae is discussed on page 879. 2

ACTINOMYCOSES Actinomyces species, despite their name, are bacteria which are often confused with fungi because of their filamentous appearance and their propensity to produce chronic suppurative infection with sinuses discharging purulent material. Most human disease is caused by Actinomyces israelii, an oral commensal, and occurs in the presence of chronic lack of dental hygiene. The characteristic pathological feature is that tissue boundaries are crossed and bone is infected. There is a fibrotic reaction around suppurating areas which accounts for the induration detected in palpable lesions. Three major forms of disease occur: cervicofacial, thoracic and disseminated actinomycosis. In cervicofacial actinomycosis the characteristic lesion is a painless swelling below the border of the mandible, bluish, fluctuant and slowly enlarging. Sometimes the abscess is painful and is usually apparent long before there is any spontaneous discharge. Thoracic actinomycosis may result from aspiration of oral debris, or may occur after thoracic surgery; it is easily

SUMMARY 13 Actinomycosis • • • • •

Cervicofacial form associated with poor dental hygiene Thoracic form may follow surgery Abdominal form is usually ileocaecal Discharging sinuses occur Prolonged penicillin treatment is needed

confused with carcinoma. The chest radiograph shows unilateral consolidation, with evidence of chest wall involvement. Nowadays the disease is usually diagnosed before discharging sinuses develop. Actinomycosis may be disseminated haematogenously and lesions may occur at almost any site. The abdomen is the most common, usually in the ileocaecal region. Because of the chronic course and the difficulty of diagnosis, sinus or fistula formation is still seen in abdominal actinomycosis. Other rare sites of infection are the pelvis, CNS and bones. Diagnosis and management Microscopy of tissues reveals the Gram-positive organisms, with surrounding acute or chronic inflammatory reaction and foamy macrophages. The organism can be cultured from tissues and, rarely, from blood. Penicillin is the drug of choice and prolonged treatment in high dose is necessary. Intravenous administration of 2.4-3.6 g every 6 hours for 4-6 weeks is followed by oral treatment with amoxicillin 500 mg every 8 hours for at least 6 months. Tetracycline, erythromycin and chloramphenicol are alternatives in penicillin-allergic subjects. Relapse is uncommon if this rigorous approach is adopted. Surgery is rarely required but may have to be considered in severely ill patients who fail to respond to medical therapy, and possibly in those with CNS disease.

MADURA FOOT This is a descriptive term for a group of chronic infections of the subcutaneous tissues and bone of - usually, but not invariably - the limbs. Two groups of organisms cause this condition: • Actinomycetes cause 60% of cases; these are bacteria and respond to antibiotics. • True fungi such as Madurella mycetomatis, which are almost totally unresponsive to antifungal drugs. These two groups of organisms are found in the soil and enter the skin by penetrating injuries. Pathology and clinical features The main features are chronic inflammation with microabscesses and granulomatous reactions around collections of the organisms. The organisms spread slowly along fascial planes and not via lymphatic or haematogenous routes.

Collections of organisms forming adjacent to bone produce local resorption, and small lytic areas visible on X-ray. 3 There are also areas of reactive sclerosis. Sinuses form and organisms are extruded. There is swelling of the affected area, with often prominent subcutaneous nodules. The time course is slow, and although there may be recollection of a penetrating injury, such as a thorn going deeply into the foot in childhood, there is often no such history and presentation may occur years after the patient has left an endemic area.

9

Investigations and management The most important investigation is an adequate surgical biopsy of diseased tissue, which must usually go down to bone. This specimen is divided to give samples for histology and culture for the two groups of organisms. Actinomycetoma responds well to prolonged courses of antibiotics. Co-trimoxazole is effective (two tablets twice daily for 9 months, with streptomycin 1 g intramuscularly daily for the first 2 months). Dapsone, 100mg daily, is an alternative to co-trimoxazole. Pigmented grains fungi may respond to ketoconazole, whereas non-pigmented species are treated with itraconazole. Madurella mycetomatis has responded to ketoconazole 200 mg twice daily, and Aspergillus nidulans, A. flavus and Fusarium species have responded to 100 mg twice daily, in prolonged courses. Surgical removal of diseased tissue in addition to antifungal agents is recommended for some fungal infections.

NOCARDIOSES Nocardiosis is a systemic infection caused by species of Nocardia (N. asteroides and, less often, N. brasiliensis and N. caviae). N. brasiliensis is the common cause of actinomycetoma in the tropics and subtropics. Nocardia are higher bacteria which exist in a filamentous and coccobacillary form, depending on culture conditions. Aetiology, pathogenesis and epidemiology Nocardia are aerobic, filamentous branching Gram-positive bacteria and grow well on standard culture media. The organisms are found in rotting vegetation and infection is acquired by inhalation in most cases, although inoculation via the skin occurs. Initial pulmonary lesions may lead to dissemination of infection by the bloodstream. Necrosis and suppuration are the typical pathological changes. Adults are most often affected. Infection occurs mainly in immunosuppressed individuals. Clinical features Acute and sometimes fulminant or chronic suppurative pulmonary disease occurs in nocardiosis. Fever, cough with viscid sputum, night sweats and weight loss are usual features. Pleural thickening and empyema may develop, and the affected lung may cavitate. Haematogenous dissemination to cause brain abscess, kidney lesions, bone lesions and subcutaneous sepsis is reported.

331

Actinomycetoma due to Nocardia presents with longstanding swelling of the extremity, most often the foot or ankle, and sinuses discharging on to the skin.

diagnosis and can give information about the response to treatment. An adequate deep biopsy will give the best chance of isolating the organism concerned.

Diagnosis Diagnosis depends on finding the organism in discharges, exudates, bronchial washing or biopsy sections. Nocardia grow well on media used for growing tubercle bacilli and on Sabouraud's medium. Serological tests are helpful in

Management Co-trimoxazole is effective (two double-strength tablets every 8 hours). Prolonged treatment is the rule, with relapse being less common after 3 months, medication. Minocycline (100-200 mg twice daily) is also effective.

SYSTEMIC MYCOSES Mycoses are infections caused by organisms that are classified morphologically as yeasts or filamentous fungi. Yeasts are round or oval and reproduce by budding. Filamentous fungi grow by branching and longitudinal extension of hyphae. This is not a simple distinction, as several organisms can grow in either form, depending on conditions. 1

CANDIDA INFECTIONS The range of diseases caused by Candida albicans is very broad. It is a common commensal of the oral and vaginal mucosae at one extreme, and at the other it can cause pneumonia, endocarditis, septicaemia and death. Candida can become a pathogen on damaged skin, in severely ill patients, in patients who have specific immune deficiency (especially AIDS), and in patients receiving

SUMMARY 14 Candida infections Skin

Secondary infection of damaged skin Mucocutaneous candidiasis (specific T cell deficiency)

Oral

Broad-spectrum antibiotic use Topical corticosteroids (e.g. inhalers for asthma) Immune deficiency (e.g. AIDS) Severe illness

Vaginal

Antibiotics or contraceptive pill Spontaneous (no immune deficiency)

Pneumonia

Due to severe illness with secondary immune deficiency

Septicaemia

Severely ill patient; multiple antibiotics Intravenous cannulae Neutropenia

1

332

Figs 9.30, 9.31

2

Fig. 9.32

broad-spectrum antibiotics when the local microbial ecology is disturbed. Both cell-mediated and humoral immunity participate in control of Candida. Inherited deficiency of T-cell function causes severe but localized disease. In combined immune deficiency the infection becomes disseminated. Severely ill patients under intensive care frequently have a mixed immune deficiency and are also receiving antibiotics that predispose to Candida infection. Disseminated infections are also seen in intravenous drug abusers. Cutaneous and mucosal surface candidiasis is described in Chapter 10. Diagnosis of systemic Candida infection Candida septicaemia is often a terminal event and over half of patients are diagnosed at postmortem. Clinical signs are usually minimal, but there may be endophthalmitis with fluffy yellow/white retinal spots resembling cottonwool spots, or small abscesses may be detected elsewhere, such as in the liver, where ultrasound-guided biopsy or aspiration can clinch the diagnosis. Pulmonary infiltrates may occur on chest X-ray, with Candida in bronchial washings. Isolation of Candida from a single site on one occasion does not necessarily indicate serious infection, even when it comes from a blood culture. The judgement that Candida is responsible for disease manifestations remains clinical. In view of the difficulty in diagnosis, treatment is given for disseminated Candida infection if Candida is cultured from more than one site in a severely ill patient in whom there is new evidence of infection (e.g. fever) but bacterial infection is judged to be unlikely. Management Fluconazole is the first choice for disseminated candidiasis. It is given intravenously, 400 mg by infusion initially, followed by 200 mg daily until oral therapy can be substituted. Maintenance treatment is needed until antibiotics are stopped. Toxicity is very uncommon. If there is no response to this drug amphotericin B is substituted, starting at 0.25 mg/kg daily infused over 1 hour, increasing rapidly to about 0.8 mg/kg daily. The maximum dose is limited by sideeffects such as fever, uraemia and bone marrow toxicity. For management of local Candida infections and the emergence of drug resistance, see pages 391 and 453.

ASPERGILLUS Aspergillus is a less common fungal pathogen which can cause disseminated disease in terminally ill patients. However, its more common manifestations are a range of respiratory diseases, the nature of which is determined by the immune response of the host (see Ch. 13). 2

CRYPTOCOCCOSIS Cryptococcosis (torulosis) is a systemic infection with the yeast-like fungus Cryptococcus neoformans. The organism is commonly found in soil contaminated by bird droppings, notably those of pigeons. Disease occurs following inhalation of the organism, but it is unusual in immunocompetent people and currently is seen most often in patients with AIDS (Ch. 11, p. 448). Clinical features CNS infection causes chronic meningitis with very lowgrade symptoms on presentation. Cranial nerve palsies may cause visual abnormalities, loss of acuity or diplopia, or facial weakness and numbness. The course is slow or rapidly progressive, depending on the patient's immune status. Corticosteroid therapy accelerates progression. Respiratory infection usually occurs in the absence of CNS involvement. It is more chronic and may regress spontaneously. Skin lesions are found in 10% of patients. The face and scalp are the most common sites of the painless lesions, which are variable nodules, sometimes ulcerating. Diagnosis Routine culture is often negative. In the CSF little abnormality is found, apart from a few cells, mainly lymphocytes, mildly reduced glucose compared with the blood, and slightly raised protein. The organisms are shown by India ink preparations, but the most sensitive method of diagnosis is by detection of antigen in CSF and sometimes in blood. Management Treatment is started with fluconazole 400 mg daily, by mouth if the patient is not vomiting, or intravenously at the same dose. If there is no response after 2 days amphotericin B is substituted, giving up to 1-1.5 mg/kg/day until there is a clinical response, and then reducing to three times weekly to a total dose of 1-2.5g (see Candidiasis).

HISTOPLASMOSIS Histoplasma capsulatum is a common environmental fungus found in the soil in a filamentous form, but when it infects humans it behaves as a yeast. It multiplies rapidly in soil containing the droppings of birds and bats, and humans are usually infected by inhalation of spores in

infected dust. Infection has been shown to be common by skin test surveys in the eastern central part of the USA, but there are also scattered areas of infection in both temperate and tropical areas of the world. H. capsulatum var. duboisii is a variant found in central Africa. H. capsulatum infects macrophages, and the pathogenesis of disease is similar to that of tuberculosis. Delayed hypersensitivity skin tests show that the majority of infected individuals are asymptomatic. Clinical features of H. capsulatum Acute pulmonary infection in normal hosts causes febrile illness with headache, often accompanied by cough and chest pain, with malaise, myalgia and weight loss. Erythema nodosum or multiforme may occur. Symptoms usually last less than a week, but occasionally severe infection is seen following a large infecting dose and it may even be fatal. Radiological changes occur in only 25% of acute infections, with scattered pulmonary infiltrates and hilar and mediastinal lymphadenopathy. The histoplasmin skin test is not helpful in diagnosis, but a complement fixation antibody test may be useful. Treatment is often not required but, now it is established that the azole drugs, ketoconazole and itraconazole, are beneficial, they can be given more easily than amphotericin B and with less risk of toxicity, and they may make patients feel better earlier than if they receive no treatment. Ketoconazole (400 mg daily) or itraconazole (200 mg daily) is given by mouth for 3-6 weeks. Amphotericin B (0.8mg/kg daily) is reserved for severe illness. Concurrent prednisolone therapy, tapering from 60-80 mg/day, is given if there is a marked hypersensitivity reaction to the infection. Chronic pulmonary histoplasmosis affects normal individuals with structural defects of the lung. Infection of emphysematous lesions causes chronic destructive disease in the lung apices, similar to tuberculosis. Fever may be accompanied by night sweats; one-third of patients have chest pain. Diagnosis requires repeated sputum cultures. Antibody and skin tests are usually unhelpful. Treatment is with ketoconazole, 400 mg daily for 6-12 months, or longer if there is evidence of relapse. Itraconazole is more expensive but may cause fewer side-effects. The need for treatment is judged on the basis of clinical severity. Disseminated histoplasmosis is a rare and often fatal manifestation of infection in immunocompromised hosts. It may be acute, subacute or chronic, depending on the age of the patient (acute is more frequent in infants) and the degree of parasitization of macrophages (more severe parasitization causes more acute disease). Fever, hepatosplenomegaly, anaemia, leukopenia, thrombocytopenia and interstitial pneumonia are most common in the acute disseminated form. Intestinal ulceration, Addison's disease, meningitis and endocarditis occur in the subacute form, and oropharyngeal ulcers occur in chronic disseminated disease. Diagnosis is by histology and culture of relevant tissues, and treatment is as for the other forms

9

333

of the disease. Amphotericin B is reserved for the most severe cases where the prognosis is poor, and it is the first choice of treatment for progressive disseminated histoplasmosis associated with AIDS, which does not respond satisfactorily to ketoconazole.

Histoplasma capsulatum var. duboisii H. capsulatum var. duboisii infection occurs in Africa. The route of infection is poorly defined; asymptomatic pulmonary infection, with later spread to the skin and bones or inoculation into the skin, is a possibility. Skin, subcutaneous tissues and underlying bone can be affected, causing: • Superficial, cutaneous granulomas (papules or ulcers) • Subcutaneous granulomas and abscesses • Osteomyelitis with discharging sinuses which heal intermittently. Smears from exudate or discharges show numerous large yeast cells. Histological sections of infected tissue show giant cell granulomas with the large yeast cells. Treatment with ketoconazole (400 mg daily) or amphotericin B may be needed in lesions that show no signs of regression.

COCCIDIOIDOMYCOSIS Coccidioidomycosis is a fungal infection caused by Coccidioides immitis. Its distribution is confined to arid areas of the western hemisphere, including the southwestern USA, Central America and some countries of South America. Dust contaminated with fungal spores is the source of infection. Granulomatous responses occur around fungal spherules, and suppurative lesions with leukocytes characterize the tissue response to endospores. The host response to the organism determines the outcome of disease. Asymptomatic infection occurs in about three-fifths of patients and is detected by skin testing. A smaller number of those infected present with respiratory symptoms, cough, fever and chest pain, with malaise, muscle pains, night sweats and chills. A macular erythematous rash, erythema nodosum and erythema multiforme also may be seen (p. 407). Radiographic appearances include pneumonitis, hilar adenopathy and pleural effusion. In most of these patients the disease resolves spontaneously. In some cases pulmonary infiltrates and systemic symptoms persist, sometimes with cavitation. The diagnosis is made by finding the endospore of the organ-

1

334

MCQ 9.22

ism in material from affected lung. There is a considerable risk of laboratory infection with this organism. Serological tests are an aid to diagnosis. Dissemination of infection from an initial site of pulmonary disease may occur. Pregnant women show an increased incidence of dissemination. All tissues and organs of the body may be affected. Meningitis is one of the most severe manifestations of disseminated infection. Amphotericin B is used in treatment. The main difficulty lies in deciding which patients will benefit from treatment. Persistent, symptomatic pulmonary infection and disseminated infection require chemotherapy.

PARACOCCIDIOIDOMYCOSIS Paracoccidioidomycosis, or South American blastomycosis, is unrelated to Coccidioidomycosis. It is caused by Paracoccidioides brasiliensis and occurs exclusively in the area between Mexico and Argentina. The natural environment of the pathogen is uncertain. Its mode of acquisition is probably through inhalation. Clinical features, diagnosis and management Adults usually present with progressive respiratory symptoms, sometimes accompanied by ulcers in the mouth and nose, dysphagia, voice changes, skin lesions and cervical lymphadenopathy. Diagnosis is usually made by culture or histology, and treatment is with imidazole (particularly itraconazole), sulphonamides or amphotericin B.

BLASTOMYCOSIS Blastomycosis is a primarily pulmonary disease caused by the dimorphic fungus Blastomyces dermatitidis. It is much less common than histoplasmosis and occurs mainly in the eastern USA. The organism has been isolated from soil, and infection may be by inhalation of dust. Clinical features Primary infection may be asymptomatic, or it may lead to an acute pneumonia which either resolves spontaneously or progresses to chronic pulmonary disease, which may be confused with tuberculosis. Extrapulmonary manifestations are caused by haematogenous spread from the lungs, and the most common is skin disease with verrucous or ulcerative lesions. Diagnosis is by demonstration of the yeasts in sputum or other specimens, or by culture of the organism. Amphotericin B is the mainstay of therapy, but imidazoles have been used successfully.

SPOROTRICHOSIS

MUCORMYCOSIS

Sporothrix schenckii is a dimorphic fungus found in soil, on plants and on plant debris. Infection is uncommon and is seen in the Americas and Africa. Disease usually results from traumatic inoculation into the skin, and the common form is a red painless papule at the site which enlarges and forms a granulomatous plaque that commonly ulcerates. Further lesions often appear proximally along the course of the lymphatic channels, which are often palpably thickened by inflammatory infiltrate such that they could be mistaken for thrombosed superficial veins, and in regional lymph nodes. Disseminated disease is rare and usually only occurs in immunosuppressed patients. Diagnosis is by culture of the organism from skin lesions. Cutaneous lesions are treated with a saturated solution of iodine, 5-10 drops t.d.s. orally and gradually increasing to 50 drops t.d.s. for up to 3 months, and extracutaneous disease with amphotericin B. Itraconazole can also be used.

Mucormycoses are a group of similar diseases caused by members of the order Mucorales. The organisms are common in the environment, often growing on fruit or bread. They cause disease almost exclusively in immunosuppressed patients.

9

Clinical features In rhinocerebral disease in an uncontrolled diabetic the fungus invades the palatal, nasal and paranasal sinus mucous membranes, and destructive lesions spread directly back towards the brain. Black necrotic lesions are visible, and from these a variety of secondary bacterial pathogens are often isolated. Pulmonary, cutaneous and intestinal disease may occur. Diagnosis is made by examination of scrapings and biopsies, and cultures are often negative. Successful treatment depends on management of the underlying cause. Amphotericin B is the only useful drug in treatment, but the response is often poor. 1

PROTOZOAL INFECTIONS

Protozoa are single-cell organisms that have adapted to a range of ecological niches in the human host (Table 9.32). They are an important cause of human disease, especially in the developing world.

PROTOZOAL INFECTION OF THE GUT

GIARDIASIS Infection with the flagellate protozoan Giardia lamblia can cause asymptomatic infection or diarrhoeal disease of variable severity, from mild to severe with malabsorption. Aetiology The parasite is a flagellate, pear-shaped protozoan (Fig. 9.24) which is found in the lumen of the upper small intestine. Some trophozoites encyst, to be passed in the faeces. The cyst is infective and can survive for 2 months at 8°C in water. It is resistant to normal levels of chlorination, and sand filtration is used to clear municipal water supplies of this and other parasites. The parasite can now be grown in culture. Distribution ana1 incidence The distribution is worldwide, although it is more common in the tropics. It is endemic in the countries of eastern Europe. Epidemics have occurred at daycare nurseries, on cruise ships and in towns in the USA. Giardiasis has

occurred in children in inner-city populations in Britain, and can cause diarrhoea in the elderly. Transmission ana1 epidemiology Spread is by the faecal-oral route. As few as 10 cysts will cause infection, and 1000 cysts consistently cause infection. Contaminated food and water are vehicles of infection. Person-to-person spread is common in childhood, particularly when children are not toilet trained. Mothers are often infected by changing the nappies of an infected child. Any circumstances in which standards of personal and public hygiene (such as water treatment) are low lead to transmission. In the tropics children are most often infected, although both indigenous and visiting adults can develop symptomatic disease. Male homosexuals and retarded children are other risk groups. Hypogammaglobulinaemia and reduced gastric acid secretion are host factors that increase susceptibility. Pathology and pathogenesis The main abnormalities relate to the function and morphology of the upper small intestine. Markedly symptomatic patients have impaired absorption of fat, o-xylose and vitamin B12, and lactose maldigestion. The jejunal mucosa is abnormal, with a ridged or convoluted mucosa, reduced villous height with increased crypt depth, and an increased infiltrate of plasma cells in the lamina propria. Subtotal villous atrophy may occur with giardiasis but is uncommon. Patients with mild or no symptoms have

335

TABLE 9.32 Classification of protozoa Gut protozoa Giardia lamblia Cyclospora cayetanensis Cryptosporidium Entamoeba histolytica Balantidium coli Isospora belli Sarcocystis Non-pathogenic gut protozoa Entamoeba coli Entamoeba hartmanni Endolimax nana lodamoeba butschlii Chilomastix mesnili Blastocystis hominis Entamoeba dispar Genital tract protozoa Trichomonas vaginalis Tissue and blood protozoa Toxoplasma gondii Pneumocystis carinii Plasmodium species

Babesia species Trypanosoma gambiense, rhodesiense and cruzi Leishmania species Free-living protozoa causing human disease Naegleria fowleri Acanthamoeba

normal jejunal morphology and function. Lactose maldigestion may be present. The pathogenesis is not well understood. The parasite itself may damage the enterocyte surface membrane to impair function and fat digestion intraluminally. Jejunal colonization with bacteria may also contribute to some of the mucosal dysfunction. Antigiardia IgA from the mucosa and in bile may control parasite numbers. Clinical features The incubation period is usually about 10-14 days, although it can be much longer. Many patients with giardiasis are asymptomatic or have minimal bowel upset. Acute giardiasis is characterized by the sudden onset of anorexia, nausea, abdominal distension, discomfort and diarrhoea with frequent yellow, offensive, frothy stools by day and night. Lethargy is often severe and weight loss is usual. After about 3 weeks there may be the beginnings of spontaneous improvement. This may progress to complete resolution over a month, but some patients remain mildly symptomatic, often because of continuing lactose intolerance. Some patients remain markedly symptomatic and fail to regain lost weight or continue to lose weight. The abdomen is distended and bowel sounds are prominent. The stools 1

336

MCQ 9.23

FIG. 9.24 Scanning electron micrograph of Giardia lamblia (Reproduced by kind permission of Dr David Warhurst.)

are yellow and offensive. Testing confirms malabsorption. Children are occasionally brought to medical attention because of failure to thrive. Giardiasis has been reported as a cause of chronic diarrhoea in elderly persons in the UK. It is not a major cause of diarrhoea in AIDS. Diagnosis Diagnosis depends on finding the parasites. Stool microscopy shows cysts in most patients, although examination of several samples may be necessary. Trophozoites may be found in diarrhoeal stools. When the parasite is not found and symptoms are marked, investigation of intestinal morphology and function is indicated. Jejunal juice and jejunal mucus obtained at the time of biopsy can be examined for trophozoites. Giardia may be seen in the intervillous space of the sections of the biopsy. Giardia antigens can be detected in stools by immunological techniques. Differential diagnosis Other causes of gastroenteritis are discussed on page 278. Cryptosporidiosis causes self-limited diarrhoea in healthy persons (see below). Cyclosporiasis causes a similar acute illness (see below) Campylobacter species can cause a small bowel type of diarrhoea but usually cause systemic upset. The differential diagnosis of malabsorption syndrome is discussed in Chapter 15. Laboratory features Laboratory investigations show malabsorption of fat, Dxylose and vitamin B12 in severely affected patients. DXylose malabsorption and lactose maldigestion may also be present in those with mild symptoms. Occasional patients are seen who have marked malabsorption and folate deficiency. Barium follow-through examinations show nonspecific changes, with dilatation of small bowel loops and thickened mucosal folds in patients with malabsorption.

SUMMARY 15 Giardia

SUMMARY 16 Treatment

Anorexia Lethargy Weight loss Diarrhoea Yellow mushy to watery stools Abdominal distension Sulphurous eructation Foul-smelling flatulence

First choice Tinidazole 2g single dose with evening meal Repeat 1 week later Precautions: no alcohol for 48 hours after dose, as it can make people 'woozy'. Also, patients should not ride a bike, drive a car or operate any dangerous machines for 48 hours after doses

Note: These are typical symptoms; in an acute severe case there is a range of severity, as indicated in the text.

Management Tinidazole is effective and can be given in a single dose of 2g (50mg/kg) which can be repeated after 1 week. A cheaper alternative is metronidazole, 2g as a single dose on 3 successive days. A second course after 10 days may increase the cure rate. Both drugs cause nausea and a metallic taste in the mouth, and have a disulfiram-like interaction with alcohol. Asymptomatic giardiasis in pregnancy need not be treated. When there is symptomatic disease in pregnancy associated with weight loss or failure to gain weight, then metronidazole (200 mg three times a day for 10 days) may be given. Symptoms due to giardiasis improve rapidly after treatment. Dietary measures are sometimes helpful for continuing gut symptoms. Avoidance of alcohol, spicy foods and lactose is often helpful. Repeat stool microscopy 6-8 weeks after treatment provides a test of cure. Abnormalities in intestinal structure and function disappear over 6-12 weeks after treatment. Prevention and control Travellers in areas where the tap water is not safe to drink should avoid salads, uncooked foods, unpeeled fruits and ice cubes in drinks. Sterilization of drinking water with 2% tincture of iodine (0.5 mL/L of water and allow to stand for 30 minutes) may be necessary. Treatment of asymptomatic cyst excreters is worthwhile, particularly in a non-endemic area, as it reduces the risk of transmission to others. 1

CYCLOSPORIASIS Cyclospora cayetanensis is a coccidian protozoan parasite which has been recognized as a cause of acute and persisting diarrhoea. The organism is an intracellular pathogen that infects the enterocytes. It has a complex lifecycle involving sexual and asexual forms, similar to that of Toxoplasma gondii. Infection occurs by ingestion of oocysts. It causes an acute small bowel type of diarrhoea with features suggesting malabsorption. It is not associated with systemic upset, dysentery or a faecal cellular exudate. It is likely to have a widespread geographic distribution

9

Second choice Metronidazole 2g single dose with evening meal for 3 days Repeat 1 week later Same precautions Third choice Mepacrine 100 mg tds for 7 days Causes abdominal discomfort: stains the skin yellow Contraindicated in patients with psoriasis If the above fail, try albendazole 400 mg bd for 5 days

throughout the world. The diagnosis is made by finding the oocysts, which contain four sporozoites in the mature form, in faeces which stain a pinkish colour with a modified Ziehl-Neelsen method. Treatment with co-trimoxazole, two tablets twice daily for 7 days, is effective.

CRYPTOSPORIDIOSIS Cryptosporidiosis is an infection of the gastrointestinal tract with the sporozoan parasite Cryptosporidium. It causes enteritis in humans and a range of other animals. It is a zoonosis with a worldwide distribution. Its life cycle is complex, with asexual and sexual cycles. Infection is by the faecal-oral route. Ingested oocysts are disrupted in the intestinal lumen, liberating four sporozoites which enter the apical regions of intestinal cells within a membrane-bounded vacuole. Trophozoites develop from the sporozoites and undergo asexual reproduction. Merozoites liberated by schizogony can undergo further asexual cycles, or become gametocytes which fuse to form oocysts which are passed in the faeces. Transmission and epidemiology Many reports recognized this organism as a gut pathogen causing persistent watery diarrhoea in immunosuppressed patients, especially those with AIDS. Other studies showed that Cryptosporidiosis was associated with diarrhoea in 4-9% of otherwise healthy children and adults. Children at nursery school may transmit infection person to person. Susceptible animals include calves, lambs, goats and birds, although a history of contact with animals is often absent. Clinical features An acute watery diarrhoeal illness, often with offensive

337

stools, is the usual history. There may be fever and vomiting. The disease is self-limiting, with a median duration of about 12 days, but prolonged diarrhoea with weight loss occurs in patients with humoral and cellular immunodeficiency states, especially AIDS (p. 446). Diagnosis and management Faecal smears stained by the modified Ziehl-Neelsen technique show the pinkish-red-staining oocysts. Other forms of the parasite are seen at the luminal surface of enterocytes in intestinal biopsies. In most patients the disease is self-limiting and symptoms resolve spontaneously in 10-12 days. There is no effective drug therapy for the chronic infection seen in AIDS patients though diarrhoea may regress with the institution of effective anti-retroviral treatment. Rehydration, nutritional support and symptomatic treatment may be needed. Handwashing is the most important factor in limiting the spread of infection by the person-to-person route. FIG. 9.25 Scanning electron micrograph of Entamoeba histolytica

AMOEBIASIS

(Reproduced by kind permission of Dr David Warhurst.)

The different manifestations of amoebiasis have long been a paradox. The following have been recognized:

not with non-invasive strains has been developed. These findings give support to the view that there are two separate species, one consistently a pathogen, E. histolytica, and the other a commensal, E. dispar.

• Non-invasive, asymptomatic and confined to the lumen of the colon, with amoebic cysts in the stools; • Invasive ulcerative colonic disease (amoebic dysentery) with amoebic trophozoites in stools and scrapes from the ulcers; • Invasive, causing amoeboma; • Invasive, causing amoebic liver abscess (ALA). Hitherto all these forms have been ascribed to infection with Entamoeba histolytica, but research over the past two decades has produced evidence for the existence of two separate species accounting for invasive and non-invasive conditions, respectively. Aetiology Entamoeba histolytica has two forms, a motile trophozoite (Fig. 9.25) and an encysted trophozoite. The motile trophozoite is found in the stools of a person with amoebic dysentery and contains ingested red cells. The cyst is the infective form. It is resistant to the chlorine in potable domestic water supplies. Sand filtration is the best method of removing cysts from water. Isoenzyme analysis of strains from patients with invasive disease and persons who excrete cysts only has shown two separate patterns that segregate invasive from non-invasive organisms. Subsequently a DNA probe which hybridizes with invasive amoebae but

1

338

Figs 9.33, 9.34

2

Distribution and incidence Amoebiasis has a worldwide distribution. It is common in those areas of the world where standards of sanitation are low. Most instances of invasive disease occur either in endemic areas or in travellers who have recently been exposed. Occasional cases are seen in people who returned from endemic areas years before, or who have never left a country such as Britain. Transmission and epidemiology Faecal-oral spread by ingestion of amoebic cysts in contaminated food and water, or on fingers or other direct personal contact, is the route of infection. With the new classification of amoebae noted above we now need information about the prevalence of Entamoeba histolytica, particularly the frequency of excretion of cysts. People excreting cysts are the source of infection to the community. There is an ELISA-based technique for detecting E. histolytica. Asymptomatic amoebiasis (E. dispar) is not uncommon in male homosexuals in Britain with transmission by person-to-person spread. It is also common in the tropics and subtropics. Amoebiasis is more common in adults than children. Men are affected by ALA more often than women, but dysenteric disease occurs in both sexes. Invasive amoebiasis can be more severe in pregnancy.

Fig. 9.35

Pathology and pathogenesis There are no pathological changes in E. dispar cyst

excreters. Their amoebae live on the colonic mucosa but do not invade. There is now a need to review the pathological findings in asymptomatic E. histolytica cyst excreters. These individuals maintain infection in communities without apparently being diseased themselves. In amoebic dysentery trophozoites invade the colonic epithelium and lyse host cells, causing necrosis. Areas of the mucosa are excavated, leaving multiple flask-shaped ulcers. 1 Amoebae are found in the slough and in the advancing edge of the lesions. The rectum and caecum are commonly affected areas, but the extent can vary from localized proctitis to a pancolitis with a diffusely inflamed ulcerated mucosa. 2 The mucosa over a variable length of the colon is occasionally sloughed off and passed per rectum, leaving exposed denuded submucosa. Amoeboma is an uncommon late complication that occurs after amoebic dysentery has resolved. The lesion consists of a mass of granulation tissue containing few amoebae. The caecum is a usual site, but any part can be involved and lesions may be multiple. ALA results from the haematogenous spread of amoebae from the colonic mucosa via the portal vein. Amoebae reach the liver, and on contact with amoebae liver cells are lysed. The affected area enlarges progressively and the necrotic cells liquefy in the abscess cavity, which may be unilocular or multilocular. A single abscess in the right lobe is usual. The pus is typically odourless, pinkish-brown, sterile on culture and with few pus cells. A carbohydrate containing lectin on the surface of amoebae helps attachment to epithelial cells. The mechanism of cell damage then appears to be the result of insertion of a large permanent ion channel into the host cell membrane, and the host cell dies rapidly. This ion channel is referred to as an amoebapore. Clinical features No symptoms are attributable to the commensal amoeba. In patients with amoebic dysentery symptoms vary according to the extent and site of colonic involvement. The usual symptoms are abdominal discomfort and increased frequency of bowel action, with softer stools. Blood may be noticed in the stools: sometimes flecks of blood or bloodstained mucus on the outside of stools, and sometimes overtly bloodstained stools. Physical signs usually consist of mild colonic tenderness. Sigmoidoscopy may show the typical appearance of scattered ulcers surrounded by mucosal erythema with normal mucosa intervening, but the mucosa may be diffusely inflamed and ulcerated. With increasing extent of colonic disease, symptoms and signs are more marked. When sloughing of areas of the mucosa occurs loss of protein-rich fluid from the denuded mucosa is marked. Acute onset of diarrhoea with systemic upset is almost certainly not due to amoebic dysentery, even if there is blood in the stools. It tends to have a more indolent onset with, in some patients, the development of more severe symptoms and signs as the extent of colonic involvement increases.

Dehydration, anaemia and septicaemia may occur in severe disease, and perforation, toxic dilatation and significant haemorrhage may complicate the clinical course. Rarely the superficial part of the mucosa may be sloughed over a variable length of colon and passed per rectum. Occasionally ulceration can extend from the anal canal to affect the perianal skin. Late complications of amoebic dysentery are colonic stricture and amoeboma, where a mass of granulation tissue forms in response to the presence of relatively few amoebae. Localized pain and blood in the stools are usual symptoms, and one or more masses may be felt in the abdomen or rectally.

9

Symptoms and signs in ALA Symptoms and signs in ALA relate to the site and size of the abscess. There may be no history to suggest past or present intestinal amoebiasis, and many patients do not have amoebic cysts in their stools. Fever, sweats, lethargy and anorexia are usual early symptoms and examination may reveal no abnormal physical signs, or only tenderness on springing the right lower ribcage. With larger abscesses pain becomes a more prominent feature related to the site of the abscess. Abscesses sited laterally or superiorly cause chest pain which may be pleuritic, or referred pain in the right shoulder indicating diaphragmatic irritation. Rigors occur with multiple large abscesses. Some patients have no systemic symptoms. The liver may be enlarged and tender, with a localized tender mass on the surface. With superiorly sited abscesses there may be dullness to percussion and impaired movement at the right lung base. Left lobe abscesses may present with symptoms and signs related to the epigastrium or left hypochondrium. Complications The main complication in ALA is extension outside the liver (Fig. 9.26). Spread to the brain may occur, producing a necrotizing abscess-like lesion. Rarely the skin of the chest or abdominal wall is involved after percutaneous rupture of ALA. Diagnosis Cyst passers are usually diagnosed when stool microscopy is carried out during screening after overseas travel or residence, investigation of diarrhoea due to some other microbial pathogen, or examination of other groups at risk of infection, e.g. male homosexuals. Serological tests for amoebiasis are negative. Antigen detection tests can be

SUMMARY 17 Amoebiasis • Cyst passers. Asymptomatic. Cysts in stools. • Amoebic dysentery. Diarrhoea, dehydration, anaemia. Motile trophozoites in faeces and scrapes from ulcers. • Amoebic liver abscess. Fever, pain, tenderness. Ultrasound shows abscess. Positive serology.

339

CASE STUDY 9.3 TRAVELLERS' DIARRHOEA

A 26-year-old man had travelled extensively in India over a 4-week period. He presented to an infectious disease unit in London 21 days after his return with increased frequency of bowel action that had begun in the third week of his stay. He was passing stools that were softer than normal and which for the previous week had contained obvious external blood. There was no abdominal pain, appetite and weight were normal and he experienced no systemic upset. Abdominal examination was normal. On rectal examination the mucosa felt slightly granular and a limited sigmoidoscopy showed an inflamed mucosa to 15 cms, with haemorrhagic areas and some obviously ulcerated areas. Questions 1. What is the differential diagnosis? 2. What further investigations should be performed?

FIG. 9.26 Sites of spread of amoebic liver abscess

1

340

Figs 9.36, 9.37

2

MCQ 9.24

Comment This history is in marked contrast to that of in Case study 9.1 (p. 279), which was invasive bacterial infection. The history is very suggestive of amoebiasis. Abdominal examination was normal. The endoscopic features are typical. Nonspecific inflammatory bowel disease (ulcerative colitis or Crohn's disease) could present like this and histological findings would be diagnostic. He could have postinfective gut symptoms and be bleeding from a colonic polyp, and so had amoebae or another source of bleeding not been found then colonoscopy would be the next step. Schistosoma mansoni affects the colon but only rarely causes rectal bleeding. This parasite occurs in Africa, the Middle East and South America, but not in India. In older patients presenting with rectal bleeding whether they have travelled to tropical areas or not colonic cancer always needs to be excluded

and colonoscopy is essential where no immediate source is found. A rectal biopsy was taken and a scraping from an ulcerated area and mounted on a slide in saline for immediate microscopy. A faecal smear was made from flecks of bloodstained mucus on the stool specimen provided, and microscopy of this showed motile trophozoites with ingested red cells. In the cellular exudate there were red cells but very few pus cells. The rectal scrape also showed motile trophozoites with ingested red cells. Histological examination showed an ulcerated mucosa with amoebic trophozoites in the surface layers. He was treated with tinidazole 2g as a single dose with food daily for 3 days, followed by diloxanide furcate 500 mg tds for 10 days.

used to identify E. histolytica cysts; most of those detected are E. dispar. Some cysts are, however, untypable using the currently available kit. Amoebic dysentery is diagnosed by finding motile trophozoites with ingested red cells in faecal smears or scrapes from ulcerated colonic mucosa. Amoebae can be seen in the edges of ulcers in rectal biopsies, where they are seen as large cells at the edges of ulcers and in slough. They are PAS-stain positive. Serological tests for amoebiasis (e.g. indirect fluorescent antibody test, or IFA) are positive in 60% of cases, and so diagnosis cannot depend on serology but rather upon finding the parasite. With amoeboma, cysts may be found in the stools. Colonoscopy allows the lesion to be visualized and biopsied. Serology for amoebiasis is positive in over 90% of cases and is a reliable test in this situation. Diagnosis of ALA depends on the clinical features suggesting an intrahepatic lesion and strongly positive amoebic serology (found in over 95% of cases). Cases seen early in the course of disease occasionally have a negative IFA but have a definitely positive test 7 days later. Additional investigations Intestinal disease With increasing extent of colonic involvement there is

increasing exudation of protein-rich fluid into the gut. Plasma albumin levels fall; normochromic, normocytic anaemia and a raised ESR develop. These investigations are normal in more localized disease. Plain abdominal X-rays are useful in giving an indication of the extent of the inflammatory process, indicated by the extent of colon devoid of faecal shadows. Barium enema studies show non-specific features with ulceration in dysentery. Amoeboma appears as one or more mass lesions causing narrowing of the colon, simulating colon cancer, and so if cancer is not demonstrated in biopsies from a stenosing lesion amoebic serology should be performed to prevent resection of something that would respond to metronidazole or tinidazole. Amoebic liver abscess There is neutrophil leukocytosis, a normochromic, normocytic anaemia and raised ESR. Bilirubin levels are usually normal, but aminotransferase levels are slightly raised; alkaline phosphatase levels are raised and plasma albumin levels are low. A chest X-ray will show elevation of the diaphragm above an abscess in the right lobe superiorly, and sometimes basal linear atelectasis or consolidation or pleural effusion. Ultrasound and CT scanning give more accurate information about an abscess, and its precise site and size. 1 Ultrasound can be repeated to assess changes and to guide the insertion of a needle for percutaneous aspiration. Differential diagnosis Intestinal disease Shigella, Salmonella, Campylobacter and Yersinia infections may all cause colitis, but the onset of symptoms is rapid in these infections and pus cells predominate in the faecal cellular exudate. Non-specific inflammatory bowel disease, Crohn's disease and ulcerative colitis must be considered. Amoebic infection should be rigorously excluded in a patient who could have been exposed to amoebiasis in the past, before treatment for inflammatory bowel disease, as steroids may promote rapid progression of invasive amoebiasis with perforation and death. Where the distinction cannot be made then treatment for both conditions may be appropriate. Carcinoma of the colon, pericolic abscess, Crohn's disease of the colon, tuberculosis of the ileocaecal region or colon, and schistosomiasis are possible causes of a colonic mass. Amoebic liver abscess Subphrenic and intrahepatic pyogenic abscess and infected hydatid cyst cause similar symptoms and signs. Often the history is more acute and the patient more toxic, with chills and rigors. Pyogenic liver abscess may be secondary to some other focus of intra-abdominal infection. Hepatocellular carcinoma and bacterial infection of a hydatid cyst may present like ALA, particularly when the carcinoma has undergone necrosis or there has been bleeding into it. Management The drug treatment of amoebiasis is set out in Table 9.33. Patients who can be shown to harbour E. dispar do not

TABLE 9.33 Drug treatment of amoebiasis Clinical situation

Drug

Cyst passer (E. histolytica)

Diloxanide furoate for 10 days 500 mg 3 times a day (8mg/kg3timesaday) Metronidazole 800 mg 3 times a day for 5-10 days (40mg/kg per day in 3 divided doses) or Tinidazole 2g (50mg/kg) single daily dose for 3 days (abscess) or 5 days, (dysentery amoeboma) followed by Diloxanide furoate 500 mg 3 times a day (8mg/kg3timesaday)

Amoebic dysentery Amoeboma Amoebic liver abscess

9

need treatment. Symptoms and signs improve rapidly within 48 hours. Surgical treatment of amoebic dysentery is rarely needed, except where large areas of mucosa are sloughed; in this case resection of the affected colon can be life-saving. When ALAs are very large, pointing or likely to rupture, percutaneous aspiration is needed. This is done using a large-bore needle inserted under local anaesthetic, if possible under ultrasound guidance. Variable amounts of fluid can be removed. Amoebae are sometimes found in the material aspirated from the edge of the abscess. Aspiration has a low morbidity. Rupture of ALA dramatically increases the mortality, which is low in uncomplicated cases. Prevention and control The provision of clean water and safe disposal of faecal waste would do much to prevent transmission. Patients with invasive amoebiasis should always receive diloxanide furoate to eradicate any pathogenic amoebae remaining in the gut lumen. Where water supplies are not treated, boiling and filtering drinking water is advised. Tincture of iodine (2%) is effective in killing amoebae, used as above. Salads, unpeeled fruits and ice cubes are possible sources of infection and should be avoided in endemic areas. 2

BLASTOCYSTIS HOMINIS The pathogenicity of this organism is questioned. There do appear to be some infected patients whose diarrhoea settles with eradication of the organism after a single dose of tinidazole 2g on two occasions a week apart. However, there are also patients excreting B. hominis whose diarrhoea persists after eradication, and in others the organisms persist despite resolution of the diarrhoea. The organism is very commonly found in healthy individuals with no gut problems.

341

BALANTIDIUM COLI INFECTION The ciliate Balantidium coli is the largest protozoan parasite in humans. It can be seen with the naked eye in cultures. It is a parasite of pigs, and humans who have close contact with pigs become infected. In most cases it causes no disease. Occasional cases of enteritis due to B. coli affecting the small and large intestine are reported. Death occurs as a result of bacterial peritonitis after ulceration and perforation of the gut. Metronidazole is effective in symptomatic cases.

Trichomonas infection is discussed in Chapter 11 (p. 462).

ISOSPORA BELLI INFECTION

NAEGLERIA FOWLERI INFECTION

Isospora belli is a sporozoan parasite of the gastrointestinal tract in humans. It has a worldwide distribution and humans are the only reservoir of infection. The life cycle has asexual and sexual phases. Infection occurs by ingestion of oocysts containing sporozoites, which are released in the small intestine and invade epithelial cells; there they undergo asexual reproduction, or schizogony. The host cell ruptures, releasing merozoites which invade other epithelial cells and reproduce asexually to form oocysts. These are released into the gut lumen by the normal desquamation of epithelial cells and passed in the faeces. Infection occurs by ingesting oocysts in contaminated food and water. Most infections cause no symptoms. Some cause a diarrhoeal illness, with weight loss, pale stools, lethargy and low-grade fever. Patients with immunodeficiency syndromes, including AIDS, exhibit chronic diarrhoea with more marked symptoms. The diagnosis is made by finding oocysts in stools, in jejunal juice, and in stained sections of jejunal biopsies. Co-trimoxazole, two tablets twice daily for 10 days, appears to be effective. Longer courses of treatment may be needed in patients with immunodeficiency.

Naegleria fowleri is a free-living amoeboflagellate, 6-9 mm in diameter, which causes the rare primary amoebic meningoencephalitis. It is found in water from hot springs as well as in ponds and pools, and has a worldwide distribution. Amoebae in fresh water are carried into the nasal passages by activities such as swimming. They adhere to areas of the nasal mucosae and then migrate through the cribriform plate along the nerve fibres. The amoebae replicate asexually and spread through the subarachnoid space. The olfactory bulbs show massive invasion with amoebae and necrotic changes. There is a superficial encephalitis of the affected area. Abscess formation can also occur. In the meninges there is a prominent fibrinous and purulent reaction that is most marked over the inferomedial regions, with particular involvement of the basal subarachnoid cisterns.

SARCOCYSTOSIS

342

of muscle cysts, and thorough cooking of meat prevents infection.

The two parasites of humans are Sarcocystis bovihominis and S. suihominis, which have cattle and pigs as their respective intermediate hosts. They have a worldwide distribution, with high infection rates in cattle and pigs. However, the infection is probably uncommon in humans. The parasites have a complex lifecycle involving both asexual and sexual reproduction. Abdominal discomfort, distension, nausea and diarrhoea occur within hours of infection, and sporocysts are found in the stools 14-18 days later. Diagnosis is by finding the small numbers of cysts shed in the faeces. Treatment with pyrimethamine and sulfadiazine or nitrofurantoin has been successful in patients with chronic symptoms. Freezing meat reduces the infectivity

GENITAL TRACT PROTOZOA

FREE-LIVING PROTOZOA CAUSING HUMAN DISEASE

Clinical features Patients are often children or young adults. Fever, headache, nausea, vomiting and lethargy are usual early features, followed by deterioration of conscious level. Neck stiffness is usual. Papilloedema may be present. Lumbar puncture in shows features suggesting acute pyogenic meningitis. The CSF is turbid, with a very high neutrophil count. Protein levels are very high and sugar levels are reduced or normal. The amoebae are readily seen on phase contrast microscopy from a direct smear of CSF. They can be grown in special culture media. Management Amphotericin B for 6 weeks at least, depending on the response, is effective in treatment, possibly in combination with miconazole. The high mortality in this condition is probably related to delay in diagnosis and initial trials of antibiotics as for pyogenic meningitis.

ACANTHAMOEBA INFECTION Acanthamoebae are free-living amoebae commonly found in soil and water in tropical and temperate regions. Infec-

tion may occur by ingestion or inhalation of cysts, which are very light and may be distributed on the wind. A considerable range of clinical features have been reported for this infection. Cerebral abscess, granulomatous meningoencephalitis, meningitis and involvement of the scalp, orbit and middle ear have occurred. Acanthamoebae have most often been identified in corneal ulcers, and the increased use of contact lenses has perhaps increased the frequency of amoebic keratitis. The amoebae are difficult to kill. They have been known to cause a radial keratitis, with spread of the organisms alongside nerve fibres, seen on confocal microscopy of the cornea. Lung involvement has also been reported. Amoebae may be identified in smears from CSF or from corneal ulcers, or in stained tissue sections. A modified Field's stain is effective in showing the amoebae. Amphotericin B is used in systemic infections. 0.1% hexamidine isethionate, dibrompropamidine isethionate and chlorhexidine have all been used with varying degrees of success for corneal lesions. There appears to be little correlation between in vitro and in vivo activity of compounds.

NON-PATHOGENIC GUT PROTOZOA There are a number of protozoa that infect humans but cause no disease. These are listed in Table 9.30. When they are found in patients with abdominal symptoms, the symptoms should not be ascribed to these parasites and further investigations should be undertaken.

TISSUE AND BLOOD PROTOZOA TOXOPLASMOSIS Toxoplasmosis is caused by Toxoplasma gondii. It is a zoonosis, the definitive host in the lifecycle being cats (Fig. 9.27). Primary human infection is acquired by ingestion of oocysts derived from animal excreta or from undercooked meats such as pork or mutton, the pig or sheep having acted as an intermediate host. Infection is common, 20-40% of adults in Britain having antibodies. The most serious effects are seen after congenital infection or in immunocompromised patients. Pathology and pathogenesis Trophozoites released at the site of entry multiply and spread via blood and lymphatics to all tissues. They can invade any cells, and intracellular proliferation leads to cell death. Development of immunity results in restriction of proliferation and the organism is eliminated, or forms tissue cysts. Cysts 10-200 mm in diameter, containing several thousand organisms, may remain in the tissues for life. Depression of CMI results in reactivation.

9

FIG. 9.27 Transmission of toxoplasmosis Infection can occur in either sex and at any age. Reactivation causes disease in immunosuppressed patients.

Clinical features Congenital infection Congenital infection usually follows primary asymptomatic infection of the mother during the third trimester of pregnancy. The effect on the fetus may be spontaneous abortion, stillbirth, premature birth, or normal birth with subsequent illness. The latter is most frequent and results in disease of the eye, usually bilateral choroidoretinitis, or of the brain, with hydrocephalus, microcephaly, cerebral calcification or convulsions. Fever, lymphadenopathy, hepatosplenomegaly and rash may also occur. Acquired infection Asymptomatic infection is common in adults, but there may be a mild chronic febrile illness with lymphadenopathy. Often a group of enlarged lymph nodes in a single area is the only finding, although more diffuse adenopathy may occur with splenomegaly. Sore throat does not occur. Choroidoretinitis is now recognized as a manifestation of acquired infection. Reactivation in the immunosuppressed patient Most toxoplasmal disease in patients with deficient immunity is caused by reactivation of latent cysts. The most common manifestation, especially in AIDS, is necrotizing lesions in the brain, often presenting as intracerebral mass lesions. Primary infections in the immunocompromised host cause disseminated infections. This has been recognized in cardiac transplantation when a seropositive donor is the source of a heart for a seronegative recipient. This is prevented by chemoprophylaxis. Myocarditis and possibly pneumonitis also occur. Diagnosis This is by serology. The Sabin-Feldman dye test, indirect haemagglutination or fluorescent antibody test (FAT) are used to measure IgG antibodies, and an ELISA is also available for measurement of IgM antibodies. Histology of affected lymph nodes shows reactive changes which are

343

not specific for toxoplasmosis. The organisms are rarely seen in tissue sections. Diagnosis of cerebral lesions is by CT or MRI scan. 1 A/loncrgfemenf

Treatment of acquired infection in normal adults does not affect outcome and is rarely required. Severe infections are treated with the synergistic combination of pyrimethamine 50 mg daily and sulfadiazine 1.0-1.5g every 6 hours by mouth for 3 weeks, together with folinic acid 15 mg daily, as pyrimethamine is a folic acid antagonist and can cause bone marrow suppression. Clindamycin 600 mg every 6 hours orally or i.v. can be used as an alternative to sulfadiazine in the event of toxicity. Patients with AIDS may need a longer course of treatment and the cerebral lesions are monitored by CT or MRI. These patients also require maintenance therapy with pyrimethamine 75 mg twice per week and sulfadiazine 1.5g twice daily once a satisfactory response has been achieved. Tissue cysts are not killed by this treatment, but they may be susceptible to azithromycin or atovaquone. 2

PNEUMOCYSTIS CARINII INFECTION

sive interstitial shadowing. The diagnosis should be suspected in immunosuppressed patients, especially those with AIDS, and confirmation is obtained from examination of bronchial washings obtained at bronchoscopy or in sputum induced using nebulized hypertonic saline. Management The treatment is adequate oxygenation and ventilatory support when necessary, and chemotherapy with either high-dose co-trimoxazole (four tablets four times a day) or pentamidine given by intramuscular injection or slow intravenous infusion in 250 mL of normal saline or 5% dextrose over 1-2 hours. Corticosteroids are added in moderate or severe cases. Fever, rash and leukopenia have been seen in AIDS patients after about 7-10 days on co-trimoxazole, and a change to pentamidine may be necessary, as treatment for at least 21 days is required. Pentamidine side-effects include renal and hepatic failure and prolonged hypoglycaemia. There is a 20% mortality in AIDS patients in their first episode of Pn. carinii, irrespective of treatment. Prophylaxis with co-trimoxazole for this infection has been effective in patients with acute leukaemia. Primary prophylaxis in AIDS patients with the combination of an antifolate and a sulphonamide/sulphone is of value, and nebulized pentamidine has also been used successfully.

Pneumocystis carinii is a common organism that is found worldwide. Although previously classified as a protozoan, evidence suggests that it may be fungal in nature. It infects the lungs and causes no problems in normal human hosts. The organism is found in autopsy material from healthy people in up to 4% of cases, and serological studies in Europe have shown up to 75% seropositivity in healthy children. The reservoirs of infection and the mode of transmission are poorly defined, but it seems likely that personto-person droplet infection occurs. It causes interstitial pneumonitis in premature and young infants, in children with protein-energy malnutrition, in immunosuppressed patients during treatment for haematological malignancies, and in AIDS patients in the west. It is uncommon in AIDS in Africa. In normal lung the organisms are found singly or in clusters on the alveolar septal wall, with no surrounding host response. When there is pneumonitis large numbers of parasites are found in the alveoli, with desquamation of alveolar cells and accumulation of proteinaceous material in the alveolar spaces. Clinical features and diagnosis The clinical features of Pn. carinii pneumonia are progressive dyspnoea, cough, fever and weight loss. Chest examination is often normal; chest radiographs may be normal initially but progress fairly rapidly to show exten-

1

344

Fig. 9.38

FIG. 9.28 World distribution of malaria, 2000 Chloroquine-resistant malaria occurs in all areas.

9

FIG. 9.29 Malaria in the UK, 1981-2000, showing the progressive rise of falciparum importations (Communicable Diseases Surveillance Centre, Colindale and Malaria Reference Laboratory, PHLS.)

MALARIA Malaria is an acute febrile illness characterized clinically by paroxysms of fever, the consequence of rupture of infected red cells due to asexual reproduction by species of Plasmodium (schizogony). Plasmodium vivax, P. ovale and P. malariae are associated with morbidity but no major mortality, but P. falciparum causes both morbidity and considerable mortality. The infection is transmitted by the bite of the female anopheline mosquito. Distribution and incidence Malaria occurs widely throughout the tropical areas of the world, in the Americas, Africa, Asia and the Pacific area (Fig. 9.28). Falciparum malaria is particularly common in tropical Africa, where it causes at least 1000000 deaths per year, mainly in children. The resurgence of malaria in the Indian subcontinent was led by a rising incidence of vivax malaria in the 1970s, and now falciparum infection is more widespread in the region. Ovale malaria is predominantly a west African disease. Malariae malaria is the least common form. Malaria is imported into temperate regions by tourists, people employed overseas, business travellers and immigrants. The numbers of cases have risen progressively over the past 20 years (Fig. 9.29). Transmission and epidemiology Life cycle The life cycle of malaria is shown in Figure 9.30. Vivax and ovale life cycles are similar, with primary exoerythrocytic schizogony (EES) in hepatocytes leading to infection of the peripheral blood with merozoites. These enter red blood cells (RBCs) and undergo erythrocytic schizogony (ES) every 48 hours; this is benign tertian and ovale tertian malaria. Some of the sporozoites produce latent forms, the hypnozoites, within liver cells, which produce EES and then ES up to 2 or 3 years after infection, i.e. relapsing malaria.

FIG. 9.30 Life cycle of malaria

Falciparum malaria has no hypnozoite form, and so the infection is cured when parasites are cleared from the blood by treatment. ES has a periodicity of less than 48 hours ('subtertian'). Malariae parasites also lack the hypnozoite stage but can cause reappearance of parasitaemia (parasites in peripheral RBCs) up to 20 or more years after infection. Small numbers of parasites persist in RBCs to

345

cause this. The periodicity of ES is 72 hours (quartan malaria). Vivax, ovale and malariae parasites invade 1-2% of RBCs at most. Falciparum parasites invade any proportion of RBCs, accounting for the severity of disease and the high mortality. Host A baby born of an indigenous mother in an endemic falciparum area will be protected against infection during the first year of life by maternal antimalaria IgG crossing the placenta in the last trimester of pregnancy. After the first year the child is fully susceptible. Without chemoprophylaxis the child experiences repeated attacks of malaria, and by the age of 4 or 5 years will have acquired protective immunity, which persists as long as they remain in the endemic area. Parasites are often found in the peripheral blood of an asymptomatic child, so that there is disease immunity without immunity to reinfection; indeed, reinfection is necessary to maintain antigenic challenge and the immune status. The immunity declines if an individual leaves the endemic area. Maternal immunity declines during pregnancy, particularly in primiparae, with transplacental transfer of IgG. Anaemia, fever and intense parasitization of the placenta make miscarriage, premature labour and low birthweight common, especially in areas such as West Africa, where there is heavy seasonal transmission with a high risk of infection. Where transmission is not so intense, effective immunity is not built up and all ages in the exposed population are at risk. Any individual, of any age, from a malaria-free area may contract severe malaria. Splenectomy enhances susceptibility to maleria to a considerable degree. Sickle cell trait haemoglobin C trait, and the heterozygous state for glucose-6-phosphate dehydrogenase (G6PD) deficiency protect against severe malaria. Vector Climatic factors have a profound influence on the transmission of malaria through effects on survival and reproduction of the mosquito population, and on the development of the parasite in the vector. Mosquitoes survive up to several months and their lifespan is not affected by malaria parasites. Ambient temperatures in the range 20-30°C, with a relative humidity of 60% or more, are ideal. In most areas of tropical Africa malaria is transmitted all year round, with upsurges of incidence with the dramatic increase of anopheline numbers during rainy seasons. In Asia transmission is seasonal with the rains. Sporogony will not occur below 16° or above 33°C. The vector species vary considerably in different localities. Anopheles gambiae is one of the most efficient vectors because of its long lifespan and preference for biting humans rather than other animals. 1

346

Fig. 9.39

Additional routes for transmission of malaria are: • Transplacental, which is uncommon • Transfusion associated, which is uncommon in Europe and North America but common in endemic areas • Syringe transmitted, among intravenous drug abusers. Mention should be made of 'airport malaria', which has been documented in a range of temperate countries, including the UK and France. Malaria can occur in people who have never been to an endemic area and who are not at risk as a result of transfusion, shared syringes etc. Such cases have occurred around airports, and it is thought that malarious mosquitoes that have been carried on international flights survive in the temperate country and bite individuals there, infecting them and causing disease. Pathogenesis and pathology The pathogenesis of falciparum malaria is complex and incompletely understood. The initial step is adhesion of the merozoite to the erythrocyte membrane. Glycophorin A, the major glycoprotein on erythrocyte membranes, is a receptor for binding specific surface proteins of falciparum merozoites. RBCs deficient in glycophorin A are resistant to invasion. The Duffy blood group antigen is a specific receptor for invasion by P. vivax merozoites, and the absence of this antigen among populations from west Africa explains the absence of vivax infections there. Changes in the parasitized RBC result in sequestration of RBCs containing mature falciparum trophozoites in the postcapillary venules. 1 Parasitized cells are less flexible than normal cells. In addition, as the parasite matures the RBC becomes deformed and knobs develop on its surface. These knobs are an expression of a parasite-derived adhesion-promoting molecule which binds to molecules such as ICAM-1, VCAM-1 and thrombospondin expressed on the vascular endothelium of the postcapillary venule. This adhesion phenomenon explains why mature trophozoites and schizonts are usually absent from peripheral blood films. Maximum numbers of adherent RBCs are found in venules in brain, liver, spleen, kidney and lung. The placenta is heavily parasitized with chondroitin sulphate A serving as a specific receptor for adhesion of infected RBCs. The trophozoite matures into the schizont and at schizogony the RBC membrane bursts, releasing merozoites, malaria antigen, malaria pigment and RBC cytoplasmic constituents. The cycle continues, progressively building up the numbers of parasitized RBCs. How these changes lead to altered consciousness in cerebral malaria is not clear. CT of the brain in patients with cerebral malaria has not shown oedema, although among children in Kenya intracranial pressure recording has shown raised pressure. Diffuse irreversible vascular obstruction is also an unlikely cause, in view of the normal cerebral flow and complete recovery without neurological deficit in most survivors. CSF lactate levels are increased in cerebral malaria, which may indicate some degree of anaerobic cerebral glycolysis. At the molecular level, high levels of TNF and other cytokines are found in the

blood of patients with the most severe forms of malaria, especially cerebral malaria. The extent to which these molecules cause severe malaria is not yet known. It is recognized that quantitative differences in cytokine production, particularly TNF, are genetically determined, indicating possible inherent predisposition to severe disease. Renal damage occurs because of prerenal and renal factors. Acute tubular necrosis is the usual effect of severe malaria on the kidney. This may occur both with and without severe intravascular haemolysis. Vessels in the heart are parasitized but cardiac function is well preserved. Reduced peripheral vascular resistance and dehydration from sweating, vomiting and reduced fluid intake may contribute to hypotension. In cerebral malaria post mortem the brain shows swelling with small haemorrhages throughout the white matter. The spleen is enlarged and has a slate-grey hue over the normal red colour. Centrilobular necrosis is seen in the liver with the accumulation of malaria pigment in Kiipffer cells. The pulmonary venules contain large numbers of parasitized RBCs. Pulmonary oedema of the ARDS type occurs but the cause is not understood. The placenta is usually heavily parasitized. The anaemia of malaria has several causes, which include rupture of parasitized RBCs in schizogony; sequestration of RBCs in tissue venules; destruction of parasitized and non-parasitized RBCs in the reticuloendothelial system (especially the spleen); haemolysis due to the presence of malaria antigen, antibodies and complement on RBCs; and marrow suppression. Abnormalities of coagulation are usual, with low platelet counts due to peripheral consumption and consumption of clotting factors. Disseminated intravascular coagulation does occur in a few patients with severe malaria but is probably not a major factor in pathogenesis in most severely ill patients. Clinical features Vivax and ovale • After an incubation period of about 13 days (vivax) or 18 days (ovale), prodromal symptoms begin with headache, fever, shivering without rigors, and general aches. These last for up to 3 days before the first paroxysm of coldness, then extreme heat, then defervescence with a profuse sweat. • Forty-eight hours later the full paroxysm occurs, with a feeling of extreme coldness and a rigor beginning in the late afternoon. Headache, nausea and vomiting are usually present. The temperature is high, the pulse rapid and low in volume, and the skin is cold. This phase lasts for 45 minutes to 1 hour. When the rigor ceases, peripheral dilatation occurs; the patient feels very hot and thirsty. The pulse is rapid and of full volume. The skin is hot and dry. This lasts about 1 hour and defervescence follows, with a profuse sweat. The symptoms settle completely and the patient will usually sleep. The following day there may be a little weakness. One day later the malaria paroxysm recurs. Daily paroxysms occur when

parasite broods are undergoing schizogony on successive days. • Untreated, the paroxysms continue for 6 weeks and die out spontaneously, only to recur 2-3 months later. By the time symptoms have been present for a week the spleen is usually palpable and there may be mild anaemia. Rupture of the enlarged spleen is reported in vivax malaria.

9

Malariae malaria The incubation period is usually about 28 days. Nonspecific prodromal symptoms last for 2-3 days before the onset of the first paroxysm, which is accompanied by a rigor. The periodicity of symptoms is every 72 hours. The spleen is enlarged when symptoms have been present for 7-10 days. Falciparum malaria • The incubation period is about 12 days, range 9-17 days. • Headache, anorexia, nausea, vomiting, weakness and fever are prominent symptoms. • The periodicity of symptoms is less than 48 hours, and periodicity is an unreliable clinical feature, being present in only 30% of cases. • The patient feels ill all the time, with exacerbation of symptoms at the time of paroxysms, which are similar to those described above but usually more severe. • Convulsions occur in children. Vomiting and diarrhoea are sometimes prominent features in the history. Herpes simplex vesicles may appear on the lips during the illness. Complications Complicated malaria occurs in falciparum infections. • Altered consciousness in the presence of falciparum malaria must be taken as a clinical indication of cerebral malaria. Neck rigidity is not a feature, although mild neck stiffness may be present. Raised intracranial pressure is not seen and focal neurological signs are uncommon. Generalized convulsions occur in children and adults. • Hypoglycaemia occurs in children, pregnant women with severe malaria and, less often, in adults with severe infection. • Jaundice may be a prominent physical sign in some patients with severe malaria. Haemolysis can cause this, but tender hepatomegaly and abnormalities of liver function suggest liver involvement. • Some degree of uraemia is common, but this resolves after treatment. Uraemia with oliguria indicates renal involvement. • Pulmonary oedema can result from overhydration, but can also occur in the patient whose infection is coming under control. • Blackwater fever, due to acute massive intravascular haemolysis, became less common after chloroquine replaced quinine for prophylaxis and treatment of malaria, suggesting that quinine itself contributed to pathogenesis. It does, however, occur in people who have not taken quinine. The urine is black, the haemoglobin

347

SUMMARY 18 Main features of malaria • • • • • • • • •

Fever Chills and rigors Headache Vomiting Diarrhoea sometimes Jaundice* Anaemia* Falling urine output* Passing black urine*

* Indicative of severe infection.

falls rapidly and jaundice appears. Hypotension and tachycardia are usual, and renal failure may follow. • Gram-negative septicaemia has been reported in patients with severe malaria and may be responsible for hypotension. Late complications Tropical splenomegaly syndrome (TSS) occurs in areas of endemic falciparum malaria. There is considerable splenomegaly, a lesser degree of hepatomegaly, anaemia and pancytopenia. Malaria antibody titres are high, with polyclonal elevation of serum IgM. Quartan nephrotic syndrome is related to P. malariae infections, and presents with gross oedema and proteinuria in children aged 4-6 years. The histological features are segmental endothelial proliferation with obliteration of capillary loops by periodic acid-Schiff (PAS)-positive material. There is segmental thickening of the basement membrane. Progression of the glomerular lesion leads to diffuse glomerulosclerosis. Immunoglobulins and complement are deposited in the glomerulus, and P. malariae antigen is detectable in 25% of biopsies. The prognosis is poor, with progressive deterioration in renal function. Treatment of malaria and chemoprophylaxis do not influence the course of established disease.

TABLE 9.34 Differential diagnosis of malaria Fever alone Any cause of pyrexia of unknown origin (see p. 274)

Fever and renal failure Septic shock Leptospirosis Viral haemorrhagic fevers

Fever and jaundice Yellow fever Typhus Leptospirosis Biliary sepsis Septicaemia

1

348

Fig. 9.40

Fever and impaired consciousness Meningitis Encephalitis Brain abscess

2

Case 9.4

Diagnosis The clinical diagnosis of malaria is based on the history of a febrile illness after exposure in an endemic region. Differential diagnosis is shown in Table 9.34. Cases in non-endemic areas present most often in the first 3 months after leaving the endemic area. Persons who have taken mefloquine as malaria prophylaxis may experience attacks of malaria up to 3 months after returning from the endemic area. Ovale and vivax infections may present later, and the latency in malariae malaria may be prolonged. The diagnosis is confirmed by the finding of parasites in stained blood films. If falciparum malaria is diagnosed, the percentage of RBCs parasitized should be determined and the film should be carefully examined for schizonts, which may be seen in the blood film from patients with severe malaria. Severe malaria is defined parasitologically as over 5% of RBCs parasitized, although complicated malaria can occur with fewer; clinical assessment is therefore very important. Because of binding of RBCs containing mature ring forms to vascular endothelium, the peripheral blood film in falciparum malaria is an uncertain guide to the severity of infection. Low initial parasitaemia can rise dramatically within hours when schizogony of sequestered parasites occurs. The blood is positive on the first occasion in most cases. The peripheral blood film gives only a guide to the severity of infection. If the parasitaemia is high the interpretation is straightforward but, if low, parasitaemia is not always indicative of mild malaria as many RBCs containing maturing parasites may be sequestered in deep venules, and so parasitaemia alone may not give a true picture of the severity of infection. A blood film should be made and examined as soon as the diagnosis of malaria is considered. 1 It is not necessary to wait for fever to occur. An antimalarial drug taken before blood films are obtained may clear the blood of parasites. If necessary, further blood films should be made and examined once or twice daily until the febrile illness either is diagnosed or resolves. During a period of observation in hospital, prophylactic antimalarial drugs should be stopped to give the best chance of demonstrating parasites. It may be necessary to start treatment as soon as blood films have been taken or on the basis of clinical diagnosis if laboratory confirmation is not readily available or the patient is very sick. Microscopy of stained blood films has been the gold standard for malaria diagnosis and in skilled hands is 100% specific and very sensitive, capable of detecting parasitaemia as low as 0.0001%. However, expertise is very variable and antigen detection methods for malaria diagnosis have been developed and can be used as an adjunct to microscopy. These depend on using monoclonal antibodies to various parasite epitopes. Compared with expert microscopy these methods have a 90% sensitivity in detecting parasitaemias in the range 0.01-0.001%, i.e. 50-500 parasites per microlitre of blood.

Laboratory features Laboratory investigations are normal or mildly abnormal in vivax, ovale and malariae malaria, but consistently abnormal to some degree in falclparum malaria. Haemoglobin levels may be normal or slightly reduced at presentation, but fall thereafter. Reticulocyte counts may be low before and after treatment, while haemoglobin levels are falling, suggesting a period of marrow suppression, followed by reticulocytosis with rising haemoglobin. Platelets are low in falciparum malaria, sometimes as low as 20 x 109/L, but other coagulation tests are usually normal. Occasional patients do have coagulopathy with spontaneous bleeding. Plasma sodium levels are low in falciparum malaria and urea may be raised. Globulin levels rise and albumin falls. Abnormalities in liver function tests are frequent, and there may be a minor elevation in both total and unconjugated bilirubin in mild infections or greater elevation in more severe falciparum. Azotaemia is common in falciparum malaria but oliguria is not. Elevation of urea and creatinine may reflect the hypercatabolic state of severe malaria, but urine output must be monitored as renal failure, most often due to acute tubular necrosis, is well recognized in severe malaria. Serological tests for malaria have no place in diagnosis of individual cases but may be helpful in assessing previous exposure. 2

9

Management Drug treatment Vivax, ovale and malariae. The aim of treatment is to stop further ES cycles (Table 9.35). There is usually a prompt response following chloroquine treatment of vivax, ovale and malariae malaria. Where radical cure of relapsing malaria is required G6PD levels should be checked and, if normal, primaquine given. When there is severe G6PD deficiency, primaquine should not be given because of the risk of severe haemolysis; chemoprophylaxis can be given to prevent relapse, e.g. pyrimethamine (25mg weekly). Primaquine is not recommended in pregnancy and a chemoprophylactic should be given. Falciparum. P. falciparum malaria is a serious infection demanding urgent treatment. Quinine is the drug of first choice for treatment because chloroquine resistance is now so widespread. The route of administration must be chosen carefully. When the patient is vomiting, is clinically ill with fever >39°C, is hypotensive or has specific complications of malaria, or has more than 2% of RBCs parasitized or schizonts in the peripheral blood, intravenous treatment is needed. The dosage regimens used are set out in Table 9.35. Impairment of renal or hepatic function will prolong

TABLE 9.35 Treatment of malaria Type of malaria

Drug

Dose

Additional treatment

Vivax and ovale

Chloroquine base

Malariae Falciparum Chloroquine-resistant or sensitivity unknown Mild

Chloroquine

600 mg (10mg/kg) stat, then 300 mg (5mg/kg) 6 hours later, then 300 mg (5mg/kg) daily for 2 days As above

Primaquine. 7.5 mg twice a day for 14 days (children 250ug/kg per day) Nil

Quinine salt

700 mg max (10mg/kg) 3 times a day for 7 days orally

or Malarone (atovaquone 250 mg + proguanil 100mg) or Mefloquine Quinine salt

4 tablets once daily for 3 days (reduced dose for children)

Fansidar. 3 tablets as a single dose for adults: reduced doses for children Nil

Severe

Known chloroquine-sensitive Mild Severe

Chloroquine base Chloroquine base

1,5g (125mg/kg) in 2 divided doses 6-8 hours apart Initial loading dose 1400 mg max (20mg/kg) in 250-500 ml of fluid in over 4 hours then 700 mg max (10mg/kg) by i.v. infusion in 250-500 ml fluid over 4 hours given 8-12-hourly until the patient can take oral quinine giving a total of 7 days treatment As above 300 mg (15mg/kg) given by i.v. infusion in 250-500 ml fluid, repeated 12-hourly Change to oral Chloroquine as soon as the patient's condition permits

Nil Fansidar as above

Nil

Note: (1) Loading doses of quinine should not be given to patients who have previously received quinine or therapeutic doses of mefloquine: in that case standard doses of i.v. quinine should be given. (2) Blood glucose levels must be monitored during intravenous quinine therapy as it can stimulate insulin release and hypoglycaemia.

349

CASE STUDY 9.4 COLLAPSE AND DYSPHASIA A 61-year-old man had travelled for 3 weeks in southern Africa visiting Zambia, Zimbabwe and South Africa. He had been in South Africa for a week before his return to UK. He had taken no antimalarial chemoprophylaxis. He became unwell 4 days before his return flight, with shiveriness, headache and anorexia. Towards the end of his flight from Johannesburg he collapsed, and on arrival was admitted to hospital. He was found to be very drowsy, with dysphasia and a divergent squint. He was febrile, 38°C, with no other localizing signs in the nervous system. Examination was otherwise unremarkable.

Questions 1. What is the differential diagnosis? 2, What further investigations would be needed?

It was thought that he had a stroke. Six hours later his sister visited and gave the information that he had just returned from an African holiday; blood films were then examined for malaria. These were positive, showing 11% asexual parasitaemia with schizonts, typical of falciparum malaria. Also on review deep jaundice was apparent, not noted previously. At this time his conscious level had declined further and he showed withdrawal responses to painful

stimuli. There were no retinal haemorrhages. Chest and cardiovascular examination was normal, BP 125/70. A urinary catheter was inserted and 59 mL of urine drained. Additional investigations showed Hbl2g/dL, with an increased white cell count at 15.3 x 109/L and low platelets, 13 x 109/L; coagulation screen normal; bilirubin 306 mmol/L, alanine transaminase 50 uL, albumin 37g/L, sodium 128 mmol/L, urea 27.6 mmol/L, creatinine 292umol/L.

Question 3. What Would be your immediate steps in management?

The patient was sedated, intubated and ventilated. Comment This case is typical of severe falciparum malaria, which very often occurs against a background of no prior chemoprophylaxis. The rapidity of progression of disease, particularly in the final hours before hospital admission, is in contrast to the much more indolent course of the vivax malaria case described below. The high level of parasitaemia is typical of falciparum, but the presence of schizonts in the peripheral blood is not usual and is indicative of more severe disease. Usually severe

the half-life of the two main schizonticidal drugs, and so the dosage regimen may need to be modified, reducing the amount given and, if necessary, the frequency of administration after 48 hours of medication. As soon as the patient's condition has improved sufficiently, oral treat1

350

MCQ 9.26

malaria is associated with high parasitaemias, but this is not invariable and severe disease can be seen in patients with less than 2%. This patient had cerebral malaria with renal failure due to acute tubular necrosis. Although platelets were low there was no bleeding. Management comprises termination of parasite replication and supportive care until the disordered physiology starts to return to normal. Great care is needed with fluid balance to avoid overhydration, with the risk of pulmonary oedema, and intraoesophageal Doppler is a valuable means of assessing fluid status, giving 200 mL boluses of colloid to optimize systolic ejection parameters. Quinine 1400mg was given as a loading dose over 4 hours, with 700 mg 12-hourly thereafter until all asexual parasites had been cleared, when a single dose of three Fansidar tablets was given. A six-unit exchange transfusion was carried out. Blood sugar was monitored during the quinine infusion. Haemofiltration was commenced. He was unconscious for 12 days before starting to wake up. Haemofiltration was required for 3 weeks. This man returned to his normal work. One of the concerns is how much irreversible cerebral damage occurs after cerebral malaria. In adults the clinical impression is surprisingly little, but 5-10% of children surviving this condition in East Africa do have residual cerebral abnormalities, e.g. cortical blindness, deafness.

ment should be given. Tinnitus, deafness and headache are common side-effects of quinine, and if they are very troublesome the frequency of administration of the drug can be reduced. Quinine alone is used in pregnancy (Table 9.35). Quinine alone will not cure falciparum malaria and is followed by a single dose of Fansidar (pyrimethamine plus sulfadoxine) as in Table 9.33. Doxycycline, 100 mg daily for 7 days, or lesser doses, is an alternative for older children and adults. Mefloquine, 20mg/kg up to a maximum of

1.5g, in two divided doses 6-18 hours apart, can be given as sole treatment. Malarone (a combination of atovaquone 250 mg plus proguanil 100 mg per tablet) is effective in mild falciparum malaria, giving four tablets as a single dose with food or a milky drink at the same time daily for 3 days. The problems of multiple drug resistance in strains of P. falciparum are maximal in southeast Asia. Compounds derived from Artemesia annua, qinghaosu, are very important in treatment. Although these drugs produce rapid defervescence and parasite clearance there is a relatively high recrudescence rate for monotherapy, and so to use them most effectively they must be given with another drug, for example artemether plus lumefantrine. Atovaquone and proguanil are combined in malarone (see above). It has been proposed that in order to try to maintain the efficacy of currently available antimalarial drugs combination therapy should become the rule, giving both selected drugs together from the beginning. Chloroquine is generally well tolerated by mouth; nausea and blurring of vision are transitory side-effects. Upper abdominal discomfort may occur after taking the drug. Pruritus is common in Africans taking chloroquine. This is not associated with a rash and is not a manifestation of allergy. Chloroquine can be used in pregnancy. The route of administration is determined by the severity of the illness, judged by the criteria detailed above. Additional measures Patients with severe or complicated malaria should be managed in an intensive care unit. Apart from effective malaria chemotherapy, management is essentially supportive. When the patient is unconscious intubation is necessary to secure the airway. Convulsions should be controlled with diazepam. There is no specific additional treatment. Dexamethasone is contraindicated and at present there are no grounds for recommending the use of mannitol. Hypoglycaemia has been reported in severe cases, both before and during treatment with quinine, particularly in young children and pregnant women. Quinine causes insulin release. Blood sugar should be monitored regularly and 50% glucose given i.v. as needed. Dialysis is needed when there is renal failure. Attention to fluid balance is essential to prevent fluid overload. Pulmonary oedema can occur because of fluid overload, but can also occur without evidence of heart failure - an adult respiratory distress syndrome picture. Mechanical ventilation may be needed when pulmonary oedema occurs. Transfusion is indicated when there is acute intravascular haemolysis and when the haemoglobin falls below 7.0g/dL. Fresh frozen plasma and platelet transfusions are given when there are major clotting defects requiring correction. Whole blood exchange transfusion, six units in adults, appears to be effective in reducing parasitaemia as well as removing plasma containing immune complexes, cytokines, malaria antigen and products of red cell schizogony in severely ill patients, and is indicated when more than 30% of RBCs are parasitized and at lower levels of parasitaemia when complications are present. Although

there are no controlled trials demonstrating its efficacy there is logic to it and it should be considered where supplies of adequately screened blood are available.

9

Management of other complications Quartan nephrotic syndrome has a poor prognosis, with progressive deterioration of renal function to end-stage renal failure. Fluid restriction, diuretics and albumin infusions may be helpful in relieving the oedema and fluid retention. The idiopathic tropical splenomegaly syndrome is treated with maintenance antimalarial chemoprophylaxis. With this, the splenomegaly regresses, haematological indices improve, and abnormal serological findings return towards normal. Cessation of chemoprophylaxis may cause relapse. Prevention and control For the traveller to endemic areas prevention comprises the avoidance of mosquito bites and chemoprophylaxis. Everything possible must be done to prevent mosquito bites (p. 272). With the increasing problems posed by drugresistant malaria, these measures occupy a central place in prophylaxis and their importance must be stressed to travellers and residents in endemic areas. This applies particularly for infants, children and pregnant women. Chemoprophylaxis in pregnancy prevents maternal anaemia, prevents the abortifacient effects of fever due to malaria and prevents intrauterine growth retardation. The use of permethrin-impregnated bed nets is very effective in reducing attacks. Permethrin has a residual insecticidal effects lasting 6 months before retreatment is needed. The drugs currently used for chemoprophylaxis are listed in Table 9.36 but up-to-date information should be obtained by those advising travellers to endemic areas. A malaria vaccine remains elusive. As indicated above, there is clear evidence for effective antidisease immunity in populations in endemic areas but this needs to be maintained by constant antigenic challenge. It is likely that any vaccine developed will comprise antigens from several stages of the parasite lifecycle, but the sporozoite stage seems to be important in stimulating immunity. Eradication is no longer perceived as an achievable goal and control is the current aim. Insecticide resistance among mosquitoes, particularly to the cheapest compound, DDT, and chloroquine resistance among falciparum parasites, have contributed to the problems of controlling malaria. The main methods used are: • Reduction of vector breeding sites • Larviciding pools that cannot be drained • Spraying insecticides inside houses and on other sites where mosquitoes rest • Treating cases in humans. 1

FURTHER READING ON MALARIA White N J et al. Averting a malaria disaster. Lancet 1999; 353: 1965-1967. World Health Organization. Severe falciparum malaria. Trans Roy Soc Trop Med Hyg 2000; 94: suppl 1.

351

TABLE 9.36 Prophylactic regimens against malaria in adults (Reproduced with permission of the PHLS Communicable Disease Surveillance Centre) Regimen Areas of chlomquine resistant P. falciparum: Mefloquine* Proguanil plus chioroquine Pyrimethamine-dapsone (Maloprim)+ plus chioroquine Doxycycline++ Atovaquone-proguanil (Malarone)** Areas without drug resistance: Chioroquine Proguanil

Dose for chemoprophylaxis

Usual amount per tablet (mg)

1 tablet weekly 2 tablets daily plus 2 tablets weekly 1 tablet weekly§ plus

250 (228 in US) 100 150 (base) 12.5(pyrimethemine) plus 100 (dapsone) 150 (base) 100 250 (atovaquone) plus 100 (proguanil)

2 tablets weekly 1 tablet daily 1 tablet daily

2 tablets weekly 2 tablets daily

150 (base) 100

All antimalarials to be avoided in severe hepatic and renal impairment. Chioroquine doses are given as the base. Folate supplements should be given to those taking proguanil or Maloprim when pregnant. * Avoid during pregnancy, during lactation and avoid pregnancy for 3 months after stopping it. Do not prescribe mefloquine if there is a history of epilepsy, depression or severe, psychiatric disorder. Appropriate for up to 1 year abroad. Caution in cardiac conduction disorders. + Best avoided in the first trimester of pregnancy. (Maloprim is only available for travellers who have epilepsy and cannot take alternatives.) + Contraindicated during pregnancy and lactation and for children less than 12 years. § Never exceed one tablet of pyrimethemine/dapsone (Maloprim) weekly. ** Licensed for up to 28 days in malarious area. Do not give in pregnancy. Begin 1-2 days before departure. Continue for 1 week after return.

FIG. 9.31 World distribution of trypanosomiases

matic infections also occur. Prodromal features include headache, malaise and gastrointestinal upset prior to the onset of fevers and rigors, with jaundice, haemoglobinuria and renal impairment in B. divergens infections. B. microti produces a more prolonged but less severe illness. Increasing age is associated with more severe clinical illness. The laboratory diagnosis is made by finding organisms similar to malaria parasites in Giemsa-stained peripheral blood films. Treatment with clindamycin, 600 mg 6-hourly intravenously or intramuscularly (20mg/kg/day in children), and quinine, 600 mg 8-hourly (10mg/kg 8-hourly in children) is effective in treatment of severe B. microti infections. For all severe cases intensive supportive therapy to maintain circulation, organ perfusion and oxygenation is essential. The efficacy of drug therapy in B. divergens is less certain.

TRYPANOSOMIASES African trypanosomiasis (sleeping sickness)

BABESIOSIS Babesia are intracellular protozoa that invade, divide asexually within, and finally rupture host erythrocytes. There are 70 species, affecting a wide range of vertebrates including cattle (B. diver gens) and rodents (B. microti). The parasites are transmitted by the bite of hard-bodied (ixodid) ticks. Babesia are widely distributed geographically, with reports of cases in humans from many countries, including Scotland and the USA. Babesia may occur concurrently with Borrelia burgdorfei infection. The incubation period is up to 4 weeks, less in patients who have had a splenectomy. There is a considerable range in the severity of infection, from mild to severe. Asympto-

1

352

Fig. 9.41

2

Fig. 9.42

Infection with Trypanosoma gambiense causes an initial febrile illness which progresses slowly to a chronic phase dominated by infection of the CNS, causing neuronal death and a variety of organic neurological syndromes. Neurological involvement occurs early in T. rhodesiense infections, within weeks of inoculation of parasites and, untreated, there is rapid deterioration and death. The two species cannot be distinguished on morphological grounds. Various species of tsetse fly, Glossina, are the vectors. Distribution and incidence The specific names of these parasites suggest a West African (T. gambiense) or East African (T. rhodesiense) origin for the infections, but there is a considerable overlap of the distribution of the two species in Central Africa (Fig. 9.31). The total numbers of cases are far fewer than malaria, but rhodesiense infections have a considerable economic impact as valuable farming and grazing lands

cannot be used because of the high risk of infection. There have been increases in both forms recently with imported cases seen in Europe.

Epilepsy occurs. In the terminal stage the patient is scarcely reusable, unable to swallow or feed him/herself. Bronchopneumonia is a common cause of death.

Transmission and epidemiology Tsetse flies become infected by ingesting trypanosomes in a blood meal. Once infected, a fly remains infective for life. About 1 % of flies in an area are infected. Humans are the reservoir in gambiense infections, whereas a variety of wild mammals, including bush buck and hartebeest, are the reservoir of rhodesiense infection, and humans are an incidental host. Transmission of gambiense trypanosomiasis occurs in wooded areas around rivers and streams that are water sources for local populations. Blood transfusion and transplacental transmission are also possible routes of infection.

Diagnosis Diagnosis depends on finding trypanosomes. Early in the course of infection with either species, parasites may be found in peripheral bloodfilms.0 There are usually many more parasites in rhodesiense than in gambiense infections. Repeated examinations must be made to detect the low levels of parasitaemia in gambiense infection. Gambiense parasites may be seen in bone marrow smears. Smaller numbers of parasites can be detected in smears of the buffy coat. Material aspirated from a lymph node can be smeared on a slide and stained. CSF must also be examined. The cell count is determined and the stained deposit examined for trypanosomes. An increased white cell count (>5/mm3) and/or an increased CSF protein (>250mg/L) indicates CNS infection in a patient with parasites in the blood or tissues. Serum immunoglobulin concentrations are increased, with IgM levels being very high. Increased CSF IgM indicates nervous system involvement. Serological tests do not have diagnostic value for individual patients.

Pathology and pathogenesis There is an early marked inflammatory response at the site of inoculation and in the lymphoid tissues of the body, lymph glands, liver and spleen. Trypanosomes may be seen in tissue sections. Rhodesiense infections progress rapidly to involve the heart and brain. Myocarditis and pericarditis are common. An obliterative endarteritis with vascular occlusion is found in the CNS, causing the degenerative changes. The neuropathological features include thickening of the meninges, which are infiltrated by lymphocytes and plasma cells. This infiltrate extends into brain substance in a perivascular distribution. The parasites continuously shed surface antigens which evoke antibody responses in the host, but the antigens change sequentially, a possible mechanism of evading host immune responses. Autoantibodies to brain and heart may cause tissue damage. Disseminated intravascular coagulation can occur in rhodesiense infections. Clinical features With both species, fever, chills and weakness are usual initial symptoms, beginning about 10 days after the bite. An inflamed lump is present at the bite site - the trypanosomal chancre 1 - and there is regional adenopathy. On white skins it is possible to see transient irregular areas of circinate erythema in both types of disease. Posterior auricular and posterior cervical adenopathy develop after 1-2 months of infection in gambiense infections. Unpleasant dysaesthesia follows squeezing of the calf muscles. Hepatosplenomegaly is often present in both infections early on. Jaundice occurs in rhodesiense infections. Purpuric rashes on the extremities are occasionally seen in severe rhodesiense infections. Neurological involvement appears within weeks of infection in rhodesiense infections, but it may be months or years later with T. gambiense. Indifference, lassitude, personality changes and altered behaviour are early features. Sleep rhythm is reversed. The basal ganglia, cerebellum and brainstem regions are most affected, producing tremor, incoordination, and problems with speech, balance and walking. Parkinsonian features may be prominent.

Management Suramin is used to kill parasites in blood and lymph nodes, and cures patients early in the disease. After an initial test dose of 100-200 mg given by slow intravenous infusion to ensure that the drug does not cause anaphylaxis or cardiovascular collapse, 1g doses are given by slow intravenous infusion on days 1, 3, 6,14 and 21. Pentamidine is the first-line drug for use in T. gambiense infections, giving 3-4mg/kg of pentamidine base daily by slow intravenous infusion or intramuscularly until 7-10 doses have been given. Sterile abscesses can occur at sites of intramuscular infection. Pentamidine stimulates insulin release from the pancreas, with the consequent risk of hypoglycaemia. Although pentamidine is used for early disease there is evidence that it crosses the blood-brain barrier and has worthwhile actions in early neurological disease in rhodesiense infections. Melarsoprol treatment regimens are complicated. Treatment is given intravenously. There is a significant incidence of arsenical encepholopathy with its use. For T. gambiense the drug is given in doses of 3.6mg/kg, maximum 180mg, for 3 days, repeating the series after a week without drug in patients with 6-19 WBC/cu mm CSF, and giving a third series for those with > 20 WBC/cu mm. For T. rhodesiense the regimen builds the dose up slowly from 0.36mg/kg (day 1), 0.72 (day 2), 1.1 (day 3), 1.8 (days 10,11 and 12), 2.2 (day 19), 2.9 (day 20), and 3.6 to a maximum of 180mg (days 21, 28, 29 and 30). The incidence of reactive arsenical encephalopathy is reduced by giving prednisolone 1 mg/kg before and throughout treatment. Eflornithine is effective in CNS disease due to T. gambiense with none of the arsenical toxicity, and should replace melarsoprol in this infection; initial treatment is with

9

353

400mg/kg/day in divided doses, 6-hourly for 14 days, followed by 300mg/kg/day in divided doses for 30 days. This amounts to a lot of drug! Examination of blood films, IgM levels, CSF and the patient's condition is helpful in assessing the response to treatment. Improvement in the patient's condition and resolution of the abnormalities indicate a good response. Relapse is indicated by a rise in the CSF protein or cell count, and further treatment should be given. When patients are treated early, complete recovery is the rule. In more advanced cases treatment arrests progression of the disease. Prevention and control Appropriate clothing in endemic areas is essential to minimize the area exposed to biting. Tsetses can bite through thin material. Pentamidine has been used as a prophylactic agent against gambiense infection in those heavily exposed. Efforts to control vectors by destruction of habitats and trapping flies may be helpful.

American trypanosomiasis (Chagas' disease) Aetiology American trypanosomiasis is an infection caused by T. cruzi, which exists in several forms during its lifecycle. The trypomastigote is found in the blood and has similar appearances to the African trypanosome. Amastigotes are found within host cells. Human blood forms are ingested by the vector species of reduvid bug, 1 when it takes a blood meal. The bug defaecates infective faeces on the skin after taking a blood meal, and the metacyclic forms enter the host through the bite or other skin abrasion, or through the conjunctiva if the faeces are carried to the eye on fingers. Distribution and incidence American trypanosomiasis is most common in the countries of South America among people living in poorly constructed housing, as the bugs live in roof thatch and cracked mud walls. Many people are infected in South America, with about 6000000 cases in Brazil, although not all those infected develop disease (see Fig. 9.28). Transmission and epidemiology Contact with infected bugs (the most important source), transplacental infection and blood transfusion are the routes of infection (Fig. 9.32). Infection in bugs is maintained from a reservoir of T. cruzi in dogs. Infection of children in the first decade of life is common. The bugs live in the thatch and cracked mud walls of basic village housing.

1 Fig. 9.43

354

4

MCQ 9.27

2

Fig. 9.44

3

Fig. 9.45

Animal reservoir (e.g. dog)

Proliferates in gut

TRIATOMID BUG HUMAN Trypomastigotes enter histiocytes and divide asexually

Cells rupture, releasing more trypomastigotes

Trypomastigotes released and invade other cells AUTONOMIC CELLS INVADED AND DENERVATION RESULTS FIG. 9.32 Lifecycle of Trypanosoma cruzi

Hundreds if not thousands of bugs can be found when these houses are demolished. Chagas' disease is essentially a disease of the rural poor. Pathology and pathogenesis Trypanosomes invade cells, including myocardial and gut smooth muscle cells. Rupture of the cell releases trypomastigotes and evokes inflammatory responses in the surrounding tissues. The trypomastigotes circulate, invade other cells, and replicate as amastigotes. Phagocytic cells in any tissue, as well as heart muscle, skeletal and smooth muscle and CNS, are among the tissues involved. Pericarditis and cardiomegaly are often present. A meningoencephalitis can occur in acute disease. The heart and the gut are the organs severely affected in chronic Chagas' disease. The heart is enlarged, with apical ventricular aneurysm commonly present. Thrombus is often present in the atrial appendage and in the ventricles. There is diffuse myocardial cell necrosis, with fibrosis and involvement of the conducting system. Parasympathetic plexuses of the gut are destroyed, causing dilatation of the intestine, frequently the oesophagus or colon (the mega-syndrome). Involvement of the conducting system of the heart is common and may be the only evidence of disease. Evidence of disease may come on at any time, even decades after infection. The pathogenesis of the disease is complex. There are several different strains of T. cruzi, some of which cause the cardiac and mega-syndromes. Other strains cause lifelong infection but no disease. Autoimmune mechanisms are probably involved in damaging tissues. Clinical features Acute infection may be inapparent in young children, but in older children and adults there may be an acute febrile illness. The patient may find a lump in the skin marking the site of the infected bite and initial parasite replication (the chagoma). Unilateral periorbital oedema indicates a conjunctival route of infection. Local adenopathy and hepatosplenomegaly are found.

Cardiomegaly, heart failure, arrhythmia and conduction defects indicate myocarditis in the acute illness. This may cause sudden death. Meningoencephalitis may occur. The local signs at the bite site, hepatosplenomegaly and fever subside in 6 weeks. Chronic Chagas' disease presents with heart and gut disease. Right- and left-sided cardiac failure are common. There may be signs suggesting pulmonary or systemic embolization from thrombus in the cardiac chambers. Conduction abnormalities are often present and may cause Stokes-Adams attacks and sudden death. Megaoesophagus causes dysphagia and aspiration pneumonia in the most severely affected. 2 Megacolon causes constipation and faecal impaction. 3 Volvulus of the sigmoid megacolon may occur. Diagnosis Acute infections can be diagnosed by finding parasites in the peripheral blood film. Trypanosomes may also be found on microscopy and culture of CSF, even when evidence of CNS involvement is lacking. Xenodiagnosis is also used to make a specific parasitological diagnosis. Laboratoryreared bugs, known to be free of infection, are fed on the patient, up to 20 bugs being used. They are then kept in the laboratory, dissected after 3 weeks and the hindgut examined for metacyclic forms. The polymerase chain reaction offers a valuable technique for detecting low levels of infectious agents and has been adapted for use in Chagas' disease. Serological tests are usually positive in both acute and chronic cases by indirect imrmmofluorescence, indirect haemagglutination and complement fixation techniques, and may be the only indication of infection. Management Treatment is unsatisfactory. Two drugs are available, nifurtimox and benznidazole. Both will clear the blood of parasites in the acute phase, but prolonged follow-up suggests that neither gives a radical cure. There is now evidence from a prospective randomized controlled study that benznidazole treatment, 8mg/kg body weight in divided doses for 3 weeks given to children with positive serology, will reduce the frequency of development of ECG changes during follow-up and, with time, be followed by a decline in antibody litres. In adults with positive serology nonrandomized studies have shown a reduction in the frequency of development of ECG changes during follow-up after benznidazole therapy. However, the extent to which these findings indicate a definite benefit for treatment is still uncertain. Control of arrhythmias and heart failure is all that can be done for cardiac disease in the chronic phase. Relief of obstruction at the oesophagogastric junction by dilatation or surgery may help the patient with megaoesophagus. A dilated segment of small or large intestine can be excised. Prevention and control Brick-walled houses with corrugated iron roofs would prevent the transmission of Chagas' disease. Residual

spraying of existing houses with y-benzene hexachloride is a useful method of vector control. 4

9

LEISHMANIASIS The Leishmania are obligate intracellular protozoan parasites that infect humans and other mammals, including dogs, foxes, gerbils and rats. Infection is transmitted by the bite of infected sandflies. The generalized lifecycle is shown in Figure 9.33. These organisms invade the mammalian host, parasitize cells of the reticuloendothelial system and evoke granulomatous responses in infected tissue. Genetic factors controlling the host cell-mediated responses determine the outcome of infection, with only a proportion of those infected developing clinically apparent disease. The geographical distribution is shown in Figure 9.34. Two forms of disease are recognized. Cutaneous leishmaniasis denotes infection with species of Leishmania that affect primarily reticuloendothelial cells in the skin, whereas visceral leishmaniasis denotes infection with a species that primarily infects reticuloendothelial cells of deeper tissues (Table 9.37). The species causing cutaneous disease are geographically distinct between the Old and the New World.

Cutaneous leishmaniasis Cutaneous leishmaniasis is a granulomatous infection of the skin caused by various species of Leishmania, resulting in one or more chronic ulcerating skin lesions (Fig. 9.35). In time, the ulcers usually heal with scarring.

Proliferation in gut

SANDFLY

Promastigotes regurgitated or contaminate mouth parts while feeding

Ingestiqn of parasitized cellwithamastigotes

HUMAN

Rupture of host cell. Other reticuloendothelial cells invaded

Reproduction by binary fission

FIG. 9.33 Generalized lifecycle of Leishmania

Promastigotes taken up by reticuloendothelial r.fills

355

FIG. 9.34 World distribution of leishmaniasis FIG. 9.35 Cutaneous leishmaniasis producing ulcerating skin lesions

(copyright S.G. Wright)

TABLE 9.37 Forms of Leishmania infection Location

Cutaneous leishmaniasis

Visceral leishmaniasis

Old World

Leishmania tropics Leishmania major

Leishmania donovani

Leishmania aethiopica Leishmania infantum Leishmania mexicana

Leishmania infantum

New World

Leishmania chagasi

Leishmania braziliensis Leishmania peruviana

Aetiology The organisms causing cutaneous leishmaniasis are listed in Table 9.37. Morphologically they are identical, but distinctions can be made using polymerase chain reaction (PCR) methods. Distribution and incidence The incidence is difficult to estimate, as many self-healing cases never come to medical attention. World Health Organization figures put the incidence at 100 cases per 10000 of the population in endemic areas (Fig. 9.31). Transmission and epidemiology Female sandflies of the genera Phlebotomus (Old World) and Lutzomyia (New World) transmit these infections by their bite. Apart from L. tropica (transmitted in towns and cities, with humans as the reservoir), these are zoonotic infections; humans are infected when they move into areas where there is a mammal-sandfly cycle. Urban infections tend to be more common among children. Pathology and pathogenesis Promastigotes change to amastigotes after inoculation into

1

356

Fig. 9.46

the skin and are phagocytosed by macrophages, where they replicate (Fig. 9.30). 1 The local granulomatous reactions produce a subcutaneous nodule and then the overlying skin ulcerates. The lesion persists as long as amastigotes proliferate within macrophages. This occurs despite the presence of sensitized T lymphocytes, shown by a positive delayed hypersensitivity reaction to a leishmanin skin test antigen. In most infections healing occurs spontaneously with time. It is likely that factors that contribute to TNF production, often genetic, determine the natural history of infection. Clinical features Persisting cutaneous ulceration on an exposed part of the body continuing for more than a month is the usual feature. The ulcer is painless and may crust over, only for the crust to separate again. There may be small nonulcerated satellite nodules around the primary lesion, and in both Old World and New World forms there can be involvement of draining lymphatic channels, which may be both palpably thickened and the sites of secondary lesions along their course. Lymph nodes on these lymphatics can be enlarged. Secondary bacterial infection is surprisingly uncommon. These infections will heal spontaneously in 2-8 months, leaving a residual scar. Recidivans leishmaniasis caused by L. tropica infection is characterized by a chronic relapsing course in spite of T-cell immunity. In contrast, diffuse cutaneous leishmaniasis causes numerous nodular or plaque-like lesions which tend not to ulcerate. These patients represent a kind of 'lepromatous' cutaneous leishmaniasis (p. 326). It is caused by L. aethiopica in Africa and L. mexicana in Central and South America. Leishmanin tests are negative. The South American species cause lesions similar to those of L. major (Fig. 9.35). L. braziliensis causes an initial ulcerating skin lesion which heals, but may subsequently produce a recurrence of disease with destruction of tissues at a mucocutaneous junction (mucocutaneous leishmaniasis). This can occur at the time of presentation with the primary skin

lesion, or months or even a year or so later. Another feature of cutaneous disease is the tendency to form lesions at the sites of chronic inflammation in the skin or at sites of trauma. For example, cutaneous leishmaniasis can develop in the perianal region, when there is recurrent perianal sepsis. Laryngeal disease due to leishmania has also been recognized. Diagnosis Parasites can usually be found in Giemsa-stained smears of material from slit skin smears, material aspirated from the edge of the lesion, or touch impression smears from the surface of a biopsy. Material from the biopsy can also be cultured in Novy-McNeal-Nicolle (NNN) medium. PCR on DNA extracted from biopsies or from cultured material can be used to make the diagnosis or to characterize the species complex involved. Serological tests for leishmaniasis are positive at low litre in most forms of cutaneous leishmaniasis. Differential diagnosis A variety of infective disorders can cause indolent ulceration. Among these are • • • • • • • • •

Yaws Tertiary syphilis Lupus vulgaris Histoplasma duboisii Diphtheritic ulcers (desert sore) Sporotrichosis Blastomycosis Malignant ulcers Tropical ulcers.

Diffuse cutaneous leishmaniasis can be mistaken clinically for lepromatous leprosy unless the causative organism is demonstrated in skin smears. Management The treatment of cutaneous leishmaniasis is unsatisfactory, not least because the commonly used drug, sodium stibogluconate, 20mg/kg, has to be given by daily intravenous injection for 20-30 days. Intralesional antimonials, in Old World disease, and y-interferon have also been used. Sodium stibogluconate (Pentostam) is the first choice of drug for both Old and New World leishmaniasis, and usually results in healing in most types of disease. Diffuse cutaneous leishmaniasis is often resistant and pentamidine is the alternative drug. Amphotericin-B is also very effective, given as a solution rather than in the liposomal formulation which is very effective in visceral disease (see below). Prevention and control Among Old World cutaneous leishmaniasis cure is usually associated with immunity to reinfection with the same strain, but vaccination is not yet available. Destruction of the burrows of rodent vectors can be successful in removing reservoirs. Removal of sandfly breeding sites around houses and spraying with insecticides will help to control

peridomestic transmission. Protection against the bites of sandflies may be helpful, e.g. mosquito nets and insect repellants, although the insects will bite through a shirt.

9

Visceral leishmaniasis (kala-azar) Visceral leishmaniasis, an infection caused by L. donovani, is spread by the bite of an infected female sandfly and can cause a prolonged febrile illness with wasting, hepatosplenomegaly and lymphadenopathy. In South America visceral disease is caused by L. chagasi, which is probably identical to L. infantum which causes visceral as well as cutaneous disease in the countries of the Mediterranean littoral. Distribution, incidence and transmission The areas affected are shown in Figure 9.31. This condition is endemic in countries of southern Europe and Mediterranean islands that are regularly visited by tourists. Infections acquired in these areas are seen in the UK. Leishmaniasis has been recognized as an opportunist infection in AIDS in southern European countries. A major epidemic has been continuing over 10 years in the southern Sudan. Children are commonly affected in the endemic areas. Subclinical cases probably exceed clinical cases by 5 to 1. Endemic, epidemic and sporadic forms of the disease occur. All ages and both sexes are susceptible to infection. The dog seems to be an important reservoir throughout the endemic areas. Humans are the reservoir for L. donovani in northeastern and eastern India. The vectors are female sandflies of the genus Phlebotomus in the Old World and Lutzomyia in the New. These insects, guided by scent plumes, bite at night. Pathology and pathogenesis Promastigotes transform to amastigotes when they are injected into the skin and ingested by phagocytic cells. Amastigotes replicate asexually within macrophages and are carried to lymph nodes draining the site of the bite; they then disseminate to macrophages in other tissues, particularly the liver, spleen, nodes and marrow. There is progressive enlargement of liver and spleen. Wasting of the muscles and low plasma albumin are metabolic responses to the infection. The malnutrition will impair immune responses to the parasite and to other diseases such as tuberculosis or pneumococcal pneumonia, which may complicate the patient's illness. Leishmanin tests (a delayed hypersensitivity skin test) are negative before treatment and become positive after cure. Clinical features The incubation period is very variable, from 2 weeks to 2 years or more. Latent infections are common and are particularly likely to give rise to disease in the context of AIDS. Early symptoms are lethargy, sweats, and fever, with a typical pattern of two peaks in a day. Within weeks of onset the liver and spleen are palpable and progressively enlarge. Body weight falls. Lymphadenopathy may be

357

found. Emaciation is obvious in more long-standing cases. Darkening of the skin of hands, feet, face, and abdomen is common in eastern India ('kala-azar' means 'black sickness' in Hindi). Poorly nourished patients should be carefully examined for coexisting tuberculosis or other infection, such as pneumonia. Investigation shows anaemia and pancytopenia due to hypersplenism, low plasma albumin and high globulin (e.g. 50g/L) with polyclonal elevation of IgG. Serological tests for leishmaniasis are positive in most cases. Parasites are readily found in patients with co-existing AIDS. Diagnosis The clinical features, together with exposure in an endemic area, suggest visceral leishmaniasis. The parasites are usually found on examination of Giemsa-stained smears of bone marrow aspirate, which should also be put into NNN culture medium. 1 Percutaneous splenic aspiration has a high yield of positive results (95%); it can be done safely in a cooperative patient whose prothrombin time is not more than 3 seconds prolonged and whose platelet count is 40 x 109/L or more. Again, material should be cultured and used for making smears. Management For many years sodium stibogluconate has been the drug

of first choice, giving 20mg/kg intravenously daily for 30 days, though longer courses may be needed. Liposomal preparations of amphotericin B are very effective, much less toxic than the conventional formulation, and can be given for shorter courses, 2-3mg/kg/day for 10 days, while 4mg/kg/day for 5 days plus an additional dose at day 10 is very effective. Relapse occasionally occurs; the options are to give more sodium stibogluconate for longer periods, or to give a second-line drug, e.g. pentamidine or amphotericin B. One of these is certainly needed if there is no response to a second course of sodium stibogluconate. Preliminary results with miltefosine, an anticancer drug, are promising. Although there may be good responses to treatment in patients with AIDS, relapse is likely as parasites are unlikely to be eradicated. However, there is little experience to date of antileishmanial treatment given with concurrent highly active antiretroviral treatment (HAART), and the improvements in CD4 counts seen with HAART may possibly allow parasite killing. Prevention and control Prevention is difficult. Where humans are the reservoir, finding and treating cases and insecticiding homes may help in control. Control of canine reservoirs can be of help. Vaccines are under study but not yet in clinical use.

HELMINTH INFECTIONS

A large number of helminthic parasites affect humans (Fig. 9.36). Humans are the definitive host in most instances, and the sexual phase of the lifecycle takes place in the human host. A classification is given in Table 9.38. Less commonly humans are an inadvertent intermediate host, infected either with larvae that do not develop any further because they are in the wrong host, e.g. Toxocara canis, or with larvae that are unlikely ever to be consumed by the definitive host, as with hydatid disease. The lifecycles of these parasites differ markedly. Faecal-oral transmission by ingested eggs or percutaneous infection by infective larvae are characteristic of the soilassociated gut nematodes. In contrast, the filariases have blood-sucking insect vectors in which they develop to infective larvae that will infect humans when later blood meals are taken. The lifecycles of the trematode platyhelminths are more complex, involving more than one intermediate host in many instances. The climatic conditions favouring transmission are warmth and humidity. These factors promote the survival of the soil-transmitted helminths and the insect vectors of

1

358

Fig. 9.46

TABLE 9,38 Classification of helminth infection Nematodes (roundworms) Human intestinal Ascaris Ancylostoma Necator Strongyloides Enterobius Trichuris

Filariai Onchocerca Wuchereria Brugia Loa loa

Animal intestinal Trichinella Toxocara

Trematodes (flukes) Schistosoma Clonorchis Opisthorcis Fasciola Fasciolopsis Paragonimus Cestodes (tapeworms) Human intestinal Taenia species Diphyllobothrium Hymenolepis

Larval tissue cestodes Taenia solium (cysticercosis) Hydatid (Echinococcus)

the filarial parasites. Local traditions in food preparation and consumption are important in the transmission of a number of parasites, such as eating raw meat or lightly cooked or raw fish and crabs. Among sheep and goat

9

FIG. 9.36 Example of helminths A Rhabditiform larvae of Stmngyloides stercoralis passed in faeces (copyright S.G. Wright). B Adult roundworms, Ascaris lumbricoides. The larger female is about 15cm long. C Taenia saginata, the beef tapeworm, after expulsion from the gut (copyright S.G. Wright).

herding communities where dogs are used, definitive and intermediate hosts of Echinococcus granulosus are brought together, and it is common for sheep and goat offal containing hydatid cysts to be fed to dogs. Dog faeces containing E. granulosus eggs contaminate the environment around houses, with the consequent risk of infection of humans, particularly children. It is apparent that the burden of infection is not evenly distributed in communities. There is a small group, often children, with persistent heavy infections, and it is among this group that the morbidity and any mortality of helminthiasis are seen. Studies have shown adverse effects on nutrition and growth, and there are associations between intestinal helminthiasis and impairment of intellectual development, although a causal link has not been shown. These effects have been well shown for Ascaris lumbricoides and for Trichuris trichiuria, particularly in the setting of the trichuris dysentery syndrome. Additional studies are extending these observations to consider a similar role for schistosome infection. In addition, the availability of ultrasound scanning and endoscopy has allowed recognition that migration of ascarids into biliary and pancreatic ducts is more common than hitherto realized. The filarial worms produce disease as a result of inflammation caused either by adults in main lymphatic channels, e.g. Wuchereria bancrofti, or by larval microfilar-

iae dying in the skin in Onchocerca volvulus infection. Additional morbidity in onchocerciasis is the result of onchocercal larvae dying in the cornea to cause the ocular complication sclerosing keratitis, which can progress to blindness. The chemotherapy of helminthic infections has become progressively simpler. The benzimidazoles - tiabendazole, mebendazole and albendazole - are broad-spectrum antihelminthics active against the soil-transmitted helminths. The incidence of side-effects decreases in the order given, the most marked side-effects occurring with tiabendazole, which should be avoided unless absolutely necessary. Praziquantel has actions in a range of cestode and trematode infections, with a low incidence of side-effects. It is the first drug to be effective in cysticercosis.

NEMATODES HUMAN INTESTINAL NEMATODE INFECTION Ascariasis Ascariasis is infection of the gut lumen with the nematode parasite Ascaris lumbricoides.

359

Aetiology The adult female worm measures up to 400mm long by 6mm wide at its thickest (Fig. 9.36B). The male is up to 300mm long by 2-4 mm. They live for about 1 year. Both are a greyish-white colour. The female lays up to 200000 eggs per day. The worms do not attach themselves to the gut mucosa but maintain their position by actively moving against the flow of intestinal contents. The eggs survive well in warm moist soil but are killed by heating to 50° C. Distribution and incidence Ascariasis is common in the tropics and subtropics, and up to 90% of people may be infected. Poor or absent sanitation and frequent faecal contamination of the environment with vast numbers of Ascaris eggs lead to a high frequency of infection, often with large worm loads in children. Transmission and epidemiology Eggs are ingested in food or water or from fingers, and the larvae are released in the small intestine. They enter the mucosa and then blood vessels to reach the lungs, where they leave blood vessels and enter air spaces. Here they undergo two further moults and then ascend the bronchial tree to the pharynx. They migrate down the oesophagus and through the stomach to reach the small intestine. An excess of infections is found in a small proportion of the population, the reasons for which are uncertain. However, among this group there is evidence of undernutrition, perhaps because of the effect of the worm bulk creating a sense of satiety, with consequent reduction in food intake. Pathology No intestinal mucosal lesions are attributable to the adult worms. Pulmonary infiltrates with eosinophils are found during the larval migratory phase of the lifecycle, but this rarely has a clinical correlate. Clinical features Symptoms due to the migrating phase of ascariasis are very uncommon. Pulmonary infiltrates on X-ray, eosinophilia and restrictive pulmonary function defects can occur. This can be seen when humans are infected with the pig ascaris, Ascaris suum. There are also few symptoms attributable to established infection. Worms may be vomited, or passed with the stools, or may even emerge from a nostril. Bolus obstruction of the small intestine occurs, particularly in children with very heavy infections. A mass of matted worms in a segment of gut can cause intestinal obstruction, volvulus and infarction. Occasionally an adult worm will migrate

1

360

Figs 9.47, 9.48

into the common bile duct, causing pain, obstructive jaundice and ascending cholangitis, or into the pancreatic duct to cause pancreatitis. Ascarids can also migrate through intestinal suture lines to cause leakage from an anastomosis. Ascaris egg granulomas are occasionally found in the peritoneal cavity, presumably related to ectopic development of an adult female.

Diagnosis Diagnosis is made by finding eggs in faecal samples. The complications of ascariasis are usually diagnosed on surgical exploration or endoscopic examination. Eosinophilia is uncommon, apart from at the stage of tissue migration. Management Mebendazole (100 mg twice daily for 3 days) or albendazole (400 mg as a single dose) is effective. Both drugs are relatively expensive. Piperazine hydrate elixir (750 mg of hydrate per 5mL) is effective. The dose depends on age and weight. Where a mixed infection of ascaris and hookworm is being treated with piperazine and tetrachlorethylene (TCE) for hookworm, it is usual to treat the ascaris first, as TCE may stimulate ascaris migration into biliary or pancreatic ducts. Prevention and control The use of latrines would prevent infection. Regular administration of drugs, e.g. albendazole, in affected communities is being used. Faecal contamination of the environment and the use of human faeces as fertilizer on crops encourage transmission. Before use as fertilizer, faeces can be treated with sodium nitrate or calcium superphosphate to kill the eggs of gut parasites. Education of the community about the control efforts, and their active cooperation, are essential to the success of control programmes.

Hookworm This is infection of the gut with either Ancylostoma duodenale or Necator americanus. Both grip the wall of the mucosa with their mouth parts and take a small amount of blood from the host each day. A relatively small number of people develop disease related to hookworm infection. Aetiology The worms are about 9 mm long. The female Ancylostoma produces 30000 eggs per day and Necator 9000 eggs per day. Eggs must be passed into warm, moist soil for the ovum to develop into the larval stage. The lifecycle is shown in Figure 9.37. Larvae of Necator grow in the lungs. Ancylostoma larvae can be ingested with food and develop in the gut lumen without tissue migration. Epidemiology Both species occur worldwide where temperature and soil conditions are suitable. Ancylostoma eggs are capable of resisting desiccation to some extent. Infection rates of 80-90% are common in endemic rural areas. The use of

Growth of larvae

Larvae penetrate alveoli, pass via the trachea to the pharynx and are swallowed

Larvae in upper small intestine, mature to adult stage

Venous circulation to heart and lungs Larvae penetrate skin HUMAN

Egg passed in taeces

SOIL Eggs mature and hatch in soil FIG. 9.37 Lifecycle of hookworm

untreated faeces as fertilizer on crops accounts for the high infection of farm workers.

(TCE), giving 0.5mL/kg to a maximum single dose of 5.0mL. Eradication of the worms is not the priority if anaemia is present. Packed cell transfusion may be needed after giving diuretics. Where there is heart failure, diuretics and cautious blood transfusion with packed cells are needed. Exchange transfusion is another way of correcting anaemia without risking the adverse effects of fluid overload. Later, iron stores will need replacing.

9

Prevention and control Safe disposal of faeces prevents transmission. Where night soil must be used to fertilize crops, prior chemical treatment kills ova (see Ascariasis, above). Wearing shoes or sandals also prevents infection. A diet with meat as a regular constituent would prevent anaemia, but in the tropics and subtropics this is often beyond the means of many families and so vegetables with a high content of ferrous iron, e.g. soya beans, should be introduced into the diet. Regular deworming is increasingly used.

Cutaneous larva migrans Pathogenesis and pathology Each worm takes a small amount of blood from the host. Nutrients are removed by the parasite and the residue passes into the gut, where it can be digested and absorbed. Iron deficiency anaemia occurs when iron losses exceed intake and iron stores are depleted. Anaemia may be severe, with haemoglobin levels of less than 5.0g/dL. Hypoalbuminaemia is usual at this stage, but malabsorption does not occur. Clinical features No symptoms are attributed to hookworm infection. Hookworm anaemia appears to cause few symptoms until oedema due to hypoproteinaemia develops. Salt and water retention exacerbates the oedema and causes some degree of exertional dyspnoea. The findings are those of severe iron deficiency anaemia, with mucosal pallor and koilonychia. Some patients present with high-output congestive cardiac failure. Diagnosis and investigation Diagnosis is confirmed by finding eggs in the faeces. Worm burden can be assessed from 24-hour faecal egg outputs. Hypochromic microcytic anaemia is present. It is important to distinguish between iron deficiency and p-thalassaemia trait, which is common in the tropics. Eosinophil counts can be normal or increased, though they are often markedly raised during the phase of migration.

Management Mebendazole (100 mg twice daily for 3 days) or albendazole (200 mg twice daily for 2 days) is effective. Bephenium hydroxynaphthoate granules (5.0g) are given on an empty stomach. The cheapest treatment is tetrachlorethylene

The hookworm parasite of dogs and cats, Ancylostoma braziliensis, produces eggs which are passed in the animal faeces; they hatch in warm, moist soil conditions to release larvae which mature to the infective stage and can penetrate the skin of humans. The larvae cannot follow their usual migration in humans and instead meander around in the subcutaneous tissues close to the site of entry until they die. This can go on for several months. The feet are the most common site for this condition, but any part in contact with contaminated soil or sand can be affected. This is often seen in people who have recently had a tropical beach holiday. Occasionally the infective hookworm larvae of Ancylostoma and Necator will produce the same appearance. The appearances are readily recognized. A red papule marks the tip of the track produced by the migrating worm, and is in fact a little way behind the migrating worm. A raised, red, serpiginous track, about 3 mm across, stretches out behind. 1 The affected skin is itchy and may be blistered. Secondary bacterial infection can occur. Sometimes when the lesion is deeper in the skin a subcutaneous nodule is all that can be felt, with none of the signs described. The lesion can be active for several months. The larvae can migrate through the fabric of a swimming costume and so lesions are seen on areas that have been covered. The diagnosis is made on the appearances. Treatment with oral albendazole, 400 mg daily for 3 days, is effective, as is a single dose of ivermectin 200u,g/kg body weight, though with more numerous worms longer treatment is needed. Tiabendazole topically as a 5% cream applied twice daily for 7 days is also effective. Following effective treatment worm migration ceases and the track will slowly disappear. Secondary bacterial infection requires antibiotic treatment.

361

Migration via blood to lungs and then to gut

Adults in gut

Autoinfection through anal mucosa and perianal skin

Rhabditiform larvae in gut

Invasive infective larvae HUMAN SOIL Free-living larvae Sexual reproduction FIG. 9.38 Lifecycle of Strongyloides stercoralis

Strongyloidiasis Strongyloidiasis is an infection of the small intestine with the nematode parasite Strongyloides stercoralis, which is distributed worldwide but flourishes particularly in a hot humid climate. Most people are asymptomatic. A small number have cutaneous manifestations, and immunosuppressed patients may develop massive systemic invasion with filariform larvae. A particular characteristic of this helminth is its capacity to produce prolonged infections for decades after exposure as a result of autoinfection (e.g. former prisoners of war of the Japanese in Thailand and Myanmar). Aetiology Strongyloides stercoralis adult females are small (2mm long) and are found in the intestinal lumen, where they insinuate their anterior ends between the villi. The lifecycle is described in Figure 9.38. A variation of the lifecycle involves rhabditiform larvae developing into filariform larvae during transit down the gut. These larvae penetrate the rectal or anal mucosa or the perianal skin, and then enter the usual migratory tissue phase. It is this variation of the lifecycle that gives rise to prolonged infection, because filariform larvae develop into new adults that maintain infection. When it occurs in immunosuppressed patients very large numbers of filariform larvae develop and give rise to the massive autoinfection syndrome. Transmission and epidemiology Transmission and epidemiology are broadly the same as those of hookworm but tend to be focal, depending on soil

362

1

Fig. 9.49

2

Fig. 9.50

4

MCQ 9.28 5 Fig. 9.51

3

conditions. The occurrence of autoinfection in Strongyloidiasis allows infection to persist for 30 years or more, successive generations of adults maintaining infection. It has wide distribution in the tropics but is particularly common in Asia. Pathology The migrating larvae in the lung evoke some degree of eosinophilic pneumonitis. Markedly raised eosinophil counts are usually seen during the phase of migration, and lesser but elevated counts persist in established infection. Adult larvae provoke a mild inflammation in the gut. Maiabsorption may occur but the mechanism is not understood. Diarrhoea, hypoproteinaemia and oedema of the small bowel wall occur in children. 1 Massive autoinfection with filariform larvae can occur in immunosuppressed patients who are taking steroids or are taking steroids with cytotoxics as part of chemotherapy for tumours, with dissemination of larvae to the gut, lungs and CNS, and concurrent septicaemia from the gut. There is a clear relationship between infection with HTLV-1 infection (see p. 1237), Strongyloidiasis and massive autoinfection. Clinical features Most patients are asymptomatic and the diagnosis is made by stool microscopy finding rhabditiform larvae (see Fig. 9.36A) or in the investigation of eosinophilia. Some patients present with a transient, linear or serpiginous urticaria that comes and goes within 40-60 minutes. Lesions are seen on the trunk, buttocks and upper thighs. This is caused by subcutaneous migration of infective filariform larvae, larva currens 2 (cf. Cutaneous larva migrans, above). Intermittently rhabditiform larvae develop into infective filariform larvae during their passage down the intestine, and these then invade the rectal wall and perianal skin causing autoinfection. The massive autoinfection syndrome occurs most often in people on steroids or immunosuppressive drugs. Fever, abdominal distension, diarrhoea, vomiting and ileus are common presenting symptoms. There may be signs of meningitis. Dyspnoea, cyanosis and pulmonary infiltrates indicate pulmonary involvement. Septicaemia with organisms from the colonic microflora is common. There is a range of disease severity, and the earlier it is recognized and treated the better the outlook though, overall the prognosis tends to be poor. Diagnosis Diagnosis is usually made by finding rhabditiform larvae in stools or juice obtained from the duodenum by intubation. Eosinophilia is usual in normal persons but absent in patients with massive autoinfection. Larvae are readily found in sputum, jejunal aspirate or faeces in massive autoinfection.

Case 9.5

Management Albendazole (400mg twice daily for 3 days) is effective and

causes few side-effects. There is evidence to suggest that ivermectin, 200ug/kg as a single dose may be more effective. Treatment with albendazole is continued for 7 days or longer in massive autoinfection. There is no experience with ivermectin given over days, but adverse reactions to the drug are not common. Occasionally in those very severely affected the parenteral veterinary preparation of ivermectin has been given with good effect, though relapse is all too common and some form of maintenance therapy is often needed, e.g. intermittent ivermectin. Intensive care is also needed for these patients. Prevention Safe disposal of faeces would prevent S. stercoralis infection, and wearing sandals would help to prevent skin-soil contacts. Treatment with cytotoxic or immunosuppressive drugs for leukaemia, lymphoma and other malignancies, reactional states in leprosy or organ transplantation should be preceded by albendazole or ivermectin treatment in patients from or in the tropics. 3 4

Threadworms Infection with the intestinal nematode Enterobius vermicularis causes pruritus ani. Children are most often affected, but adults can also become infected. The lifecycle is described in Figure 9.39. The full cycle can take 14 days. Transmission Autoinfection maintains the parasite in the host. Personto-person spread is the likely route of infection in children and other family members, or by handling pyjamas, sheets or towels with adherent eggs. Clinical features Intense pruritus ani is the usual symptom of this infection and often disrupts sleep. When symptoms are present the worms may be seen on the perianal skin. In females vulvovaginitis and vaginal discharge may occur. Diagnosis Diagnosis is made either by identifying a female worm obtained from the perianal skin or the faeces, or by finding eggs. The perianal skin should not be washed before the swab is taken. Eosinophilia does not occur. Management A number of drugs are available and it is usual to treat the whole family. Mebendazole (100 mg as a single dose, fol-

SUMMARY 19 Effects of nematode infection on child health • Growth impairment • Increasing evidence for impaired intellectual function

Airborne dust

9

INGESTION Sexual reproduction in small bowel

Scratching

Adult worms Deposit of eggs on perianal skin

Gravid females migrate down colon and through anus

FIG. 9.39 Lifecycle of threadworm

lowed by a second dose 2 weeks later) is effective. Piperazine citrate (65mg/kg to a maximum dose of 2.0g/day for 6 days, repeated after 3 weeks) can be used. Pyrvinium pamoate (5mg of base/kg as a single dose) is also used. In addition to chemotherapy it is helpful to prevent ringer sucking and nail biting. The importance of handwashing after going to the lavatory and before eating should be stressed. Occasionally one child is chronically afflicted by enterobiasis, and this causes both child and family great distress. This may occur in families that are particularly careful about washing. Reassurance is needed that this is not uncommon, and that enterobiasis is occasionally difficult to treat.

Trichuriasis Trichuriasis is infection of the intestinal lumen with the nematode Trichuris trichiura. Most patients are asymptomatic. Malnutrition is associated with heavy infections, causing symptomatic disease which can include stunting of growth and rectal bleeding. Eggs are swallowed and the larvae hatch. They mature during their transit down the small intestine, and adults are found in the terminal ileum and colon. The eggs mature in soil and are ingested. The adult worm has a long thin anterior end and a short thick posterior end. This appearance is described by the common name of the parasite, the whipworm. The thick posterior end grips the colonic mucosa. 5 Clinical features Most people are asymptomatic and eggs are found on faecal microscopy. Malnourished children may become infected with large numbers of these worms, and diarrhoea with blood in the stool may result. These children often have stunted growth. On proctoscopy large numbers of worms can be seen adhering to the rectal mucosa. Rectal prolapse is sometimes seen in these malnourished children. It is difficult to know if the worm burden is the cause or the result of malnutrition. Diagnosis and management Diagnosis is made by finding eggs in faecal smears or concentrates. Mebendazole (100mg twice daily for 3 days) is effective. A marked catch-up in growth occurs in stunted children after their infection is cleared.

363

Microfilariae penetrate stomach wall. Migrate to thoracic muscles. Mature to infective larvae in fly's thoracic muscles

FILARIAL INFECTION The filariae are a group of helminthic parasites whose adult and larval forms are found in humans, the sole reservoir for all but one filarial species. Transmission occurs by the bite of blood-sucking insects. The first-stage larvae, microfilariae, develop into infective third-stage larvae by two further moults in the vector. The infective microfilariae migrate out from the insect's mouth parts on to the skin when the insect is taking a blood meal, and then enter the skin through the puncture site. Development from infective larva to the adult takes up to 6 months. The adult worms are very long-lived; shedding of microfilariae begins about 6 months after infection and lasts for many years. The severity of disease relates to the burden of adult worms in the host.

BLACKFLY HUMAN Microfilariae in skin and eyes

Sexual reproduction in subcutaneous nodules containing male and female microfilariae released

Third-stage larvae mature in subcutaneous tissue to adult form

FIG. 9.40 Lifecycle of Onchocerca

Onchocerciasis Infection with the filarial parasite Onchocerca volvulus causes chronic intense pruritus with a papular eruption due to the death of microfilariae in the skin. Death of microfilariae in the eye causes a range of ocular disease; microfilarial death in the cornea over a prolonged period results in scarring of the cornea and unilateral or bilateral blindness. 1 Microfilarial death in the retina causes choroidoretinitis, and in the optic nerve optic atrophy. 2 The latter two complications are relatively rare. Aetiology and transmission Infective larvae escape on to the skin from the proboscis of an infected Simulium blackfly when it takes a blood meal, and enter the skin through the bite site (Fig. 9.40). The infection is maintained in the vectors when adult blackflies ingest microfilariae in blood meals. These develop into infective larvae over about 9 days. In humans the larvae develop into adults over 6 months and then start to shed microfilariae into the skin. These migrate out from the adult, and only when the larvae die do symptoms start to occur. Adult worms survive for up to 15 years and may shed microfilariae for most of that time. Distribution Tropical Africa and Central America are the main endemic areas, with foci of disease in Yemen, Colombia, Venezuela and the northern Amazonian region of Brazil. In Africa, two distinct areas are affected: the savannah area throughout West Africa, where ocular involvement is common, and the tropical rainforest area where ocular involvement is less common (Fig. 9.41).

364

1

Fig. 9.52

2

Figs 9-53-9.54 3 Fig. 9.55

4

Fig. 9.56

5

Figs 9.57, 9.58

FIG. 9.41 World distribution of filariasis Onchocerciasis distribution is very local and not uniform over the areas shown.

Epidemiology The distribution of the disease relates closely to the ecology of the Simulium vector, which requires fast-flowing freshwater streams and rivers for breeding. Most human cases occur in people who live close to such water sources and use them to collect water, wash clothes or play. Pathology and pathogenesis The changes in the skin and eyes are caused by the death of microfilariae. Following an acute response to microfilarial death there is granuloma formation and fibrosis. Over a number of years the result is destruction of elastic fibres, fibrosis, and thickening in the affected skin. Microfilariae die in the cornea, leading to irreversible corneal scarring. Worms also may die in the iris and ciliary body, causing uveitis with secondary glaucoma as a further complication. Choroidoretinitis and optic atrophy are also recognized. Some heavily infected patients have no symptoms, whereas others have marked symptoms and signs. Anti-

body and eosinophils are the important effector cells in killing microfilariae.

Clinical features The earliest symptom is pruritus, which may be made worse by a hot bath. Examination shows a papular eruption with excoriation. When the arm or leg is affected the limb may be obviously swollen in the early phase of the infection. As this process goes on over years there is destruction of the elastic fibres in the skin, giving the appearance of premature ageing. 3 In Yemen a variant of onchocerciasis called sowda occurs in which there is marked hyperpigmentation of the affected limb. 4 Regional glands are often enlarged and non-tender. The lateral corneoscleral junction should be examined with a slit lamp for keratitis, which appears first as small diffuse white spots. Slit-lamp examination may show microfilariae in the anterior chamber. The posterior segment of the eye should be examined for retinal lesions. The adult worms are found in nodules that consist of the adult female surrounded by fibrous tissue of host origin. 5 These nodules are felt over bony prominences such as the skull, shoulder blades, ribs, pelvic girdle and knees. When they are not palpable they must lie deep to scapulae and muscles of the gluteal region, for example. The papular rash is less apparent in chronic cases where fibrosis has produced inelastic thickened skin. There may be patchy destruction of the pigment layer in a black skin, giving a 'leopard skin' appearance. Destruction of the elastic fibres of skin in the inguinal region creates pouches containing enlarged inguinal glands, and predisposes to inguinal hernia.

bamazine (DEC). Ivermectin is given in a single dose of 200 ug/kg body weight; most adults receive 9 mg. There may be some local swelling of affected skin, but much less than that with DEC. Like DEC this drug only kills microfilariae, not adult worms, and so relapse is likely in about 50% of cases over the next year; these patients are then retreated. Diethylcarbamazine is given with a starting dose of 50 mg, increasing the dose on alternate days to a maximum of 200 mg thrice daily for 21 days. When there is marked local reaction with exacerbation of ocular involvement, topical or systemic steroids may be needed. Severe and even fatal reactions are seen in malnourished and sick patients given DEC, and so it is essential that intercurrent infection and nutritional status are improved before DEC is given. Suramin (Antrypol) is effective against adult worms and microfilariae but is used mainly for its ability to kill adult worms when ocular involvement is present. Anaphylaxis to the drug can occur and the drug is nephrotoxic. Excision of onchocercal nodules is recommended, particularly those on the head and shoulders. Early trials suggest that doxycycline may have a role in treatment, as it kills Wolbachia, bacteria within Onchocerca volvulus filariae, on which they are dependent for survival (Figs 9.42 and 9.43).

9

Prevention and control The Onchocerciasis Control Programme in West Africa has been very successful in reducing transmission of the

Diagnosis The parasitological diagnosis is confirmed by examining skin shavings from affected areas for microfilariae. These can be obtained using a needle and scalpel blade, but skin snips taken using a corneal punch is perhaps easier for the patient. Eosinophilia and a positive serological test for filariasis are usually found in the peripheral blood but are not diagnostic for onchocerciasis. Differential diagnosis Scabies and lichen planus are common or relatively common itching skin diseases in the tropics and should be considered. Nodules and ocular disease are not found, and the appearance and distribution of lesions in both conditions is different. Loa loa infection (p. 368) causes transient soft tissue swelling that produces short-lived pruritus. The lymphatic filarial infections cause unilateral limb swelling with regional adenopathy but no pruritus or papular rash. Treatment Ivermectin is now the drug of first choice because it exacerbates clinical signs much less than does diethylcar-

FIG. 9.42 Immunohistology, midbody cross-section of 'live' female Onchocerca volvulus filaria in onchocercomas Numerous bacteria are stained red in the hypodermis (arrows); some are visible as individual red dots. Note that the bacteria are located over the whole width of the hypodermis, in contrast to mitochondria (see Fig. 9.43), which also stain with antihsp60 antibodies because they have the chaperone hsp-60. Note the presence of normal embryogenesis, such as morulae (single arrow head), coiled embryos (double), and microfilariae (transverse sections, triple arrowheads). (Reproduced with permission of Dr A. Hoerauf and Professor D.W. Buttner.)

365

Microfilariae mature in mosquito thoracic muscle to become infective larvae

MOSQUITO HUMAN

Microfilariae shed by female in lymphatics enter bloodstream FIG. 9.43 Immunohistology, midbody cross-section of 'live' female Onchocerca volvulus filaria in onchocercomas Worm from a doxycycline-treated patient 8 months after the end of doxycycline treatment (100 nig P.O., daily for 6 weeks). No bacteria are detectable in the hypodermis. There is only some staining in areas towards the cuticle (c), where mitochondria (also expressing hsp-60) are located (arrow). Uteri (u) contain only remnants of degenerated embryonic stages (coiled embryos, single arrowheads, degenerated mf, double arrowheads). (Reproduced with permission of Dr A. Hoerauf and Professor D.W. Biittner.)

disease by the application of insecticides to vector habitats and by treating cases. The success is evidenced by a striking reduction in infection rates among children, but the costs involved are considerable. Insecticide resistance may become a serious handicap to further progress.

FURTHER READING Hoerauf A, Volkmann L, Hamelmann C et al 2000 Endosymbiotic bacteria in worms as targets for a novel chemotherapy in filariasis. Lancet 355:1242-1243

Lymphatic filariasis Lymphatic filariasis is a group of filarial infections in which adult worms, found in major abdominal lymphatic vessels, shed microfilariae which are carried to the bloodstream. In contrast to onchocerciasis, it is the adults that cause the pathological changes resulting in lymphoedema. This occurs in a relatively small proportion of those infected. The three species involved are Wuchereria bancrofti, Brugia malayi and B. timori. Geographical distribution of W. bancrofti is shown in Figure 9.41.

1

366

Fig. 9.59

2

Adults develop in lymphatic channels

FIBROSIS AND INFLAMMATION CAUSED BY ADULTS IN LYMPHATICS -> ELEPHANTIASIS FIG. 9.44 Lifecycle of lymphatic filariasis

Aetiology and transmission W. bancrofti (Fig. 9.44) females measure 80-100 mm long by 0.3mm. The males are smaller, 40-50 mm long. The adults live in lymphatic channels and the female sheds about 50000 sheathed microfilariae per day; these are carried to the bloodstream in lymph. They circulate in the blood and are taken up by mosquito vectors of the genera Culex, Anopheles and Aedes. The infective larvae moult twice after they enter a human host when the insect bites. W. bancrofti exhibits periodicity of microfilaraemia throughout most of its geographic distribution, releasing microfilariae during the night hours; maximal counts are at about 1 am, in close relation to the nocturnal feeding habits of the mosquito vectors. Subperiodic strains release microfilariae throughout the 24-hour period. Brugia malayi and B. timori have similar adult morphology and a similar lifecycle. Distribution ana1 incidence (Fig. 9.41) Bancroftian filariasis has the widest geographical distribution, covering tropical Africa, Asia east of Pakistan as far as eastern China, Indonesia, the Philippines, Papua New Guinea, the Caribbean and the northeastern regions of South America (all periodic forms). Subperiodic forms are widely distributed in southeast Asia and the Pacific. Current estimates indicate that about 905 million people live in endemic areas, about 10% of whom are infected. Of these, 90% have W. bancrofti. Two-thirds of all those infected live in China, India and Indonesia.

Fig. 9.60

Pathology and pathogenesis Obstructive lymphoedema of the arms, breasts, genitalia

or legs is the main outcome of inflammatory changes in the main lymphatic channels elicited by the presence of adult worms. Adult worms may be found in lymphatics or lymph nodes. Distal to this obstruction the lymphatics are dilated. Inflammatory changes in the epididymis can be an early manifestation of the host response to W. bancrofti. Similar histological features occur in infection with Brugia species, but these most often affect the legs. In established infections episodic non-bacterial lymphangitis and lymphadenitis occur, evoked by allergy to filarial antigens. Eosinophils, giant cells, dead adult worms and fibroblasts are seen in the affected lymphatics and glands. Established lymphoedema predisposes to and is exacerbated by bacterial cellulitis and lymphangitis. Minor cuts and abrasions and tinea pedis are routes for entry of streptococci and staphylococci. Bacterial infections add to the irreversible changes caused by the filarial obstruction of lymphatic channels. In the chronic phase lymphatic channels are obstructed by fibrosis with distal lymphoedema. There is fibrosis in the lymphoedematous tissue. Clinical features Many infected people have no symptoms related to the disease. The earliest feature is often generalized painless swelling of the leg. There is no rash, and inguinal glands may be normal. The swelling is worst in the evening and declines while the patient is recumbent overnight. This occurs within months of infection, before microfilariae are shed, which indicates a response to adult worms. Without treatment this swelling may progress to the chronic lymphoedema of filariasis. Filarial lymphangitis is another presentation of filariasis. Lymphatics are red and tender, and inguinal glands are swollen and painful. The lymphangitis may spread distally in a limb, which is the opposite of spread in bacterial lymphangitis. Epididymitis, which may or may not relapse, is another manifestation of the same process. The end stage is non-pitting lymphoedema with thickened, lichenified skin. 1 The penis, scrotum or labia may be involved. The breast is sometimes involved. Infection with Brugia species tends only to affect the lower limbs. Filarial abscesses can develop along the line of lymphatic channels, more often proximally in the limb. These rupture and discharge, and the ulcerated area heals well. W. bancrofti infection may cause intermittent chyluria, in which anastomoses between intestinal and renal lymphatic channels are open; these are due to more proximal obstruction in main lymphatics leading to the cisterna chylae and thoracic duct. An uncommon presentation of filariasis is referred to as tropical pulmonary eosinophilia (TPE). It occurs with both bancroftian filariasis and with Brugia malayi infections. It takes 6 months before TPE develops with bancroftian infections, but only 3 months with Brugia. The manifestations of this condition are almost entirely pulmonary. Dyspnoea with cough, predominantly nocturnal, is the major

symptom. There may also be some night sweats. Wheeze is not a usual feature. Chest X-rays may show a fibronodular pattern and peripheral blood eosinophil counts are markedly raised (6.0 x 109/L is not uncommon), with strongly positive filarial serology. Day and night bloods are free of microfilariae. It is thought that this is a manifestation of hypersensitivity on the part of the host's immune system to the microfilariae, resulting in retention and destruction of the nocturnally released larvae within the pulmonary vasculature.

9

Diagnosis Diagnosis is usually made on the history, physical signs and marked peripheral blood eosinophilia when patients are seen at the early lymphoedema stage. Later, when there is relapsing lymphangitis or epididymitis, microfilariae are found in the blood, 2 maximally at 1 am in the periodic forms, and by day and by night in the subperiodic forms. When there are many microfilariae they are readily found in stained blood smears. Filtration of blood through a filter to retain the microfilariae may demonstrate parasites in lighter infections. In patients with scrotal nodules, ultrasound scanning can sometimes visualize adults which are motile, the 'filarial dance sign'. The adult parasites may have died out in patients with end-stage disease, and so parasites cannot be found. Serological tests are not specific for filarial infections and cross-reactions with Strongyloides are common, so that a positive test should stimulate a further search for parasites. In TPE the history, with relevant exposure in endemic areas plus marked eosinophilia plus strongly positive filarial serology, are the features required for diagnosis.

Differential diagnosis Tuberculosis, malignancy and chronic bacterial lymphangitis may all cause lymphoedema in the tropics. Management Diethylcarbamazine (DEC) is used, as for onchocerciasis (p. 364) though 12 days' therapy is sufficient. Reactions due to death of worms do occur. Lymphoedema is best controlled by elevation of the affected limbs at night and pressure bandaging. Albendazole and ivermectin are also effective. Surgery to excise the lymphoedematous tissue is not satisfactory and is only a final resort. Prophylactic penicillin V, 250 mg twice daily, is helpful in patients who have had repeated attacks of bacterial lymphangitis. Meticulous care of the feet, and particularly the skin between the toes, is essential. Staphylococci and streptococci are the common causes of bacterial infection in the affected limbs. TPE responds to DEC in the same dosage regimen, though it may be necessary to give prednisolone as well to inhibit exacerbations of pulmonary symptoms. Prevention and control Traditionally these have depended on vector control and

367

treating cases. Recent work in Indonesia and Kenya has shown that low-dose DEC can be used to control the infections.

Loiasis Loiasis is a filarial infection characterized clinically by transient soft tissue swellings at sites of adult Loa loa worm death or migration of adult worms across the eye, subconjunctivally. Chrysops flies are the vectors. Aetiology Loa loa adults measure about 60 x 0.5mm (female) and 30 x 0.4mm (male). They migrate freely in subcutaneous tissues. 1 Six months after infection adults have developed and fertilized females start to shed microfilariae, which are found in peripheral blood during the daytime. Chrysops flies ingest microfilariae in blood meals. Development to adult worms takes place in subcutaneous tissues, and shedding of microfilariae begins by 90 days after infection. Adults live for up to 15 years. Epidemiology The endemic areas (Fig. 9.41) cover the tropical rainforest regions of West and Central Africa, extending as far east as the southern Sudan and Uganda. The vectors live in the forest canopy and descend to lower levels to feed. The prevalence in some populations is 100%. Animal reservoirs do not play a part in the cycle of transmission. Clinical features Transient itchy soft tissue swellings up to 7cm across, on the limbs or less often the face, are a common presentation; these are called 'Calabar swellings' and are most often seen in people who have lived in endemic areas for a relatively short time. In indigenous residents of endemic areas a more typical presentation is migration of an adult worm across the globe of the eye beneath the conjunctiva, causing local irritation, a feeling of movement and some alarm. 2 Occasionally a worm dies immediately beneath the skin, producing a linear swelling in the skin. Problems can arise in patients who have large numbers of microfilariae in the peripheral blood (over 25-50/mm3), beginning a short time after starting treatment with DEC. Encephalitis or encephalomyelitis can occur, probably because of dead microfilariae occluding vessels of the brain and spinal cord. Cerebral oedema and granulomatous reactions around microfilariae are found in fatal cases. Diagnosis The diagnosis is confirmed by finding typical sheathed microfilariae in peripheral blood samples taken during the day. 3 When there are few microfilariae, millipore filtra-

368

1

Fig. 9.61

2

Fig. 9.62 3 Fig. 9.63

4

MCQ 9.29

5

Fig. 9.64

tion of blood may show parasites (see above). Eosinophilia in the peripheral blood is usual and a positive serological test, e.g. filaria indirect fluorescent antibody test, provides only indirect evidence of infection. Management DEC (p. 365) is the most active drug in the treatment of loiasis. It kills microfilariae well and adult worms more slowly. While the patient is on DEC, Calabar swellings may appear. Also, worms dying close to the skin may be seen. When microfilariae counts are above 25-50/mm3 DEC cannot be given alone because of the danger of precipitating cerebral complications. Removal of microfilariae from the circulation is recommended. Whole blood exchange transfusion, or apheresis using a blood separator, can be used to remove large numbers of microfilariae. This makes it safe to give DEC. Prednisolone in doses of 40-60 mg/day, beginning 24 hours before starting DEC and continuing till the top dose is reached, is usually more convenient. Prevention and control Measures for the control of loiasis have not been successful. Vector control is difficult. Mass chemotherapy is not appropriate because of the need for individual supervision of treatment in people with heavy infections, and because of the coexistence of onchocerciasis in the same endemic areas. Other measures include protection from biting by clothing and the use of insect repellants. 4

ANIMAL INTESTINAL NEMATODE INFECTION' Trichinosis Aetiology Trichinella spiralis is a nematode parasite with very low host specificity that will infect over 100 species of animals, particularly carnivores. Human infection occurs by consumption of undercooked meat containing encysted larvae (Fig. 9.45). These hatch in the human gut and mature to the adult stage. Fertilized females shed viviparous larvae for up to 14 weeks until the adults are expelled. The larvae penetrate the gut mucosa to reach the circulation and are disseminated to all the tissues, where they excite inflammatory responses. Larvae encyst and survive in skeletal muscles. 5 Distribution, incidence and transmission T. spiralis is widely distributed in nature through the Americas, Asia, Africa and the Arctic. Sporadic cases and epidemics occur. Outbreaks also occur in Europe. All ages and both sexes are susceptible but children seem to have milder attacks, probably related to a lower infecting dose of larvae in their smaller meals. Exposure to infection is more likely among those nationalities whose cuisine involves eating raw or lightly cooked meats. Smoked sausages made from wild boar are common sources of infection in European countries.

positive. Differential diagnosis includes polyarteritis and other causes of eosinophilia. Eats infected food

Encysted larvae eaten in raw or undercooked pork

A/loncrgemenf Albendazole 400 mg twice daily for 10 days will eradicate adult warms from the gut and thus cut off the production of invasive larvae. There is evidence that anthelminthics affect larvae in the tissues. Steroids may be needed when inflammatory responses in the tissues are very marked.

Larvae moult in intestinal mucosa to become adults in gut

Prevention and control Thorough cooking or freezing (-15°C for 20 days) of meat will prevent trichinosis. Public health measures include boiling pig swill and inspecting meat.

PIG

HUMAN

Larvae encyst in striated muscle Larvae invade intestinal mucosa and travel to muscles via lymphatics and blood vessels FIG. 9.45 Lifecycle of Trichinella spiralis

Pathology Granulomas form around worms dying in the tissues. Focal interstitial myocarditis is found in the heart, with a marked infiltrate of eosinophils. A non-suppurative meningitis with granulomas and capillary thromboses is seen in fatal cases. Invasion of the eye muscles may contribute to the periorbital oedema. Skeletal muscles show encysted larvae, myositis with marked eosinophil infiltrate, and patchy degeneration of muscle fibres. Calcification may occur in the capsule of the cyst or in the larvae by 6 months after invasion, but larvae can remain viable in muscle for 10-15 years. Clinical features Four main features of the disease are fever, orbital oedema, myalgia and eosinophilia. Bowel upset is variable. Vomiting occurs, and the alteration in bowel habit may be diarrhoea or constipation. In the phase of muscle invasion the affected muscles are tender on movement. Fever, urticaria, splinter haemorrhages in fingers and toes, difficulty with swallowing and difficulty with breathing may all be present. Invasion of the nervous system can cause a meningitislike picture. Fits, paralysis, disturbed conscious level, difficulty with balance and personality disorders may occur. Heart failure can result from severe cardiac involvement. Symptoms due to reactions to larvae in the tissues start to resolve gradually from about 14 days onwards in milder cases, but can take several months to settle completely. Diarrhoea and a malabsorption syndrome have been described. Diagnosis Trichinosis is suspected clinically and confirmed by examination of muscle biopsy for encysted larvae. Muscle biopsies can be obtained under local anaesthetic using a Trucut needle and are examined immediately, squashed between microscope slides. Eosinophil counts are very high and muscle enzymes are raised. Trichinella serology is strongly

9

Toxocariasis Aetiology Toxocariasis is caused by infection with larval forms of Toxocara canis or T. cati, which are primarily parasites of dogs and cats. The adult worms are found in the gut lumen and eggs are shed in the faeces. Humans are infected by ingesting eggs, and the larvae released burrow into the gut wall to enter vessels and disseminate. Larvae mature outside the host over 14 days and then are infective to the definitive host or to humans. Distribution, incidence and transmission The distribution is worldwide in dogs and cats, although there is considerable variation in the frequency of animal infection and contamination of the environment. Studies in Britain showed that 17% of soil samples were contaminated with toxocara eggs. Moist soil conditions are more suitable for transmission than hot, dry conditions. Children are most often infected. Infection is likely to occur when fingers are contaminated with soil or sand containing eggs during play and then put in the mouth. Dog faeces deposited in public parks, especially around playgrounds, represent a potential source of infection. Dog breeders and people who work in kennels are at risk of infection. Toxocariasis due to T. cati is much less common. Pathology and pathogenesis Disease in humans relates to the number of infecting larvae and the host response. Clinical manifestations are due to dying and dead larvae, which evoke granuloma formation with eosinophils, macrophages and lymphocytes. The eye, brain, liver, spleen and lungs may be involved in toxocariasis, but granuloma formation may occur in any organ of the body. Clinical features Ocular toxocariasis and visceral larva migrans (VLM) are two clinical presentations of this disease, which is often a subclinical infection. Unilateral visual impairment is the usual symptom in ocular toxocariasis. Lesions directly on the visual axis will cause severe impairment. A child may

369

develop a squint. The granuloma can form in relation to the lens and ciliary body, or on the retina itself. A cataract may develop secondary to a granuloma affecting the ciliary body. Visceral larva migrans is due to a heavy infection with larvae. Fever, anorexia, chills, night sweats and weight loss are usual features. Examination shows hepatosplenomegaly as the main physical sign. Pneumonitis may also be present. There is usually a marked eosinophilia in the peripheral blood in VLM, but eosinophilia is less common with ocular disease. Serological testing is valuable, the ELISA technique using a toxocara secretory antigen being a sensitive and specific test. Differential diagnosis Toxoplasmosis usually causes bilateral choroidoretinitis with destruction of the retina. 1 Lymphoma, tuberculosis and sarcoidosis are usually considered in the differential diagnosis of VLM, although the gross eosinophilia is against the former conditions and supports a helminthic infection. The toxocara antibody test is strongly positive. Management DEC is the usual treatment, giving initial doses of 50 mg and doubling the dose on alternate days till the maximum (lOmg/kg/day in three doses for 21 days) is reached. Ocular disease is not specifically affected by DEC because the worm is dead. There may be some spontaneous improvement as inflammation and granuloma size reduce. DEC is given in these cases to kill any worms that are still migrating. Prevention and control Regular deworming of dogs, particularly pregnant bitches and puppies, reduces the numbers of eggs contaminating the environment. Dog owners should try to ensure that their dogs defaecate somewhere where the faeces will not contaminate open spaces, playgrounds and parks.

Dracunculiasis (Guinea worm infestation) Dracunculus medinensis, the Guinea worm, is a tissue nematode widely distributed through Africa and Asia. Humans are infected by drinking water containing minute crustaceans of the Cyclops genus infected with Guinea worm larvae. The larvae penetrate gut tissues and migrate through host tissues, maturing to the adult stage. The female is fertilized by the male, which dies, and the gravid female migrates out into a limb, producing a painful, fluid-filled blister about 3cm across which then bursts and about 5 cm of the female protrudes. 2 Vast numbers of larvae are released from the worm. Patients often put

370

1

Fig. 9.65

2

4

Figs 9.68, 9.69

Fig. 9.66

3

Fig. 9.67

the affected limb into cold water to relieve the pain; this stimulates the discharge of larvae, providing the opportunity for larvae to continue the lifecycle by infecting other cyclops. The diagnosis is made on the clinical appearances. Treatment comprises relief of pain, treatment of secondary bacterial infection with antibiotics, and administration of tiabendazole. This does not have any direct effect on the worm but reduces inflammation around it, allowing it to be gently wound out of the subcutaneous tissues on a stick. O Control of Guinea worm is a realistic prospect using very simple methods. Cyclops can be filtered from drinking water using simple filters with nylon mesh of appropriate pore size. These can be made locally from materials that can be supplied by Ministries of Health. Village communities can all be involved in disseminating information about the disease, making the filters, and how to avoid disease by simple water filtration.

TREMATODES Schistosomiasis Infection with digenetic flukes of the genus Schistosoma affects the bladder and urinary tract (S. haematobium) or intestine (S. mansoni, S. japonicum). These worms cause disease because of the host's response to eggs retained in the tissues. The severity of disease relates to the number of eggs in the tissues, which is proportional to the worm burden. Aetiology Eggs are passed in stools or urine; those deposited in still or slow-flowing fresh water hatch to release the ciliate miracidium, which can survive for up to 48 hours before it dies (Fig. 9.46). During this time it must find an aquatic snail of the appropriate genus: Biomphalaria for S. mansoni, Bulinus for S. haematobium, and Oncomelania for S. japonicum. Cercariae are released from the snail and these penetrate the skin of a suitable host, almost always human, becoming schistosomules during penetration. These migrate via the blood vessels to the pulmonary vasculature, where some traverse the pulmonary circulation to enter the systemic circulation. Schistosomules of mansoni and japonicum mature in the hepatic branches of the portal vein, and by 6 weeks after infection they are mature. They migrate out of the liver against the flow of blood in the portal vein to small veins around the colon and small intestine. The colon is mainly involved in mansoni infections, and both colon and small intestine are involved in japonicum infections. The lifecycle of S. haematobium is similar, but maturation takes place in pelvic veins. The worm pairs migrate into small branches of the internal iliac vein, principally around the bladder, although other pelvic structures such

Schistosomules reach venous circulation via lungs. Become sexually mature and mate

9l

Male and female migrate to lay eggs in vesicle veins (S.haematobium) cr intestinal veins (S.mansoni, S.japonicum)

Blood and lymphatics

Eggs reach water in urine (S.haematobium) or faeces (S.mansoni, S.japonicum) HUMAN

Cercariae penetrate skin

FRESH WATER Eggs hatch, liberating ciliate miracidia SNAIL

Sporocysts rupture on to external surface of snail, releasing cercariae

Miracidia penetrate foot piece of one of several species of snail (intermediate hosts), where sporocysts form

Cercariae develop within sporocysts FIG. 9.46 Lifecycle of schistosome parasites

as the prostate and seminal vesicles in men, and the uterus and adnexal structures in women, may be involved. Adult worms survive for about 7 years on average, but survival of over 30 years has been reported. Distribution and incidence The distribution is shown in Figure 9.47. There is considerable overlap in the endemic areas of haematobium and mansoni, but japonicum has a distinct Asian distribution. The prevalence and incidence of infection vary considerably, with rates up to 70% or more in the endemic areas. Infection rates tend to be highest in children and decline with increasing age. Transmission and epidemiology Freshwater contact is the major factor in infection and maintenance of transmission. Children are infected early in life by playing in infected water. They are also likely to urinate and defecate in and around pools and streams, further enhancing the local intensity of transmission. The highest rates and intensities of infection are found in the second and third decades of life. Pathology and pathogenesis The pathological changes relate to the presence of eggs in the tissues. Eggs laid in the small branches of veins may: • pass through the wall of the viscus to reach the exterior in faeces or urine

FIG. 9.47 World distribution of schistosomiasis

• be retained in the tissues • embolize through vessels to lodge in presinusoidal branches of the portal vein in mansoni and japonicum infections, or in pulmonary arterioles in haematobium infections. The egg excites a granulomatous response with macrophages, lymphocytes, plasma cells and eosinophils. Granuloma formation is important for the transition of the egg through the mucosa to the lumen of the viscus. The chronic inflammation they may cause in the epithelium may be a factor in carcinogenesis of the bladder (haematobium). Eggs retained in the tissues within granulomas gradually break down until all that may remain is remnants of the egg case. The granuloma heals by fibrosis. Early in the course of infection exuberant granuloma formation can cause colonic polyps in mansoni and japonicum infections, and bladder polyps in haematobium infection. These resolve after treatment and may also resolve spontaneously. Fibrotic polyps and strictures are not often seen in mansoni infection, although they are more common in japonicum infection. Haematobium infection causes fibrosis of the bladder, which is often shrunken, with a thick wall. Bladder stones may form and cystitis is common. Unilateral or, less commonly, bilateral ureteric strictures cause obstructive uropathy. Stones may form in the dilated ureters. Distortion of the ureterovesical junction may allow vesicoureteric reflux of urine with associated recurrent pyelonephritis. Squamous cell carcinoma of the bladder is common in highly endemic S. haematobium areas. It is suggested that schistosomiasis acts with dietary factors to produce malignancy. This has not been found in S. mansoni infection or in S. japonicum, despite earlier reports from China. Mansoni and japonicum eggs embolize to the liver. The granulomas that form around eggs 4 add to the volume of the liver and increase its size, and obstruct the flow of blood through the portal circulation in the liver, causing congestive hepatosplenomegaly in those with heavy worm burdens. Granulomas resolve with local fibrosis, and when

371

this process goes on over years periportal fibrosis results, causing irreversible portal hypertension. The anatomy and architecture of the hepatic lobules are not altered and there are no regenerating nodules, therefore cirrhosis is not caused. In the lungs granulomatous reactions in the vessels cause fibrosis and pulmonary hypertension. Involvement of the CNS occurs in S. japonicum, S. mansoni and S. haematobium, in descending order of frequency. Disease of the central nervous system is most probably the result of an ectopic worm pair laying eggs in or around the brain or spinal cord. Mansoni and haematobium infections are associated with spinal cord and cauda equina lesions. Clinical features Cercarial invasion of the skin may cause a local irritant papular eruption, but this is uncommon. Most patients have no symptoms related to the phase of migration and maturation. Occasionally non-immune people develop the Katayama syndrome with the phase of worm migration. There is malaise and lethargy, and fever, profuse sweats, muscle pain, abdominal pain, joint aches, unproductive cough, urticaria, swollen eyelids and hepatosplenomegaly occur in more severely affected patients. These symptoms begin 3-4 weeks after exposure and persist for up to 3 months, with reducing severity. Chest X-rays may show coin lesions caused by worms dying in vessels and evoking local inflammatory reactions. Marked eosinophilia is common. Eggs are not found until 3-6 weeks after infection. S. mansoni and S. japonicum Many infected patients go through life without any symptoms related to this infection. Rectal examination is normal; sigmoidoscopy is normal or shows scattered mucosal haemorrhages. Occasional patients present with anaemia and oedema because of bleeding and protein loss from schistosomal polyps. The most common physical sign in infected patients in endemic areas is hepatomegaly, which correlates with the intensity of infection in the first two decades but not in older age groups. Similar findings are noted in S. japonicum infection. Abdominal pain and subacute intestinal obstruction occasionally occur in the rare patient who develops fibrotic strictures in the colon or, less often, the small intestine. Hepatosplenomegaly with congestive splenomegaly suggests end-stage schistosomal hepatofibrosis. These patients often present with haematemesis and ascites. Portosystemic encephalopathy is not a usual feature after variceal bleeding in schistosomiasis. Severe pain over a grossly enlarged spleen may indicate splenic infarction.

372

1

Fig. 9.70

2

Fig. 9.71

4

Fig. 9.74

5

MCQ 9.30

3

Fig. 9.72, 9.73

S. haematobium Dysuria, frequency and haematuria occur in some patients at the start of egg laying. More commonly, the presenting symptom is terminal haematuria, i.e. the last drops of urine passed are bloodstained. Physical examination is usually normal. Most people have light infections and suffer no long-term adverse effects. Reversible granulomatous polyps and obstructive uropathy occur in the early stages of heavier infection. This appears to be spontaneously reversible, and is certainly reversible after treatment. Those with heavy infections develop fibrotic and obstructive complications which can cause recurrent urinary tract infections, stone formation, and finally renal failure. 1 In these patients the infection is associated with an increased risk of bladder cancer. Excretion urography often shows bladder polyps and obstructive uropathy, either unilateral or bilateral, in older children and teenagers. Chronic infection causes 'sandy' patches on the trigone as a typical cystoscopic finding. 2 Calcification in retained eggs is seen in the bladder wall as a rim of calcium in the pelvis on a plain abdominal X-ray. Tramline calcification in the ureter may also be seen. Other manifestations Pulmonary involvement may be subclinical up to the stage when significant pulmonary hypertension develops. Right heart failure develops, and this is difficult to treat. Cutaneous involvement is indicated by a collection of subcutaneous papules. In haematobium infection the skin of the scrotum and perineum may be affected. 3 Symptomatic involvement of the CNS is uncommon. Intracranial disease occurs with S. japonicum infections, with a frequency of 2-4%. It is much less common in mansoni and haematobium infections. Focal signs and jacksonian epilepsy may occur with involvement of the cerebral hemispheres, whereas posterior fossa lesions are associated with raised intracranial pressure, cerebellar signs and brainstem compression. Lesions of the spinal cord may present with features suggesting cord compression, transverse myelitis or spinal artery occlusion. Granulomas may also form around the cauda equina. The occurrence of focal signs in the neuraxis in a patient in or from an area endemic for schistosomiasis should prompt a search for schistosome eggs in stools and urine that would indicate active infection. Eosinophils in the CSF suggest the possibility of schistosomiasis affecting the CNS. The empirical use of antischistosomal drugs plus steroids also need to be considered. Diagnosis Diagnosis is made by finding schistosome eggs. Eggs of S. mansoni and S. japonicum are found in concentrates of faecal samples. Eggs of S. haematobium are found in the centrifuged deposit of terminal urine samples. Eggs of all three species may be found in rectal snips. Squash preparations can also be made from fragments of bladder mucosa obtained at cystoscopy.

Eosinophilia is a common finding in the peripheral blood but occurs in several helminthic infections. Serological tests can be helpful. An ELISA for antibodies to schistosome egg antigens is useful because it indicates that infection has progressed to the stage of ovideposition. The test takes 18 months at least to revert to negative after treatment, and so is not helpful in assessing cure. Excretion urography, isotopic renography, ultrasound and CT scanning give anatomical and functional detail of the urinary tract in chronic schistosomiasis. Colonoscopy and barium enema can be used to examine the colon in more detail. Ultrasound scanning of the liver demonstrates the presence of periportal fibrosis. 4 Chest radiographs show dilatation of the pulmonary arteries and right ventricular hypertrophy in pulmonary hypertension. MRI or CT allow localization of schistosomal lesions of the neuraxis. Eosinophilia in the CSF supports the diagnosis of CNS involvement in schistosomiasis. Management Three safe, effective drugs are available for use in schistosomiasis (Table 9.39). Their safety in pregnancy has not been confirmed, and so treatment should be delayed until after delivery unless there are urgent indications for prompt treatment. All cases should be treated, even the advanced ones, as further ovideposition causes further urinary tract, intestinal, liver or pulmonary damage. Ideally, stools and rectal snips should be examined for viable ova 3 months after treatment. Prevention and control Control measures are expensive and require a change in the behaviour pattern of the exposed population. Education regarding the reasons for the measures is essential. Dams and irrigation projects for the improvement of the

SUMMARY 20 Schistosomiasis

9

• Infection by exposure to freshwater pools, lakes • Invasion usually inapparent: occasionally itching at sites of invasion (swimmer's itch) • Worm burden determines morbidity and mortality • Tissue damage is caused by host granulomatous response to retained eggs • Bladder (S. haematobium): polyps, strictures, fibrotic and contracted bladder squamous cell carcinoma associated with heavy prolonged infections • Intestine (colon S. mansoni; colon and small intestine S. japonicum) Polyps and stricture in gut uncommon • Liver: schistosomal hepatofibrosis causing portal hypertension, congestive splenomegaly, hypersplenism with variceal bleeding and ascites • Ectopic disease related to the site of a worm pair with CNS disease is the most serious manifestation

economy of an area may create new habitats for snail hosts of schistosome parasites. Molluscicides can be used, but are expensive and may have detrimental effects on other water creatures. Regular, community-based treatment with praziquantel has reduced morbidity. 5 Clonorchiasis

Aetiology Clonorchis sinensis is a hermaphroditic trematode parasite of the biliary tract in humans. The adult fluke is 10-25 mm long by 3-5 mm wide. The lifecycle is shown in Figure 9.48. Transmission ana1 incidence Fishponds into which human faeces are poured as fish feed are an important source of infection, as eggs, snails

TABLE 9.39 Drug treatment of schistosomiasis Drug

Active against

Dose

Side-effects/contraindications

Oxamniquine*

Schistosoma mansoni

West Africa and South America 15mg/kgx1 (adult) 20mg/kgx1 (child) Elsewhere 20mg/kg per day x 3

Occasional febrile episode 5 days post treatment

Metrifonate*

S. haematobium

7.5mg/kgx3 (2 weeks between doses)

Anticholinesterase actions may prolong neuromuscular blockade after surgery. Avoid elective surgery for 3 days post treatment and provide ventilatory support after emergency surgery within 48 hours of administration

Praziquantel*

S. mansoni S. haematobium S. japonicum

40mg/kgx1

Dizziness, nausea and occasional vomiting beginning 1 hour after dosing, lasting up to 4 hours

20mg/kgx3 (4 hours between doses)

*The safety of these drugs has not been confirmed in pregnancy; treatment should be delayed until after delivery unless there are urgent indications for prompt treatment.

373

FIG. 9.48 Lifecycle of Clonorchis and Opisthorcis

and fish are conveniently found together. The disease occurs in Asian countries. Dogs and cats are also important hosts. Pathology The worms obstruct small branches of the biliary tree, and bile accumulates in cysts proximal to the obstruction. Hypertrophy of the bile epithelium with adenomatous proliferation occurs. Later there is cellular infiltrate of the bile duct wall with chronic inflammatory cells. Pigment stones and biliary sludge form. Flukes obstructing the pancreatic ducts can cause pancreatitis. Long-standing heavy infections are associated with malignant change in the adenomatous hyperplasia, leading to cholangiocarcinoma. Clinical features Clinical features relate to the worm burden. Mild infections cause no symptoms. Upper abdominal discomfort, diarrhoea and weakness occur with heavier infections; with the heaviest worm burdens upper abdominal pain, diarrhoea, weight loss, obstructive jaundice, hepatomegaly, cholangitis and Gram-negative septicaemia occur. Diagnosis The eggs are found in faeces or in duodenal aspirate. Eosinophilia is usual in the peripheral blood. Endoscopic retrograde cholangiopancreatography or percutaneous transhepatic cholangiography shows dilated branches of the biliary tree, and the worms appear as filling defects.

1

374

Fig. 9.75

FIG. 9.49 Lifecycle of Taenia species

Management Praziquantel is the treatment of choice, giving 25 mg/kg on three occasions 4 hours apart on one day. When there is obstructive jaundice with cholangitis and septicaemia, relief of biliary obstruction, vigorous treatment of the bacterial infection and circulatory support for the patient are needed. Prevention and control Transmission can be interrupted by safe disposal of faecal wastes. Usually this involves a major change in social habits, which is difficult to achieve. Thorough cooking of fish before it is eaten prevents infection, as does freezing fish, -10°C for 5 days, and salting in 10% saline.

Opisthorciasis Opisthorcis viverrini and O. felineus have similar lifecycles to that of C. sinensis, and in humans they parasitize the intra- and extrahepatic branches of the biliary tree (Fig. 9.49). Raw, pickled, smoked and undercooked fish are the sources of infection. The pathological changes and consequences are those of chronic irritation and obstruction of the biliary tree, with ascending cholangitis in severely affected patients. Cholangiocarcinoma occurs in O. viverrini patients. The diagnosis is made by finding eggs of the parasite in stools or duodenal aspirate. Praziquantel is used for treatment as for clonorchiasis.

Fasciolidsis Fascioliasis is an infection of the biliary tree with the

hermaphroditic trematode Fasciola hepatica. It is primarily an infection of sheep. The lifecycle is similar to that of Clonorchis (Fig. 9.46). Aetiology and distribution The parasite measures 3 x 1.5cm. 1 Eggs are released in the biliary passages and pass to the exterior in the faeces. The miracidium hatches in fresh water and penetrates the tissues of a freshwater snail (Lymnaea species). After a cycle lasting 4 months in the snail, cercariae are released. These encyst on the leaves of watercress and other types of aquatic vegetation. Humans are infected by eating watercress with encysted metacercariae. Maturation to the adult stage in bile ducts takes about 4 months until egglaying starts. Sheep are the normal hosts of this infection, and human infections are likely wherever sheep are herded in wet pasturelands. Pathology There is hepatic parenchymal necrosis in the phase of migration through the liver. The liver tissue regenerates. The worm burden determines the severity of disease. In the bile ducts there is chronic inflammation, with regenerative hyperplasia of the epithelium and some degree of duct obstruction. Clinical features Pain in the liver area, fever, chills, tender hepatomegaly and marked eosinophilia are usual features during the tissue migration phase. Upper abdominal discomfort and hepatomegaly are present in mild to moderate infections, whereas heavy infections may cause obstructive jaundice and cholangitis as well. Diagnosis Eggs of the parasite are found in stools or duodenal aspirate, though they can be difficult to find in relatively light infections. Endoscopic retrograde cholangiopancreatography may show filling defects in the biliary tree and flukes may be extracted from ducts using a stone basket. Eosinophilia is usual and serological tests for this infection are usually positive. Management and prevention Triclabendazole, 10mg/kg b.d. x 1 day, or bithionol, 40 mg/kg in 2 divided doses on alternate days for 15 dosage days, are effective. Watercress production must be separate from sheep pastures. Wild watercress should not be sold for human consumption.

when these are eaten uncooked. After about 3 months the adult has matured in the small intestine; egg production begins and lasts for most of the 6-month lifespan of the adult. Upper abdominal discomfort, diarrhoea, preprandial upper abdominal pain relieved by food, and symptoms and signs of intestinal obstruction may occur, more marked symptoms being found with heavier infections. Eggs are found in the faeces. Tetrachloroethylene is effective in treatment. Praziquantel may prove to be effective.

pj

Paragonimiasis Paragonimus westermani is a hermaphrodite fluke usually found in the lungs but occasionally in other sites, e.g. CNS. It has a wide distribution in South America, West Africa, south east Asia and the Pacific. Human infection occurs when raw or undercooked freshwater Crustacea with metacercariae encysted in their muscles are eaten. The development cycle outside the host involves a freshwater snail: from the snail the cercaria is released and then finds a suitable crustacean host. In humans the metacercaria excysts and penetrates the full thickness of the gut wall to enter the peritoneal cavity. The phase of tissue migration continues with the parasites reaching and crossing the diaphragm to enter the lungs. Where the worm finally comes to rest in lung tissue it elicits first a cellular response and then fibrosis, so that the worm is enclosed but for a connection to an airway. Dead worms calcify. Radiological appearances comprise pleural thickening, infiltrates at the site of recent invasion, nodular lesions (encysted adults), fibrosis and finally calcification. Eggs released by the worms are coughed up or swallowed. Cough productive of sputum, dyspnoea and chest discomfort are usual features. Bacterial infection of the affected area causes fever, systemic upset, purulent sputum and haemoptysis on occasions. Abdominal organs and brain are affected less often. Diagnosis and management Eggs are found in the sputum in most cases. Serological tests are available. Praziquantel is effective in doses of 25 mg/kg given three times daily for 3 days.

CESTODES

HUMAN INTESTINAL CESTODE INFECTION Taenia saginata

Fasciolopsiasis Fasciolopsis buski is the largest trematode parasite, measuring up to 7.5cm long by 3.0cm wide. It has a wide geographic distribution throughout eastern Asia, with a reservoir of infection in pigs. The metacercariae are found on bamboo shoots or water chestnuts, and infection occurs

Human infection with T. saginata is common in Africa, the Middle East, Asia and South America. The infective stage is the cysticercus in beef, and humans are infected by eating rare or undercooked beef (Fig. 9.46). The cyst is released in the gut and the scolex everts from the cyst. It adheres to the gut wall and proglottids start to develop. Each segment

375

has testes and ovaries, and is self-fertile. The worm may be 5-10 m long. The distal segments are full of eggs. Lengths of worm are passed in the stools, and when defaecation occurs in pasture, cattle may eat the gravid segments and develop cysticercosis bovis. The infection causes no symptoms in humans although often a variety of vague abdominal symptoms have been attributed to it. Occasionally a patient will present with pruritus ani and find segments in underwear, the motile segments having wriggled out of the anus. The diagnosis is confirmed by examining segments. T. saginata segments have 15-20 lateral branches to the uterus. The eggs are occasionally found in the stools or on the perianal skin, but are identical in saginata and solium infections (see below). Treatment with praziquantel (10-20 mg/kg as a single dose) is effective in 96% of patients. Niclosamide is also effective. Both praziquantel and niclosamide destroy the scolex and it is not present in the expelled worm remnants. If segments have not been passed by 4 months after treatment the infection is cured. Thorough cooking of meat and freezing meat to -10°C for 10 days kills cysticerci. Meat inspection is also an effective control measure.

Taenia solium The lifecycle is the same as that of T. saginata, except that the pig is the intermediate host (Fig. 9.46). Infection with T. solium in adults causes no symptoms. The segments are distinguished from those of T. saginata by the small number of lateral branches of the uterus: 7-13. Treatment is the same as for T. saginata, with the addition of an antiemetic prior to dosing and a purge 2 hours after dosing. Thorough cooking of pork or freezing to -10°C for 10 days prevents infection. Cysticercosis is discussed below in the section on larval cestodes.

Diphyllobothrium iatum

376

The fish tapeworm has a more complex lifecycle than the other intestinal cestodes: coracidia released from eggs infect water fleas (copepods). Further development takes place in fish that eat the copepods, and in bigger fish that eat plerocercoid-infected fish. Mammals are infected by eating the plerocercoid-infected fish. Digestion releases the plerocercoid, which adheres to the gut wall and starts to produce proglottids. Proglottids break up, releasing eggs which are passed in the stools. Humans are infected by eating uncooked fish. A variety of abdominal and general symptoms are attributed to this infection, including abdominal discomfort, fatigue, weakness, sensations of hunger and diarrhoea. It is associated with deficiency of vitamin B12 and, less commonly, with overt megaloblastic anaemia. When the worm is sited in the upper gastrointestinal tract it can take up both free and intrinsic factor-bound vitamin B12. Features of vitamin B12 deficiency appear when body stores are exhausted.

The diagnosis is made by finding eggs or, less often, typical segments in faecal samples. Purging may provide a sample of segments. Praziquantel gives a high cure rate. Niclosamide is also effective. Thorough cooking of fish and freezing fish to -10°C for 48 hours kills plerocercoids.

Hymenolepiasis Hymenolepis nana and H. diminuta are two species of small tapeworm. H. nana is predominantly a parasite of humans, with person-to-person transmission, whereas H. diminuta is predominantly a rodent parasite, with humans as an occasional host. Children are most often infected with H. nana. Heavy infection causes abdominal pain, anorexia, diarrhoea, irritability, pruritus ani and urticaria. Eosinophilia is found in heavy infections. Treatment with praziquantel (15 mg/kg as a single dose) is effective. H. diminuta has a more complex lifecycle involving an insect intermediate host that ingests eggs with rodent faeces. Most infections cause no symptoms and are diagnosed when eggs are found in faeces. Treatment with niclosamide is effective.

LARVAL CESTODE INFECTION Cysticercosis Infection with the intermediate state of the T. solium parasite (see above) can occur in two ways: • By ingestion of the eggs of the worm, which hatch in the gut to release the cysticercus; • By regurgitation of gravid segments into the stomach to initiate the process of egg digestion and release of cysticerci. The cysticerci released by either route of infection then invade host tissues. The first route is more likely. Occasionally patients are seen with a vast number of cysticerci in a wide range of sites, in those instances it is most likely that the patient has swallowed a proglottid. The development of subcutaneous lumps is a common presentation. Uniocular disturbances of vision occur with cysticerci in the eye. The most serious consequences occur in cerebral cysticercosis, causing epilepsy, raised intracranial pressure and localizing signs related to space occupation. Frequently numerous cysts are found in the brain. The diagnosis is based on the geographic history, a positive serological test for cysticercosis on serum, and the finding of cystic lesions by CT or MRI (Fig. 9.50). In endemic areas where CT is not available the diagnosis would be made on the clinical picture. Antibodies to cysticerci can be detected in CSF, which contains increased amounts of protein, a normal glucose and a normal or increased cell count. Skeletal muscles are often involved in cysticercosis. Usually this causes no symptoms, and calcified cysticerci are seen on X-rays. There may be muscle pain at the

9

FIG. 9.50 MRI scan of cerebral cysticercosis, showing cystic lesions

time of invasion. If cysticerci are found at any site in the body a CT brain scan should be done to detect cerebral involvement. Praziquantel, 50mg/kg/day in divided doses for 10-14 days, and albendazole, 20mg/kg/day in divided doses for 14 days, are both effective in killing cysticerci, and when this happens surrounding inflammation may increase enhancing neurological signs. For this reason steroids are started prior to treatment with either drug to reduce these effects. This should be carried out in a centre where neurological expertise is available. Two-thirds of cerebral cysts disappear after treatment, and clinical improvement follows. The efficacy of drug treatment in this condition has not been proven in prospective studies.

FURTHER READING ON CYSTICERCOSIS Del Brutto O et al 1993 Therapy for neurocysticercosis: a reappraisal. Clin Infect Dis 16:730-735

Hydatid disease Aetiology and distribution Echinococcus granulosus is a tapeworm of dogs. The lifecycle is shown in Figure 9.51. The ingested eggs release the onchosphere, which penetrates the gut mucosa and spreads by vascular or lymphatic channels to other organs, most often the liver, with lung, spleen, brain, eye, bone and other tissues infected less often. Cyprus, Turkey, Middle Eastern countries, the Turkana area of Kenya, and South America are among the endemic areas. Clinical features The clinical presentation is very variable. Unexplained painless hepatomegaly is fairly common, but an expanding cyst can cause right upper quadrant abdominal discomfort.

FIG. 9.51 Lifecycle of Echinococcus granulosus

Rupture of the cyst may produce pleuritic discomfort, abdominal pain and tenderness, and allergic manifestations such as urticaria and anaphylaxis. Rupture of hydatid cysts is infrequent, although the consequences of release of daughter cysts to seed the peritoneal cavity are serious. Pulmonary hydatids may be symptomless and may be found on chest X-ray. They may rupture spontaneously. Hydatid material, membranes and scolices with booklets may be coughed up and disseminate endobronchially to seed the lungs bilaterally, causing numerous small cysts in both lung fields. Bone and joint hydatids cause pain, swelling, and sinuses that discharge hydatid material. The latter may follow surgical exploration. The cyst wall may calcify. Diagnosis There may be blood eosinophilia and positive serological tests, but a viable cyst may be present without eosinophilia or positive serology. Ultrasound and CT scanning (Fig. 9.52) allow accurate localization and measurement of hydatid cysts in internal organs. Bone hydatid produces areas of lucency surrounded by sclerosis. Joints are destroyed. Management Management is difficult. Albendazole is given in doses of 400 mg twice daily for 12 weeks and appears to be effective. However, surgical treatment is probably needed for most cases. There is increasing experience with percutaneous aspiration of cysts under ultrasound guidance. Solitary pulmonary hydatids can often be excised with the relevant segments or lobe. Excision of hepatic hydatids is

377

able, vigorous antibiotic treatment is needed first, followed by safe surgical drainage of the cysts. Hydatid disease of bone is very difficult to treat, as it causes pain and destruction of affected areas. Neither drug treatment nor surgery is very effective, although on occasion, when disease is limited to the femur, total excision of the affected bone with prosthetic replacement, including hip and knee joints, can be carried out.

FIG. 9.52 CT scan showing hydatid cysts in the liver

more difficult. Surgery for hydatids of the CNS requires great care to avoid rupture. Complications Secondary bacterial infection of a hydatid cyst is not uncommon. Amoebic abscess, pyogenic abscess and perihepatic sepsis must be considered in the differential diagnosis when liver cysts are concerned. Gallium scanning may help to indicate bacterial infection and blood cultures should be taken. Amoebic serology is strongly positive in amoebic liver abscess. When infected hydatid cyst is prob-

378

Prevention and control Regular deworming of dogs, killing of stray dogs, and safe disposal or boiling of offal are important control measures. These measures, plus effective communication of the dangers of hydatid disease among populations in Iceland and New Zealand, have controlled the disease.

FURTHER READING ON INFECTIOUS, TROPICAL AND PARASITIC DISEASES Cook G C 1996 Manson's tropical diseases, 20th edn. London: W B Saunders Lambert H P, Farrar W E 1982 Infectious diseases illustrated. London: W B Saunders Mandell G L, Douglas R G, Bennet J E, Dolin R 1994 Principles and practice of infectious diseases, 4th edn. New York: Churchill Livingstone Strickland G T, ed. 2000 Hunter's tropical medicine, 8th edn. London: W B Saunders

Skin Disease Cameron T C Kennedy and Robin A C Graham-Brown

Biology of the skin 379 Examination of the skin 381 Dermatological diagnosis 381 Dermatological therapy 382 Bacterial diseases 384

Internal malignancy and the skin 420 Sunlight and the skin 426 Pigmentation 427 Pressure sores 428 Leg ulceration 429

description of those skin diseases that are either common or important. Diseases presenting as changes in the skin but whose major impact is on other systems, are discussed in other chapters. Because of the great importance of the visual aspects of dermatology, a colour atlas or, whenever possible, patients, should be viewed in conjunction with the text. Therapy is considered in terms of general principles.

BIOLOGY OF THE SKIN The skin is a protective barrier between a hostile environment and internal tissues. Its principal functions are to reduce the loss of water, electrolytes and other solutes, and the entry of unwanted molecules, microbes and radiation. It has a complex and specialized sensory innervation and contains a network of antigen-processing cells. The cutaneous vasculature and sweat glands are vital to heat regulation. Vitamin D is synthesized in the skin. By virtue of its smell, colour and texture, the skin has psychological and sexual roles. The skin is composed of three layers: epidermis, dermis and fat (Fig. 10.1).

Superficial fungal diseases 389

Acne vulgaris 431

Viral diseases 393

Hyperhidrosis 431

Arthropods and the skin 394

Hidradenitis suppurativa 432

Psoriasis 395

Rosacea 432

Epidermis

Eczema and dermatitis 399

Pruritus 433

Ichthyosis 405

Alopecia 433

Reaction patterns 405

Hirsutism and hypertrichosis 434

Blistering diseases 410

Nails and disease 435

Miscellaneous disorders of unknown cause 413

Psychological causes of skin disease 436

Skin changes in diabetes 415

Genodermatoses 436

The epidermis is a stratified epithelium comprised mainly of keratinocytes, which are of ectodermal origin. These divide in the basal layer, where they are separated by the basal lamina from the dermis, progressively produce the sulphur-rich fibrous protein keratin, and become flattened and dead by the time they reach the surface. These cells have a lipid-rich envelope and are linked to one another by specialized cell junctions, notably the desmosomes, and in health the horny layer is tough, flexible and relatively impermeable. At the surface they are constantly shed as squames. Removal of the horny layer virtually destroys the barrier function of the skin with respect to water and solutes. The effectiveness of the horny layer is reduced if its production is faulty (e.g. in psoriasis), if the water content drops below a critical level (e.g. chapping), or if it is damaged (e.g. by detergents and lipid solvents). Resident in the epidermis are the melanocytes, Langerhans' cells and Merkel's cells.

Drug eruptions 416

Skin disorders are a common reason for seeking medical advice (Table 10.1). Apart from diseases that primarily affect the skin, many systemic diseases may produce dermatological manifestations. When a skin disease is widespread and interfering with skin functions, systemic consequences follow - sometimes, as in untreated pemphigus vulgaris, with fatal results. More limited skin diseases can cause distress or hardship. The psychological impact of readily visible skin lesions, even if biologically trivial, should not be underestimated. Another common problem associated with skin disorders is fear (often not expressed) that a skin disease is contagious, or a manifestation of cancer. This chapter gives a brief account of the biology of the skin, the principles of diagnosis and therapy, and then a

• Melanocytes migrate early in fetal life from the neural crest and become established along the basal layer, so that there is approximately one melanocyte per 10 basal keratinocytes. Via their dendritic processes, the yellow or brown-pigmented protein melanin is passed in membrane-bound particles, called melanosomes, into the keratinocytes; here most of it is distributed over the surface of the nucleus that faces the sun. Melanin absorbs ultraviolet radiation and thus helps protect against DNA damage. In its absence, in albinism, skin cancer on exposed sites is common at an early age.

379

TABLE 10.1 Prevalence of skin diseases* Rates of consultation (general practice) per 1000/year All skin diseases Alopic dermatitis and related conditions Contact dermatitis and other eczema Viral warts Psoriasis Urticaria Chronic ulcer Malignant neoplasms Some comparisons Asthma Osteoarthritis Irritable bowel Diseases of stomach (functional) Rheumatoid (and other inflammatory) arthritis

229 43 27

18 11

9 4 1.4

91 58 32 22 11

* Date from Morbidity Statistics from General Practice, Fourth National Study 1991-2. OPCS, HMSO, 1995.

FIG. 10.1 Cross-section of the skin

RECENT ADVANCES IN MEASURING DISABILITY IN SKIN DISEASE An important new approach to the assessment of skin disease has been the application of techniques that measure the impact of the process on the quality of life (see Further Reading). Such methods are being used in other branches of medicine too, and can help doctors: • Gain an understanding of what having a chronic skin disease such as psoriasis or eczema is really like; • Gauge the usefulness of treatments in terms of improvements in life quality (which may improve the ability to judge therapy in terms of risk/benefit ratios and cost-effectiveness); • Compare the effects of skin disease with other medical problems and apportion resources.

FURTHER READING Finlay AY 1997 Quality of life measurement in dermatology: a practical guide. Br J Dermatol 136:305-314. Anderson RT, Rajagopalan R 1997 Development and validation of a quality of life instrument for cutaneous diseases. J Am Acad Dermatol 37:41-50.

1

380

MCQ 10.1

• Langerhans' cells, also dendritic, are of bone marrow origin. They are located in the mid-epidermis, function as antigen-trapping and antigen-presenting cells, and can migrate to regional lymph nodes. Together with lymphocytes, they constitute a vital outpost of the immune system. • Merkel's cells, best recognized on electron microscopy by their dense-cored granules, are closely associated with sensory nerve endings, and may represent a mechanoreceptor system. They are most numerous on the lips, on digital pads and in hair follicles. The epidermis is separated from, and attached to, the dermis by a basal lamina, across which all nutrients must pass. A component of the basal lamina is the target of immunological attack in the blistering disease bullous pemphigoid (p. 411).

Dermis The dermis is a three-dimensional fibrous tissue network of collagen and elastin associated with water-rich glycosaminoglycans, in which are embedded the blood and lymphatic vessels, neural elements and epidermal appendages - sweat glands, sebaceous glands and hair follicles (Fig. 10.1). The dermis is manufactured by fibroblasts, and other cells seen in normal tissue include dermal phagocytes and mast cells. The dermis has viscoelastic properties,

allowing body movements without permanent distortion, yet considerable mechanical strength. It is essential to the wellbeing and repair of the epidermis. The superficial part of the dermis, containing capillaries and nerve endings, is more loosely woven than deeper layers, and projects into the undersurface of the epidermis as papillae. Blood flow through the skin can be greatly in excess of nutritional demands, and skin vasculature contributes to thermoregulation and the maintenance of blood pressure. There are networks of blood vessels at the levels of the deep fascia, the fat-dermis junction and in the papillary dermis, where capillary loops follow the contour of the dermoepidermal junction. There are also numerous arteriovenous shunts that allow blood to bypass capillaries, thereby conserving heat. Vascular control is by autonomic nerves and chemical mediators. The skin is also provided with an extensive plexus of lymphatics, which act both as a drainage system for lymph and in the recirculation of lymphocytes. There are afferent sensory nerves subserving touch, pain, temperature and itch, and also the specialized Meissner and Pacinian corpuscles, which are probably mechanoreceptors. The autonomic supply is sympathetic, with adrenergic, cholinergic and purinergic terminals. Collagen and elastin disorders are described on page 71.

Epidermal appendages Eccrine sweat glands consist of secretory coils connected by straight ducts to the surface. There is a rich blood supply and a predominantly cholinergic sympathetic innervation. Because of the sodium-resorptive properties of the ducts, sweat is hypotonic. The basal level of fluid loss through the skin of about 600mL/day can be greatly increased in response to the need for heat loss. Apocrine sweat glands are found in the axillae, anogenital regions, on the breast and on the scalp. They open into hair follicles and on to the skin surface. Their secretion is low in volume, and after bacterial alteration it has a characteristic odour. Very similar glands produce wax in the ears and occur on the eyelids. The pilosebaceous follicles produce both hairs and sebaceous glands. In humans hair is largely of cultural and sexual significance. There are three types: lanugo, which covers the fetus; vellus, which is fine and short; and terminal hair, which is coarse and occurs on the scalp, eyebrows, axillae and pubic regions. A residual protective value of hair is seen when its absence in bald males is accompanied in later life by sun-induced neoplasms. An important characteristic of hair follicles is their cyclical activity: for each follicle a period of active growth (anagen) is followed by a transition period (catagen) and then a resting phase (telogen), when the hair is shed and a new hair begins. Individual follicles are normally asynchronous, hence humans do not moult. Nails are hard plates of keratin which protect the ends of the digits and facilitate many of the functions of the hands. 1

FURTHER READING ON BIOLOGY OF THE SKIN

10

Eady R A J , Leigh I M, Pope F M 1998 Anatomy and organisation of human skin. In: Champion R H et al, eds. Textbook of Dermatology, 6th edn. Oxford: Blackwell Science, pp. 37-111.

EXAMINATION OF THE SKIN It is essential to examine all the skin, together with the mouth, genitalia, hair and nails, in a good light. The skin should be palpated as well as inspected, to appreciate the thickness of lesions and of the whole skin. Being touched can help some patients to shed a 'leper complex'. Skin lesions should be described in terms of: • The individual characteristics (shape, colour, etc.) • Their arrangement in relation to each other (if any) • Their distribution. The terms used have precise meanings (Table 10.2). Distribution The following points should be noted: • Is the rash localized, regional or generalized? • Is there symmetry? Endogenous disorders are often symmetrical. • Is there localization by exposure to light, cold or some other environmental factor? • Has non-specific trauma, e.g. scratching, caused new lesions (Koebner's phenomenon), as in psoriasis and lichen planus? • Is there an underlying anatomical basis, e.g. the vascular or nerve supply, the flexures or hair follicles? Many conditions have characteristic patterns of distribution, the basis for which is obscure, e.g. dermatitis herpetiformis on elbows and knees, scalp, shoulders and buttocks.

DERMATOLOGICAL DIAGNOSIS There is a great variety of dermatological diagnoses. However, the practised observer can describe changes that suggest a diagnosis or differential diagnostic group. The history is important, narrowing the possibilities (Table 10.3). Diagnostic procedures and investigations may be useful. Diascopy When the skin is pressed with a microscope slide or Perspex spatula, purpura (which does not blanch) can be distinguished from erythema (which does). Nikolsky sign The Nikolsky sign involves the production or extension of a blistering process by the combination of pressure with a

381

TABLE 10.2 Terms used to describe skin lesions Skin lesion Appearance Erythema Macule Papule Plaque Nodule Vesicle Bulla Pustule Weal Angioedema Purpura

Telangiectasia Atrophy Sclerosis Ulcer Excoriation Crust Scaling Burrow

Definition Redness due to vasodilatation Flat, circumscribed, impalpable area of colour change. There may be scaling, as in pityriasis versicolor Small (less than 0.5cm), solid, elevated lesion, most of which is above the plane of the skin Elevated lesion with relatively large surface area in relation to height Large (greater than 0.5cm diameter) solid lesion Blister less than 0.5cm diameter Blister greater than 0.5cm diameter Elevation of the skin containing fluid and leukocytes, usually yellow or green in colour, but not necessarily infective Elevation due to dermal oedema, lasting minutes to several hours Massive oedematous reaction in loose dermis or subcutaneous tissue Extravasated red cells (see Diascopy, p. 381). Small purpuric lesions are called petechiae, large ones ecchymoses. They are not obliterated by pressure. The colour varies from red, through purple and brown, to yellow, depending on the age of the lesion Permanent dilatation of capillaries Thinning, often with loss of normal skin markings, and increased transparency Hardening due to changes in the dermis Loss of (at least) the epidermis Scratch mark Dried exudate Desquamating horny layer Irregular linear elevation of horny layer, characteristic of scabies

Shape and arrangement Linear In a line Annular Ring-shaped Erythematous ring or rings separated by relative pallor with a Target purplish centre Herpetiform Clustered, like herpes simplex Zosteriform Clustered and following a dermatome Reticular Net-like

sliding action. It is characteristic of pemphigus and some other blistering diseases.

382

Wood's light Wood's light is produced by a long-wave UV lamp, which shows red fluorescence with erythrasma, green fluorescence with some common scalp fungal infections, and pale yellow fluorescence with pityriasis versicolor. It enhances the pallor of vitiligo and is helpful in screening for the ash-leaf macules of tuberous sclerosis.

TABLE 10.3 Taking a history History of the skin eruption When and where did the problem start? Has it spread, and how? Has it changed in character? What was it like if different before? Does it itch? Does it come and go? What has been used or taken for it (prescribed and self-medication)? Have any other medications - oral, injected or topical - been used? Is there any relation to sunlight? Any past history of skin disease? Does anyone the patient knows have anything similar? What does the patient think is the cause? Background Age, sex, race. Pregnant? Pubertal? General medical, past and family history Personal and family history of eczema, asthma or hay fever Social history: occupation, hobbies, travel Any known allergies?

Biopsy Biopsies are carried out for histopathology and, when relevant, immunological studies, microbiological culture and electron microscopy, cell culture and molecular biological studies such as gene analysis. Mycological techniques Mycological techniques are described on page 389. Patch testing Patch testing is described under 'Contact

dermatitis'

(p. 400). Scabies mite scraping Scabies mite scraping is described on page 395.

DERMATOLOGICAL THERAPY Dermatological treatment includes topical and systemic pharmacy, physical modalities such as liquid nitrogen cryotherapy, ultraviolet radiation, radiotherapy, laser and surgical procedures, and dietary measures. Topical therapy is the traditional province of the dermatologist.

Vehicles Active ingredients are applied to the skin in a vehicle (Table 10.4). A vehicle may be used therapeutically: a lotion to cool and dry, a cream to soothe, an ointment to moisturize dry skin, and a paste to protect. When an active ingredient is included, the vehicle must

TABLE 10.4 Composition and properties of vehicles for topical treatment Vehicle

Composition

Properties

Lotion

Liquid (water or alcohol)

Cools inflamed skin. Useful in hairy areas. Miscible with exudate

Shake lotion

Water + powder

Powder increases area for evaporation, so has cooling properties

Cream

Water + grease + emulsifier

Aqueous phase gives some cooling effect and miscibility with exudate. Grease increases hydration. Properties will vary depending on whether the emulsion is oil-in-water or water-in-oil

Ointment

Grease

Moisturizes dry skin by trapping water passing through the epidermis. Hydrophobic (e.g. vaseline) or hydrophilic (e.g. lanolin)

Paste

Grease + powder

Moisturizing and protective

Dusting powder

Powder

Reduces friction. Absorbs water

deliver it appropriately and maintain its integrity against oxidation and bacterial contamination. Therefore vehicles, especially creams, often contain preservatives and stabilizers. Some preparations contain substances to enhance drug penetration, e.g. urea. When adverse reactions to topical agents occur, it should be remembered that not only the active ingredient but also the components of the vehicle should be considered as possible causes. For most active ingredients the major barrier to penetration is the horny layer. Penetration is greater through facial and genital skin and flexures (e.g. axillae, crural folds) and is increased by occlusion (e.g. by rubber gloves), diseases that alter the horny layer (e.g. eczema and psoriasis), and pharmaceutical penetration enhancers such as urea. In contrast, when the horny layer is very thick, as on the palms and soles, there is reduced penetration.

TABLE 10.5 Potency of topical corticosteroids Group

Potency

Examples

1 2

Very high High

3 4

Moderate Low

Clobetasol propionate Betamethasone valerate Hydrocortisone butyrate Clobetasone butyrate Hydrocortisone

10

TABLE 10.6 Adverse effects of topical steroids Adverse effects

Factors increasing risk

Pituitary-adrenal axis suppression; growth suppression in children

Potency group 1 Potency groups 1 and 2 with occlusion Widespread skin inflammation Infancy Hepatic failure

Spread of skin infection

Overt infection Body folds

Atrophy (thin, red, fragile skin)

High risk sites, e.g. face, body folds Occlusion Prolonged and continuous use

Striae (stretch marks)

Potency groups 1 and 2 Body folds

Increased hair growth (hypertrichosis)

Potency groups 1 and 2 Prolonged use

Peri-oral dermatitis

Face

Exacerbation of skin disease on withdrawal

Psoriasis Facial dermatoses

tration, the vehicle, and also by factors favouring increased absorption. Adverse effects (Table 10.6) are very unlikely with topical 1 % hydrocortisone.

Other topically used drugs

Topical corticosteroids

A selection of drugs used topically is shown in Table 10.7.

Topical corticosteroids have anti-inflammatory and other actions, with useful suppressive effects in a number of skin disorders, but they do not cure disease. They can worsen skin infection and modify physical signs, and the more potent preparations easily produce side-effects. In general, the efficacy of topical steroids goes hand in hand with the ability to cause side-effects, and a preparation of the lowest potency that is effective should be used (Table 10.5). Applications are usually once or twice daily, with bland creams or ointments at other times if necessary. Efficacy is determined by the steroid molecule, its concen-

Cryotherapy Many benign and some malignant skin lesions can be quickly and successfully treated by accurately applied low temperature. The most satisfactory agent is liquid nitrogen, whose temperature is -196°C, applied on a cottonwool swab, by a spray or a probe. The procedure can be painful and is often followed by swelling and blistering, but longterm cosmetic results are generally good. Hypopigmentation and some sensory loss can be troublesome consequences.

383

TABLE 10.7 Some topical preparations used in skin diseases Topical drug

Use

Tars Dithranol Calcipotriol Benzoyl peroxide Adapalene Metronidazole Calamine Aminobenzoic acid esters Salicylic acid

Chronic eczema, psoriasis Psoriasis Psoriasis Acne Acne Rosacea Pruritus Sunscreens Viral warts, other localized hyperkeratotic conditions Viral warts Scabies Some dermatophyte infections Some dermatophyte and yeast infections; seborrhoeic dermatitis Dermatophyte infections Candidiasis Localized superficial Gram-positive bacterial infection Some herpes simplex and zoster infections

Podophyllum Permethrin Compound benzole acid ointment Antifungal imidazoles Terbinafine Nystatin and amphotericin B Antibiotics, e.g. fusidic acid, mupirocin Aciclovir

BACTERIAL DISEASES Bacterial infection can produce skin lesions at or near the portal of entry, as in impetigo. Lesions may also be due to organisms deposited in the skin during septicaemia, bloodborne spread of a toxin, and hypersensitivity phenomena, such as the vasculitic Osier's nodes and splinter haemorrhages of subacute bacterial endocarditis's. Normal intact skin is colonized by diphtheroids and nonpathogenic staphylococci and, in some sites such as the anterior nares, axilla and perineum, by potentially pathogenic Staphylococcus aureus. In moist areas there are often some Gram-negative bacilli. Overt infection nearly always follows some kind of injury, albeit trivial. If it is associated with some obvious pre-existing skin lesions, such as eczema, it is referred to as secondary infection. Host factors, particularly the immunological status, are important in determining the degree and extent of infection. Some skin diseases, such as hidradenitis suppurativa, are a non-specific reaction to bacteria, which, although contributory, are not the primary cause. The types of staphylococcal and streptococcal skin infection are shown in Table 10.8. 1

384

1

MCQ10.2

2

Fig. 10.1

3

Fig. 10.2

4

Fig.10.3

RECENT ADVANCES IN DERMATOLOGICAL THERAPY There have been a number of important advances in the treatment of skin diseases in the last 5 years. • The introduction of newer 'azole' derivatives (itraconazole, fluconazole) and the allylamine terbinafine has resulted in a wider range of more effective and safer systemic options for the treatment of cutaneous (and other) fungal infections. • Ciclosporin, which suppresses T-lymphocyte activity, has an established role in the therapy of severely affected patients with psoriasis and atopic dermatitis, and can be effective in several other skin diseases in which immune dysfunction is important, e.g. pyoderma gangrenosum. Other drugs with immunomodulatory effects useful in dermatology include tacrolimus (similar in action to ciclosporin) which can be used topically (e.g. for atopic dermatitis) and thalidomide (helpful in Behcet's syndrome and cutaneous lupus erythematosus). • The vitamin D3 analogues can now be regarded as first-line agents in the management of plaque psoriasis. Examples include calcipotriol and tacalcitol. • Retinoids are the most important cell-signalling molecules in dermatology. Two newer retinoids of clinical value are tazarotene for psoriasis and adapalene for acne vulgaris. • Better laser technology has improved the results of treatment for vascular lesions such as port wine stains. Other lasers are used to eliminate tattoos and some pigmented lesions.

Impetigo Impetigo (Fig. 10.2) is a primary superficial bacterial skin infection, initially vesicular or bullous and later crusted. It may only involve the openings of hair follicles and is caused by Staph. aureus, Streptococcus pyogenes group A, or both. In some tropical regions the streptococci can cause nephritis. Bullous impetigo is usually due to phage group II staphylococci. Impetigo occurs predominantly among children and is highly contagious. Spread is mainly from skin to skin, although in streptococcal cases infected individuals may acquire the disease from pharyngeal colonization. Trauma, for example from scratching insect bites, and infestation commonly predispose to the infection. Exposed parts are mainly affected. The early lesions are clear or turbid blisters which soon become crusts, often golden yellow, with a surrounding zone of erythema. In bullous impetigo there are larger blisters with no erythema. The lesions tend to heal centrally but spread peripherally. Regional lymphadenopathy is common. Diagnosis is by Gram stain and culture of a swab, which will help to distinguish impetigo from other lesions such as tinea and varicella.

10

FIG. 10.3 Staphylococcal scalded skin syndrome This shows characteristic crusting and radial fissuring around the mouth. FIG. 10.2 Impetigo: ruptured flaccid blisters and crusts on a red base

TABLE 10.8 Staphylococcal and streptococcal bacterial infections of the skin Due to direct bacterial infection Staphylococcus aureus Impetigo Ecthyma Follicular infections Streptococcus pyogenes Impetigo Ecthyma Erysipelas Cellulitis Necrotizing fasciitis

Due to toxin

Due to hypersensitivity

Staphylococcal scalded skin syndrome Toxic shock syndrome Staphylococcal scarlatina Scarlet fever

Erythema Vasculitis Erythema nodosum

Treatment Gentle removal of crusts is helpful. If localized, a topical antibiotic, such as fusidic acid or mupirocin, is adequate. For widespread impetigo, oral flucloxacillin (or erythromycin if the patient is allergic to penicillin) is required. Improvement of hygiene may be advisable, e.g. not sharing towels.

Ecthyma Ecthyma is a vesiculopurulent condition similar to impetigo, but the infection extends deeper, producing ulceration 2; it is caused by Strep, pyogenes, often with Staph. aureus. The condition is common in hot, humid climates, affects mainly children and the elderly, and is often associated with poor hygiene and/or insect bites. The initial vesicular lesion soon becomes a punched-out ulcer with an overlying crust. The legs are commonly affected. The differential diagnosis includes ecthyma gangrenosum 3, caused by Pseudomonas septicaemia (see below),

and other causes of ulcers (p. 430). Treatment is with a penicillinase-resistant antibiotic, such as flucloxacillin or erythromycin. If there is surrounding cellulitis penicillin should be added.

Staphylococcal scalded skin syndrome Staphylococcal scalded skin syndrome is a generalized skin reaction to bloodborne Staphylococcal epidermolytic toxins, which cause a split in the superficial epidermis; unlike impetigo, there are no organisms in the accumulating fluid. It is caused by a focus of infection of Staph. pyogenes of phage group II, which may be cutaneous but can be elsewhere. Children, especially infants, are most susceptible (Fig. 10.3). Clinical features A few days after the initiating Staphylococcal infection there is a sudden onset of widespread tender erythema with fever. Twelve hours or so later there may be widespread flaccid bullae, and the horny layer begins to detach in sheets. 4 Differential diagnosis Toxic epidermal necrolysis due to drugs (p. 417) can be distinguished by the histology of a blister roof, which shows a deeper level of split in the epidermis in the drug-induced type, and by the history. Treatment Treatment is with parenteral flucloxacillin (or equivalent), and careful attention to fluid balance and temperature regulation.

Erysipelas Erysipelas (Fig. 10.4) is a dermal infection, usually with group A streptococci. There is often a recognizable portal of entry, such as a fungal infection between the toes or a fissure at the corner of the mouth. There may have been a previous upper respiratory tract streptococcal infection. The onset is often sudden, with fever, malaise and rigors,

385

FIG. 10.5 Cellulitis Spreading soft tissue infection due to B haemolytic streptococcuslor which the portal of entry was a venous leg ulcer.

Cellulitis

FIG. 10.4 Erysipelas: tender erythema and oedema

together with a bright red, tender swelling. The face and lower legs are common sites. The skin becomes oedematous and a well-defined, raised, sometimes blistering edge advances rapidly. There are often red streaks due to lymphangitis. Diagnosis Leukocytosis is usual. Surface culture is useless, but the organism can sometimes be cultured from aspirated fluid. There is usually an antibody response to streptococcal proteins, with rising litres of antideoxyribonuclease B; antistreptolysin O is less often elevated. Differential diagnosis Acute contact dermatitis and angioedema of the face may simulate erysipelas but do not produce fever. Herpes zoster, before the outbreak of vesicles, is associated with pain or paraesthesiae. Erysipeloid (p. 387) occurs in meatand fish-handlers, and systemic symptoms are much less severe. Complications Erysipelas is usually self-limiting but can produce suppurative and widespread infection. There is involvement of the lymphatics, and this predisposes to persistent swelling (lymphoedema) and recurrent attacks. Treatment The treatment of choice is penicillin, parenterally in severe infections, unless contraindicated by hypersensitivity, in which case erythromycin or another macrolide should be suitable.

1

386

Fig. 10.4

Cellulitis is a deeper dermal/subcutaneous infection than erysipelas, usually with group A streptococci. In practice there is no absolute distinction between erysipelas and cellulitis. There is usually an obvious portal, e.g. a leg ulcer (Fig. 10.5) or wound, or oedema of lymphatic, venous or renal origin. There is spreading erythematous oedema as in erysipelas, but without the sharply denned edge or vesiculation. There may be similar constitutional symptoms and lymphangitis. Untreated, gangrene may occur. Treatment is the same as for erysipelas, unless the cellulitis is caused by bacteria other than group A streptococci.

Necrotizing fasciitis Necrotizing fasciitis (streptococcal gangrene) is a necrotizing process usually caused by group A streptococci; it involves the deep fascia and vessels within it, with secondary death of overlying skin. This uncommon process can occur in healthy subjects, but arterial insufficiency and diabetes predispose. Necrotizing fasciitis usually begins like cellulitis, but after about 2 days the area becomes purplish, haemorrhagic bullae appear, and there is tissue death. 1 Untreated, there is a high mortality. The diagnosis is often based on clinical suspicion alone, but streptococci can commonly be found in exudate or the blood. Treatment by urgent and adequate surgical debridement is essential. There has been some debate over the antibiotic of choice in necrotizing fasciitis, with some recommending vancomycin or clindamycin. However, most authorities recommend high-dose intravenous benzylpenicillin, with vancomycin being reserved for the penicillin-allergic individual.

Progressive bacterial synergistic gangrene Progressive bacterial synergistic gangrene is gangrenous ulceration due to synergistic infection with microaerophilic

streptococci and Staph. aureus, usually associated with an abdominal or thoracic operation wound. One to 2 weeks after surgery there is a slowly spreading area of ulceration which has a rim of gangrenous skin surrounded by purplish erythema. The main differential diagnosis is pyoderma gangrenosum (see p. 407). Treatment is by surgical debridement plus i.v. highdose penicillin, gentamicin and metronidazole. For the penicillin-allergic patient, clindamycin and gentamicin are recommended.

10

Follicular infections A furuncle (boil) is an acute necrotizing infection of a follicle. A carbuncle is more extensive, involving contiguous follicles and the tissues around them. Infection of the hair follicle opening is termed superficial folliculitis. The usual infecting organism is Staph. aureus, and there is often nasal, axillary or perineal carriage of the organism. Friction, as at the nape of the neck, and moisture, as in the flexures, are important predisposing factors. Obesity, poor hygiene, widespread skin disease, immune deficiency and diabetes mellitus may predispose in severe cases. Superficial folliculitis can be non-infective; causes include contact with oils and other irritant chemicals. Furuncles are tender red papules or nodules, which become pustular centrally and often heal with some scarring. Furunculosis may be chronic, as in sycosis barbae, a pustular eruption due to Staph. aureus that occurs on the male face. Carbuncles are often associated with fever and malaise, and with an underlying systemic illness such as diabetes or immune suppression. Follicular infections can produce bacteraemia and, rarely, bony and cerebral abscesses, and endocarditis. Swabs should be taken from pus and from the anterior nares, axilla and perineum of both the patient and household contacts (potential reservoirs of infection) in chronic cases. Differential diagnosis The differential diagnosis includes dermatophyte infection (the fungi can be demonstrated by microscopy and culture of hairs) and pseudofolliculitis, a non-infective papulopustular condition on the face and neck caused by cut hairs growing back into the skin. The latter is very common in black people because their hair is tightly curled. Acne vulgaris (p. 431) can usually be distinguished by the presence of comedones, and the pustules are sterile. Hidradenitis suppurativa is discussed on page 432. In anthrax (see Ch. 9) there is a haemorrhagic crust and vesicular margin, and a swab will establish the diagnosis. Treatment Furuncles and carbuncles may need incision and drainage. Superficial staphylococcal folliculitis should respond to flucloxacillin (or an equivalent) for 1-2 weeks, but chronic cases may need treatment for longer, and treatment of

FIG. 10.6 Erythrasma Well-defined brownish erythema which fluoresces pink by Wood's light.

reservoir sites, e.g. chlorhexidine for the skin and mupirocin for the nose.

Erythrasma Erythrasma (Fig. 10.6) is a common surface infection caused by Corynebacterium minutissimum. Sharply defined red or reddish-brown patches, sometimes with slight scaling, are seen in the axillae, groins, toe webs, and sometimes other body folds. There are usually no symptoms. The diagnosis is readily made with Wood's light; the affected skin fluoresces pink. Most cases respond to a topical imidazole, e.g. clotrimazole, but if the condition is widespread oral erythromycin for 10 days may be more effective.

Erysipeloid Erysipeloid is an acute cutaneous infection with Erysipelothrix insidiosa. This Gram-positive bacillus causes infection in several animals and in salt-water fish. Transfer to the skin occurs mainly to veterinary surgeons, butchers, fish-handlers and housewives.

387

The hand is the common site. A few days after inoculation there is a well-defined, slowly spreading, dusky erythematous oedema with little or no general upset (in contrast with erysipelas). Treatment is with penicillin or tetracycline.

Pseudomonas infection of the skin Pseudomonas aeruginosa is not normally found on the skin but can become pathogenic, with the production of a characteristic bluish-green pus and a fruity odour in circumstances of increased moisture. Examples may be found at the base of the nail in those whose hands are often immersed in water, in toe spaces, beneath dressings for ulcers, and very importantly on raw surfaces following burns. Prior eradication of Gram-positive flora may also predispose. A pruritic pseudomonas folliculitis can be acquired from contaminated whirlpool baths. Localized infection usually responds to simple antiseptic measures, such as acetic acid, silver nitrate or silver sulfadiazine, together with drying. Pseudomonas septicaemia can produce solitary or widespread vesicles with surrounding red haloes. Subsequent necrosis, often with haemorrhage into the tissues, producing black crusts, is known as ecthyma gangrenosum. The lesions, as well as the blood, should yield Pseudomonas on culture, and prompt treatment can be life-saving.

Mycobacterial and treponemal diseases Tuberculosis (see also p. 640) leprosy (p. 325) and syphilis (p. 458) all have important cutaneous manifestations. Cutaneous tuberculosis Involvement of the skin by Mycobacterium tuberculosis is relatively uncommon in developed countries but is often seen in the developing world, where the overall prevalence of tuberculosis is much higher. The condition still needs to be considered in the west, particularly in those who have migrated from the Indian subcontinent and Africa. Many descriptive terms have been applied to the wide variety of clinical presentations that can occur, and to some extent these reflect the mode by which the skin may be affected: haematogenous spread from a deep source (most lupus vulgaris); direct spread from underlying organs, e.g. lymph nodes, bones, respiratory or gastrointestinal tract (scrofuloderma, tuberculosis cutis orificialis); direct inoculation from outside (some lupus vulgaris, tuberculosis verrucosa cutis and warty TB). The commonest form is lupus vulgaris, which presents as an area of indurated, reddish-brown skin, characteristically on the head and neck (Fig. 10.7), although lesions may

1

388

Fig. 10.5

FIG. 10.7 Typical lupus vulgaris involving ear and cheek

occur elsewhere. The lesions gradually spread and have a tendency to produce scar tissue, which results in distortion of normal anatomy and loss of hair follicles in hair-bearing areas. When the microorganism is inoculated directly into the skin, lesions sometimes have a warty surface. Tuberculosis may result in a number of skin changes in which the microorganisms are not directly involved (known as tuberculides}: • Erythema nodosum (see p. 407), in which the lesions usually appear in association with primary infection; • Erythema induratum (or Bazin's disease), in which the nodular, ulcerative lesions appear on the lower legs, involving the calves as well as the shins; 1 • The rare papulonecrotic and lichenoid forms. Investigation of suspected cutaneous tuberculosis involves biopsies for histology and culture and a search for involvement of internal organs (see Ch. 13). Treatment should be with standard antituberculous chemotherapy (p. 645), even if a deeper focus is not found. Atypical mycobacterial infections The skin may become infected by direct inoculation of other mycobacteria, especially Mycobacterium marinum, an organism responsible for a systemic infection in fish. An indurated nodule or nodules appear, usually on the hand or arm of a fish-fancier (Fig. 10.8). In some patients, further nodules appear along the lines of lymphatic drainage (socalled sporotrichoid spread). Occasionally lesions may be more widespread, especially in those who catch the organ-

10

FIG. 10.8 'Fish-tank granuloma' These lesions are caused by mycobacteria acquired from fish.

ism from swimming. Biopsy for histology and culture should be performed, but the tissue must be incubated on special media at low temperature for the organism to grow. Lesions often resolve spontaneously, but this is a slow process and therefore treatment with antibiotics may be indicated. Various drugs have their advocates: minocycline, co-trimoxazole and rifampicin are all considered to be effective. Treatment should be for about 6 weeks.

Fungal infections involving the skin are broadly classified into superficial and deep. The former include the ringworm fungi (dermatophytes), candidiasis and pityriasis versicolor. Except for Candida, these fungi rarely invade deeper than the horny layer of the epidermis. Nails and hair, as well as the skin surface, may be involved (see Fig. 10.12, p. 391). The clinical manifestations depend partly on the infecting species, on the body part affected, and on the host response to substances diffusing from the fungi into the skin. Deep fungal infections are discussed in Chapter 9. Diagnosis Diagnosis is by microscopic examination of scrapings to show fungal hyphae, and species identification is by culture. The dermatophytes are grown on Sabouraud's medium. With skin lesions scale is collected by scraping with a blade, hair is plucked and clippings are taken from nails. Material is sent to the laboratory dry in folded paper. Microscopy is performed on small samples after clearing with a few drops of 30% potassium hydroxide. For some species of scalp dermatophyte infection and for pityriasis versicolor the Wood's light is valuable (p. 382).

Tinea The fungi that cause the characteristic annular lesions

FIG. 10.9 The annular lesions of tinea (ringworm)

known as tinea or ringworm (Fig. 10.9) are also known as dermatophytes, as they have the ability to digest keratin. There are three genera - Trichophyton, Microsporum and Epidermophyton - and numerous species, which are recognized by their culture characteristics. Humans are the primary host for some (the anthropophilic species), but with others the fungi are incidental pathogens to humans, being primarily animal parasites (zoophilic) or found in soil (geophilic). The immune response is most marked against zoophilic species, and these tend not to recur. With very inflammatory ringworm there can be so-called 'id' reactions, owing to the immune response. These are widespread, often follicular, papular lesions which do not contain fungi. Probably for similar reasons, a vesicular eruption on the hands can be associated with ringworm of the feet. The distinctive appearance of ringworm infections can be greatly modified by topical steroids. Treatment Successful treatment requires attention to sources of reinfection and measures to deal with acutely inflamed macerated skin, such as aluminium acetate soaks, as well as specific antifungal preparations. Topical agents, including Whitfield's ointment, and various imidazoles are suitable for localized skin infection. The most rapid, effective and expensive agent is terbinafme. Oral terbinafine is the treatment of choice for hair, nail, widespread and chronic skin infections, but is not yet licensed in the UK for use in children. Itraconazole and griseofulvin are alternatives. A weak topical steroid, together with antifungal therapy, may be useful if there is much inflammation. Tinea pedis Tinea pedis is a dermatophyte infection of the feet, some-

389

FIG. 10.10 Tinea pedis Macerated hyperkeratosis and scaling in the lateral toespace.

times with concurrent bacterial infection. This very common, often chronic condition is frequently acquired where bathing facilities are shared, and infection is favoured by maceration. It is much commoner in males than in females. The most common pattern is an itchy, inflamed, fissured, moist toe space, usually between the fourth and fifth toes (Fig. 10.10). It is often unilateral. A vesicular pattern may occur involving the instep, dorsum of the foot or sides of the toes. A very chronic pattern may be seen, with hyperkeratosis and fine white scaling accentuating skin creases on the soles. 1 Tinea pedis may be accompanied by cellulitis. Tinea pedis can resemble eczema, contact dermatitis, candidal infection and psoriasis. Endogenous eczema and psoriasis are more likely to be bilateral. Most toe space infections respond to a topical antifungal agent, but extensive involvement of the soles needs oral treatment. Regular cleansing of communal bathing facilities and prophylactic use of antifungal powders help reduce transmission of infection. Tinea cruris Tinea cruris is a dermatophyte infection of the groins and adjacent skin. Like tinea pedis, with which it is often associated, tinea cruris is much more common in males. Itching is very common. There is inflammation, with a well-defined margin which is scaly, vesicular or occasionally pustular, often with central clearing. The differential diagnosis includes candidiasis, in which there is a less well defined edge and outlying tiny pustules. Erythrasma is not usually inflammatory. Intertrigo, an inflammatory process common in the obese, in which friction sweating and minor bacterial infection cause erythema

1

390

Fig. 10.6 2 Fig. 10.7

FIG. 10.11 Tinea corporis Ring-shaped lesions with erythematous scaly borders and some central clearing.

TABLE 10.9 Types of tinea capitis Representative species

Wood's light fluorescence

Patchy baldness, with broken-off hairs and scaling of the scalp

Microsporum canis

Blue-green

Microsporum audouini

Green

Puppies, kittens Humans

Patchy baldness with black dots

Trichophyton tonsurans Trichophyton violaceum Trichophyton verrucosum Trichophyton schoenleinii

-

Humans

-

Humans

-

Cattle

Dull green

Humans

Clinical pattern

Kerion Favus

Source

in body folds, can mimic tinea cruris, but scrapings will be negative. Eczema and psoriasis can usually be recognized by their characteristic appearances on non-flexural skin elsewhere. For limited infection treatment with a topical imidazole is sufficient. For extensive involvement, topical steroidtreated cases, or when therapy fails, however, oral therapy is needed for 3-6 weeks. Tinea corporis Tinea corporis is often of animal origin, and is then selflimiting. Lesions are usually multiple and often, but not always, ring-shaped (Fig. 10.11). Because the appearances are so variable, scrapings should be taken from any red, scaly rash that cannot be readily diagnosed. If the disease is limited in extent, a topical agent is usually sufficient; if widespread, then oral terbinafine or itraconazole is preferable. Tinea capitis Many different fungal species can infect hair and scalp skin (Table 10.9). In the UK, tinea capitis is nearly always a

10

FIG. 10.12 Distal thickening and yellow discoloration due to fungal infection FIG. 10.13 Candidiasis, a superficial fungal infection

disease of children. Affected hairs may break off a few millimetres above the scalp, producing short stubble, or flush with the scalp, giving bald patches with black dots in dark-haired patients. 0 When there is a marked degree of inflammation, producing an appearance like that of a carbuncle, the lesion is called a kerion, and in this circumstance there may be concurrent staphylococcal infection. An appearance known as favus is now very rare in the UK. In this form yellowish cup-shaped crusts develop, together with hair loss. Both kerion and favus can produce permanent alopecia. Fungal infection can be distinguished from other causes of patchy hair loss by microscopy and culture.

Tinea of the hand A diffuse hyperkeratosis with fine powdery accentuation of the palmar creases, usually of only one hand, is the commonest pattern of tinea of the hand. If there is diagnostic doubt, scrapings should be taken.

Tinea of nails In many cases nail infection (onychomycosis) occurs in conjunction with fungal disease elsewhere. Predisposing factors include previous trauma and poor peripheral circulation. The toenails are more commonly affected than fingernails. The changes spread proximally from the free edge. The nailplate becomes discoloured and usually thickened, and may crumble away. It may also separate from the nailbed (onycholysis) (Fig. 10.12). Confirmation of the diagnosis by microscopy and culture of clippings should precede treatment. The differential diagnosis includes psoriasis, in which there is often a distinctive pattern of fine pitting. Candidal infection produces changes that begin proximally. Eczema can simulate fungal infection. Treatment is with oral terbinafine or griseofulvin.

Candidiasis Most human candidal infection is caused by Candida albicans (Fig. 10.13). This yeast is a common commensal in the

TABLE 10,10 Some factors facilitating candidal infection Local Maceration of skin Topical steroids Dentures Poor oral hygiene High oral carbohydrate levels

General Debility Extremes of age Immunosuppression by drugs or disease, particularly HIV Broad-spectrum antibiotics Corticosteroids Diabetes mellitus Iron deficiency Cushing's syndrome Hypocalcaemia/hypoparathyroidism Pregnancy

gastrointestinal tract, mouth and vagina, but not on the skin. However, it may become pathogenic (Table 10.10). Unlike the dermatophyte fungi, Candida invades living tissue. In some circumstances serious systemic infection can occur (see also p. 332). Some of the clinical patterns of candidal infection of the skin and mucosal surfaces are shown in Table 10.11. When infection is widespread and refractory to treatment this may be due to an immune deficiency. The skin lesions in such patients often resemble ringworm, and nails as well as nailfolds tend to be affected. There are several clinical/genetic types. In group I there is a well-defined major primary immune defect, e.g. Swiss-type agammaglobulinaemia. Group II comprises those cases without such a clear-cut primary immune defect. Subgroups include: • The Candida endocrinopathy syndrome, with hypoparathyroidism and Addison's disease being the commoner associated organ-specific autoimmune diseases • Autosomal recessive • Autosomal dominant

391

TABLE 10.11 Clinical patterns of Candida infection Type Oral Thrush Atrophic candidiasis Chronic hyperplastic candidiasis Angular stomatitis Skin Flexural

Anogenilal Vulvo-vaginitis Balanitis

Nails Paronychia Onychia

Pattern of infection

Features

White patches (Fig. 10.9) Painful red atrophic mucous membrane Thickened white adherent plaques Soreness and fissuring at the angles

Can be scraped off to reveal inflamed mucous membrane Common with antibiotic treatment and dentures Needs differentiating from leukoplakia Occurs in folds, usually due to ill-fitting old dentures

Sore, red, marginated skin, often with outlying pustules

Due to maceration

Itchy, sore with curd-like discharge Tiny red papules and pustules on glans and prepuce

Commoner in pregnancy

Painful swollen nail folds with loss of adhesion of cuticle Thickened, discoloured nails

Maceration predisposes. May have poor peripheral circulation Uncommon, often associated with immune detects

Usually from sexual partner

• Diffuse (severe) mucocutaneous candidiasis • Late-onset candidiasis, in which thymoma and HIV infection should be considered. Treatment of Candida infections Any remediable predisposing factors should be dealt with. Locally active agents include nystatin, amphotericin B and the imidazoles. For chronic oral candidiasis prolonged treatment is often necessary. Oral agents, including itraconazole and fluconazole, are usually required for chronic mucocutaneous candidiasis, a common problem in patients with symptomatic HIV infection. Long-term treatment is required.

Pityriasis versicolor Pityriasis (tinea) versicolor (Fig. 10.14) is a superficial infection caused by the mycelial form of the commensal yeast Pityrosporum orbiculare (also known as Malassezia furfur).

1 MCQ 10.3 2 Figs 10.8,10.9 392

FIG. 10.14 Slightly scaly macules of pityriasis versicolor; brown on a white skin (A) and pale on a tanned skin (B)

For most patients who have the disorder it is assumed that there has been a change in the pathogenicity of their surface commensal yeasts. Young adults are the most affected, and it is seen more in the tropics and subtropics. The distinctive hyper- and hypopigmentation is due to the diffusion into the epidermis of azelaic acid, a fatty acid from the fungus, which affects the melanocytes. Clinical features The upper trunk, arms and neck are the commonest sites. Lesions are macular and sharply demarcated, and have fine scaling. They may become confluent over large areas. On sun-protected 'white' skin the patches are skin-coloured or pale brown, but after sun exposure and in darker races they are hypopigmented. Scrapings show a mixture of clustered spherical yeasts and short mycelia. Under Wood's light the affected areas

fluoresce, and it is often apparent that the infection is more widespread than can be seen by visible light. Treatment Selenium sulphide 2.5% in a detergent base applied to affected areas and the scalp (a reservoir of yeasts), left on overnight and repeated a week later, usually clears the condition. Alternatives include topical imidazoles and oral itraconazole. 1 FURTHER READING IN SUPERFICIAL FUNGAL DISEASES Elewski B E. Hay R J, eds. 1999 Novel treatment strategies for superficial mycoses. J Am Acad Dermatol 40: Sl-42. Evans G V, Dodman B, Williamson D M, Brown G J, Bowen R G 1993 Comparison of terbinafine and clotrimazole in treating tinea pedis. Br Med J 307: 645-647.

VIRAL DISEASES

dysplasia verruciformis, predispose to squamous carcinoma. The hands, feet, face, around the knees, perianal and genital skin are the commonest sites for warts (Table 10.12). Because most warts resolve spontaneously, caution is needed in assessing therapy, and methods such as scalpel surgery and radiotherapy should be abandoned. For many non-genital warts, salicylic acid paints (or plasters for large plantar warts) are effective. Cryotherapy (p. 383) is uncomfortable but often more rapidly successful. Recently, intralesional bleomycin has proved effective. In selected cases curettage and minimal cautery may be needed. Genital and anal warts are frequently associated with other sexually transmitted diseases, requiring appropriate investigations. These lesions usually respond to liquid nitrogen cryotherapy or careful applications of podophyllin paint at weekly intervals, but this agent should be avoided during pregnancy because of the enhanced risk of absorption and fetal damage.

10

Molluscum contagiosum Skin lesions are a common feature of many viral infections and may represent the direct consequences of virus replication (e.g. warts), the immune response to a virus, or interaction between the two. Only those where the skin or adjacent mucous membranes are the sole target are discussed here; the remainder, including herpes simplex and zoster, 2 are considered in Chapter 9.

Warts Warts arise because of hypertrophy of the prickle cell layer of the skin (or adjacent mucous membranes) induced by human papilloma virus. Warts are transmissible, inoculation being favoured by mild trauma. They are commoner in children and young adults, and are usually self-limiting and benign. However, some subtypes of virus (defined by DNA hybridization techniques), particularly those found in the genital tract and the rare condition epidermo-

Molluscum contagiosum (Fig. 10.16) is a contagious papular eruption caused by a DNA poxvirus. It occurs mainly in children but is also common in AIDS patients. The lesions are smooth, dome-shaped papules, often craterlike, with a central keratinous plug or depression. When multiple, they are typically in groups. Lesions are ultimately self-limiting. Direct microscopic examination of the cheesy contents or an entire curetted lesion reveals the characteristic eggshaped molluscum bodies. (These are epidermal cells filled with virus particles.) Electron microscopy can also be used. Single lesions can mimic benign or malignant epithelial tumours, e.g. keratoacanthoma or basal cell carcinoma. In young children it is often best to leave the lesions. Treatment is local destruction, e.g. with liquid nitrogen or by mildly traumatizing the papules and then painting with iodine.

TABLE 10.12 Viral warts Type of wart

Main site

Appearance

Common Plane (flat)

Hand Face and backs of hands Face and neck Scalp and neck Soles

Dome-shaped, papilliferous surface Flat-topped smooth papules (Fig. 10.15) Elongated Elongated with horny cap Circumscribed, often painful papules or nodules with altered surface pattern, sometimes with black dots Aggregation of numerous adjacent plantar warts Soft, velvety, vegetative papules and plaques

Filiform Digitate Plantar (verruca)

FIG. 10.15 Plane warts Multiple reddish-brown barely elevated lesions with a rough surface.

Mosaic plantar

Soles

Anogenital

Anal and genital regions

393

FIG. 10.17 Scabies burrows on wrists

FIG. 10.16 Domed umbilicated papules of molluscum contagiosum

Orf and milkers' nodules Orf is a poxviral disease of sheep, usually acquired directly from the perioral lesions that occur on lambs. Milkers' nodule is a similar condition acquired from infection of the teats of cows and ulcers in the mouths of calves. It is caused by a parapoxvirus. Several days after animal contact, a red papule appears and evolves into a 1-2 cm diameter smooth, bluish-red nodule, the centre of which may develop a crusted surface. There is often a whitish ring around the central crust with orf nodules 1 There may be lymphangitis and lymphadenopathy, and sometimes a rash on the limbs resembling erythema multiforme (p. 407). Resolution occurs in 3-4 weeks. Diagnosis is usually evident from the history but can be confirmed by electron microscopy. No treatment is required.

ARTHROPODS AND THE SKIN Numerous arthropods can inflict skin lesions on man, by the mechanisms listed below:

394

1

Fig. 10.10

4

Fig. 10.12

2

Fig. 10.11

3

MCQ10.4

• Mechanical trauma (e.g. horsefly, tungiasis) • Injection of toxins (e.g. some spiders, bees, wasps, ants) • Hypersensitivity (local and generalized) to injected materials (e.g. many arthropods, such as fleas, bedbugs); contact reactions (e.g. locusts); to an invading parasite (e.g. scabies); to retained mouth-parts, etc. (e.g. tick bite granulomas) • Secondary infection (e.g. many insect bites, scabies) • Transmission of disease (e.g. leishmaniasis, erythema chronicum migrans - caused by a spirochaete). Insect bite reactions should be suspected when there are grouped or linear arrays of itchy weals or papules. The lesions last longer than urticaria. Fleas and mites, often from domestic pets, mosquitoes, gnats, midges and bedbugs are frequent causes of this common problem. A careful history and examination of brushings from pets, their bedding and house dust can help identify the cause.

Scabies Scabies (Fig. 10.17) is an itchy dermatosis caused by the human mite Sarcoptes scabiei. Clinical features The female mite, just visible to the naked eye, makes the necessary journey from one human to another during close bodily contact and burrows into the horny layer, producing a serpiginous greyish elevation 1-10 mm long. Burrows are most commonly seen on the finger webs, wrists, elbows, around the nipples in women and on the male genitalia. In children the palms, soles and even the face may be colonized, and vesicles and pustules may be evident. In addition there is a widespread papular rash and extensive excoriation 0, probably the result of an allergic reaction to mite products. Generally there are only a few burrows at a time, but in so-called crusted or Norwegian scabies there are vast numbers, particularly on the hands. This mainly occurs in patients with learning difficulties, the immunosuppressed, the elderly and those who are unable to scratch.

With an initial attack there is a latent period of about 2 months before itching begins, but this period is reduced in subsequent episodes. The itching is often severe and tends to be worse at night. The distinctive burrows are usually overshadowed by widespread excoriations, eczematization and secondary infection.

The lesions following their bites often do not appear until well into the next day. They occur mainly on the face and upper limbs. The environment should be treated with an insecticide.

Diagnosis Mites can be extracted from burrows with a needle; somewhat easier is scraping the area after applying a drop of mineral oil or 30% potassium hydroxide and viewing the material microscopically for mites, eggs, faeces etc.

Lice are obligate human parasitic wingless insects. The itchy reactions are a result of sensitization. There are two species: Pediculus humanus, which occurs in two varieties - the head and the body louse - and Phthirus pubis, the pubic or crab louse.

Treatment The patient and all close contacts should be treated with a scabicide such as permethrin, lindane (gamma benzene hexachloride), malathion, or benzyl benzoate from the neck downwards. Properly applied, one dose is usually curative, but sometimes two applications 3 days apart are needed. Pruritus may continue for up to 2 weeks and can be helped by crotamiton (a less effective scabicide which also has antipruritic properties) or a topical steroid. 3

Other mites Dogs, cats, birds, grain and stored foodstuffs are sometimes sources of mites that can attack humans, producing itchy papules, often topped by tiny vesicles, and weals. Harvest mites, minute red creatures found in grass and other low vegetation, can produce similar and often very florid lesions. Diagnosis involves finding the offending mite, e.g. from brushings of pets.

Fleas Fleas include varieties that can burrow into skin (e.g. Tunga penetrans, found only in the tropics) and those that feed on the skin. In general the human flea is unusual, but occurs in overcrowded communities with poor hygiene. Bites from fleas living on pets, birds and occasionally wild animals, such as hedgehogs, are very common. The reaction to a flea bite is a hypersensitivity response. Individual lesions are like urticaria, but often have a haemorrhagic central punctum and last a week or two. They characteristically occur in groups, and often leave hyperpigmentation as they resolve. Troublesome itching can be treated with an oral antihistamine and secondary infection with an antibiotic. When the domestic pet is the source, the diagnosis can often be confirmed by examination of brushings from the animal. Flea faeces in this material stain damp white blotting paper red. Both the pet and the environment need appropriate treatment.

Bedbugs Bedbugs (Cimex lectularius) are nocturnal, wingless, bloodsucking insects that live in crevices in the walls, floor and furniture, and can best be seen in the middle of the night.

10

Lice

Head lice Head louse infection is very common in children, presenting as intense irritation, particularly at the back of the scalp. Secondary infection is frequent and may be mistakenly diagnosed as impetigo of the scalp. The insect is 3-4 mm long, but usually the eggs and egg cases (nits) are much more easily seen, firmly cemented to the hairs. Treatment is with malathion, a pyrethroid or carbaryl. Contacts should also be treated. Body lice The body louse is rare in developed countries, except in cases of self-neglect. There is widespread irritation, with papules and excoriation. The lice and their eggs are found on the clothing, particularly the seams, and are best treated by hot tumble drying. Pubic lice Pubic lice have legs somewhat like crabs' pincers, with which they are often found clinging to hairs. Secondary infection is common, and as these creatures are often acquired during sexual intercourse other venereal disease should be excluded. Treatment is with lindane (gamma benzene hexachloride), malathion or a pyrethroid.

PSORIASIS Psoriasis is a non-infective, usually chronic inflammatory skin disease with a number of clinical manifestations, the most common of which is red plaques covered by silvery scales (Fig. 10.18). 4 A distinctive arthritis may occur, and nails are often involved (Fig. 10.19). Genetic and environmental factors interact to account for the capricious natural history.

Aetiology Psoriasis occurs in about 2% of the population in Britain, but is less common in other countries, notably West Africa and Japan. The condition may present at any age, but onset is most common at puberty and the menopause. Additional triggering factors, of both onset and recurrence, are infection, especially streptococcal pharyngitis, which character-

395

Pathology and pathogenesis In the well-established psoriatic lesion there is epidermal thickening with increased mitoses, retention of nuclei in the horny layer (parakeratosis) accounting for the silvery scales, small collections of neutrophil polymorphs in the epidermis, tortuous dilated dermal capillaries and a dermal inflammatory infiltrate. Cell kinetic studies show that epidermal turnover occurs in 5-7 days, compared with the normal 30-45 days. Epidermal maturation is also abnormal, not just as a consequence of the rapid throughput. There is no comprehensive theory to account for the many abnormalities demonstrated in the epidermis, dermis, circulating granulocytes and lymphocytes, and the humoral immune system. CD4 helper lymphocytes in the dermis produce cytokines which initiate many of the inflammatory and proliferative events. From work in which psoriatic skin is explanted on to athymic nude mice it seems that the increased proliferation is intrinsic to the skin. It is likely that circulating factors determine the overall activity of psoriasis and maintain the inflammatory process in the plaques.

FIG. 10.18 Psoriasis Well-defined red plaque with silvery scales.

Clinical features The different patterns of psoriasis result from variations in expression of the disease (guttate, erythrodermic and pustular types) and modification of the features at certain sites, notably the scalp, flexures, palms and soles. The extent and duration are usually unpredictable, although some patterns are more likely to improve spontaneously, e.g. guttate psoriasis. Psoriasis means a condition (-iasis) of itching (psor-), but pruritus is not inevitable. 1 FIG. 10.19 Psoriatic arthritis Red swollen tender terminal interphalangeal joints with associated dystrophic nails.

istically provokes guttate psoriasis; trauma to the skin, generating new psoriatic lesions (Koebner's phenomenon, see below); and some drugs, notably lithium salts, systemic corticosteroids followed by their rapid withdrawal, and antimalarials. In some subjects UV radiation can initiate psoriasis on exposed skin. Rarely, hypocalcaemia provokes psoriasis. The importance of stress is disputed, but sudden psychological trauma can play a part in initiating the disease and causing relapse. The tendency to develop psoriasis is genetically determined but the mode of inheritance remains unclear. There is a close association with HLA-CW6 and HLA-DR7, the former conferring a 5-10-fold increased risk. HLA-B27 is seen in 70% of those patients with psoriasis and an ankylosing spondylitis pattern of spinal arthritis.

1 396

MCQ 10.5

2

Fig. 10.13

3

Fig. 10.14

Discoid or plaque psoriasis The well-defined, silvery-scaled red plaques may be found at any site, but the knees, elbows, extensor surfaces of limbs and the lower back are commonly affected. Scratching the scales reveals many tiny bleeding points. Flexural psoriasis Psoriasis of the axillae, groins and beneath the breasts is less readily recognized because the plaques are smooth, but usually remain more sharply defined than eczema. Lesions are typically symmetrical. Scalp psoriasis Scaling often becomes very thick, beneath which are telltale areas of well-defined erythema. 2 Psoriasis of palms and soles This may be more hyperkeratotic, without the silvery scale, but usually remains well defined. Guttate psoriasis There are numerous scaly papules and small plaques. 3 This type of psoriasis mainly occurs in children and adolescents. Erythrodermic psoriasis Total involvement of the skin can cause hypothermia and hypoproteinaemia (see also p. 410).

Pustular psoriasis Pustules occur when the neutrophil collections become clinically visible. The main varieties are a form localized to the palms and soles, in which yellow, white or greenish pustules (sterile) arise on erythematous scaly skin (Fig. 10.20), and a generalized type. In the latter the patient is ill, with fever and leukocytosis, and large areas of erythema develop myriad small sterile pustules which soon dry up and desquamate, to be followed by further waves of pustules. As with erythroderma, management entails rest, the prevention of hypothermia, and maintenance of fluid, electrolyte and protein balance, as well as control of the disease process. Localized pustular psoriasis of the palms and soles is probably genetically distinct, there being no association with HLA-CW6 or HLA-DR7, and often occurs without psoriasis elsewhere. Tar and dithranol are contraindicated in generalized pustular psoriasis, and generally unhelpful in palmoplantar pustular psoriasis. Psoriatic nail changes Common patterns are: • Multiple small pits, resembling a thimble (Fig. 10.21) • Onycholysis (separation of the nail), often with an adjacent zone of orange discoloration • Subungual (nailbed) keratosis • Thickening and distortion of nails. Nail changes are very common with psoriatic arthropathy involving the fingers.

FIG. 10.20 Pustular psoriasis of the soles Circumscribed areas of erythema, some scaling and pustules.

Psoriatic arthropathy Probably about 5% of psoriatics have a seronegative arthritis, although a much higher proportion have abnormalities detected by isotopic bone scans. Different patterns include:

10

• Distal arthritis - terminal interphalangeal joints of fingers and interphalangeal joints of toes • Simulation of rheumatoid arthritis, but more asymmetrical • Sacroiliitis and ankylosing spondylitis • Arthritis mutilans, involving multiple joints with marked bone resorption. Diagnosis and management are discussed on page 1151.

Management The choice of treatment will depend on many factors. These include the capabilities and lifestyle of the patient, as well as the type, location, extent and severity of the psoriasis. The options shown in Table 10.13 will not be appropriate for every patient, and are often used in combination. Some patients will prefer not to have any treatment. Nails do not respond well to topical therapy, but may improve with the rest of the psoriasis, especially when systemic treatments are used. Topical therapy Rest and bland applications are important in acutely inflamed psoriasis, erythrodermic and generalized pustular psoriasis. Many cases with discoid lesions respond well to tar and dithranol preparations, although, particularly with the latter, it is wise to begin with a weak preparation and gradually increase the potency to minimize irritation. Topical steroids are less likely to clear psoriasis completely, but with due attention to their potential hazards they are useful for areas that are readily irritated, such as the flexures, genitalia, ears and face; they are often more effective on the palms and soles. The most potent preparations are

FIG. 10.21 Psoriatic nail showing thimble-like pitting

397

CASE STUDY 10.1 A WIDESPREAD RASH IN A 35-YEAR-OLD WOMAN FOLLOWING AN UPPER RESPIRATORY TRACT INFECTION A 35-year-old married woman presented complaining of the sudden onset of a widespread rash, beginning 2 weeks after a severe sore throat for which she had received a course of penicillin. There was no previous history of skin disease but her mother and older sister both had psoriasis. On examination there were numerous well circumscribed, discrete red scaly plaques on the trunk, elbows and knees. There were also lesions in the scalp (Case Fig. 10.1.1).

Questions 1. What is the most likely diagnosis? 2. What other condition should be considered? 3. What treatment should be instituted?

Discussion The most likely diagnosis is guttate psoriasis. The sudden onset following a sore throat (probably streptococcal) is typical and the patient has a positive family history. Pityriasis rosea can also present in this way and can sometimes be difficult to distinguish, especially in the first week or two. However, lesions on the limbs and the scalp are rare in pityriasis rosea. The most effective treatment in this case would probably be the combined use of topical tar preparations and ultraviolet B (UVB) phototherapy, although some dermatologists would also use moderately potent topical steroids and/or vitamin D analogues (calcipotriol or tacalcitol), either as well as or instead of the tar. The phototherapy would normally be administered in a specialist facility

1

398

MCQ 10.6

twice weekly, initially for 6-8 weeks, before an assessment of the response was made. If the condition failed to settle satisfactorily, further measures might need to be considered. Questions 4. What therapies might be offered to this woman if firstline therapy fails? 5. What additional factors need to be considered in reaching the appropriate decision? Discussion It may be necessary to commence topical therapy with dithranol, an anthracene derivative which can eradicate psoriatic plaques completely. However, the treatment is messy, time-consuming and can, if mishandled, result in burns and staining of both skin and clothes. This woman was a busy executive with an international marketing company and was unable to accept the inconvenience dithranol would have caused. There is a case to be made for at least a short course of

CASE FIG. 10.1.1 Plaques of psoriasis

systemic therapy with acitretin (a retinoid), methotrexate or ciclosporin. It is important to discuss the implications of each of these drugs with the patient and her husband. They need to understand that acitretin is teratogenic. It may also cause hyperlipidaemia and a (usually mild) hepatitis. It is important to avoid pregnancy while on the drug and for 2 years after stopping. Methotrexate can, in exceptional circumstances, cause severe bone marrow suppression, but its main problem is hepatotoxicity if used over long periods. Ciclosporin is expensive and may cause renal and cardiovascular damage. In the circumstances this woman and her husband agreed that it was sensible, after all, to try to use dithranol at home under the supervision of a specialist nursing team from the hospital. Although progress was slow, the psoriasis reduced considerably in extent and the woman was able to manage her disease effectively thereafter with a combination of topical steroids and vitamin D analogues.

10

TABLE 10.13 Treatment choices in psoriasis Form

Site

First choice

Second choice

Third choice

Plaque

Face, flexures genitalia Scalp Trunk, Limbs: few plaques Trunk, limbs: many plaques

Topical steroid Tar gel or pomade Dithranol or calcipotriol Calcipotriol Dithranol ± UVB UVB Topical steroid Tar Dithranol

Mild tar or weak dithranol Dithranol or calcipotriol Tar PUVA

Topical steroid Topical steroid Systemic therapy, e.g. methotrexate cyclosporin

PUVA

Oral retinoid

Guttate

Calcipotriol ± UVB

UVB ± tar

Topical steroid

Erythrodermic with systemic complications

Oral retinoid

Methotrexate

Systemic steroid

Topical steroid Oral retinoid

PUVA Methotrexate

Oral retinoid Other systemic therapy

Palms and soles

Pustular psoriasis

Palms and soles Generalized

best avoided, except on the palms and soles, and even there they should be restricted to short courses. The vitamin D analogue calcipotriol is effective and provides an effective alternative to tar and dithranol. Ultraviolet (UV) radiation and photochemotherapy Ultraviolet B (p. 426) is used both alone, e.g. for guttate psoriasis, and together with other modalities. Photochemotherapy (also known as PUVA) is the combination of long-wave UVA and the photosensitizing drug psoralen, usually taken orally 2 hours previously. It is a powerful treatment for psoriasis, but prolonged use is likely to age the skin prematurely and to induce non-melanoma skin cancer. Photochemotherapy is valuable for extensive plaque psoriasis when topical therapy is poorly tolerated or has failed, and can be successful for erythrodermic and pustular forms. Systemic therapy For the most severe forms of psoriasis, antimitotic and immunosuppressant drugs are used. Methotrexate is effective and reasonably safe. In long-term use hepatic fibrosis can occur and liver biopsies are recommended every 1-2 years to identify this before it is irreversible. Alcohol should be avoided. Ciclosporin has recently been introduced for the treatment of cases where methotrexate is unsuitable. This drug is not antimitotic, and probably works by modulating the immune system. The main hazard is nephrotoxicity. Vitamin A derivatives (retinoids) with less toxicity than vitamin A itself have been found to be of value in pustular and erythrodermic psoriasis, and useful in combination with other treatments, e.g. photochemotherapy. Retinoids have a number of side-effects and, like methotrexate, are teratogenic, but they are not cytotoxic or immunosuppressant. They probably influence the disturbed maturation of the epidermis in psoriasis. 1

FURTHER READING ON PSORIASIS Greaves N W, Weinstein G D 1995 Treatment of psoriasis. N Engl J Med 332:581-588. Stern RS 1997 Psoriasis. Lancet 350:349-353. Workshop of the Research Unit of the Royal College of Physicians of London: Department of Dermatology, University of Glasgow, British Association of Dermatologists 1991 Guidelines for management of patients with psoriasis. Br Med J 303: 829-835.

ECZEMA AND DERMATITIS Eczema is a pattern of inflammation with many possible causes, rather than a disease. The different types are unified histopathologically by the presence of spongiosis in the epidermis. The term 'dermatitis' is often used to imply an eczema of external origin, but it should always be qualified, as in 'contact allergic dermatitis', as 'dermatitis' is used in many other contexts, e.g. dermatitis artefacta, dermatitis herpetiformis etc. Eczema is usually itchy and red; when the inflammation is intense, vesicles are seen and there is oozing from the surface (Fig. 10.22). If the process continues, scaling and thickening occur, and with rubbing and scratching lichenification (the development of leather-like plaques) is common. The earliest and most distinctive pathological change is spongiosis, the accumulation of oedema fluid within and between keratinocytes. Viewed in microscopic section, this produces an appearance somewhat like a sponge. In acute eczema the process extends to produce intraepidermal vesicles. In the dermis, and to a lesser extent the epidermis, there is an inflammatory infiltrate in which lymphocytes

399

contact allergen. The face is often affected, even when not directly exposed to a causative agent.

EXOGENOUS ECZEMA Irritant contact dermatitis When the skin changes are a direct result of exposure and allergic mechanisms are not involved, the rash is termed irritant contact dermatitis. Not all reactions to irritants are eczematous; for example, the agent may produce necrosis or urticaria. Because obvious irritants tend to be avoided, the main problems occur with those that require prolonged and repeated contact, e.g. detergents, alkalis, mineral oils and organic solvents. Atopic individuals are particularly susceptible, as are housewives, hairdressers and those in a number of industrial occupations. Even when a period of avoidance produces healing, chronic irritant dermatitis tends to recur readily with re-exposure.

Allergic contact dermatitis

FIG. 10.22 Eczema With intense inflammation there are vesicles and oozing. If the process continues, scaling and thickening occur.

are prominent. With time, both the epidermis and dermis become thickened. Aetiological factors in eczema include: • • • • • • •

400

Skin irritants Contact allergens Friction, low humidity, UV light Infections (bacterial, dermatophytes, yeasts) Atopic constitution Drugs and foods Local factors: venous stasis, ichthyosis.

Several factors may act in concert, particularly with eczema of the hand and lower leg, and not infrequently treatment is an inadvertent contributory factor. A careful history is essential in the evaluation of eczema. Several different clinical patterns of eczema and dermatitis (with eczematous features) are recognized. The basis for these patterns is well understood in some instances (e.g. contact allergic dermatitis) but obscure in others (e.g. discoid eczema). Because successful management depends on recognizing external factors that are contributing to the skin disease, it is helpful to classify according to whether the disease process is primarily exogenous or endogenous, although these distinctions are somewhat artificial. Whatever- the cause, while active, eczema tends to spread, often beyond any recognizable stimulus such as a

Allergic contact dermatitis is the response to a substance to which the individual has become sensitized. Most such substances are of low molecular weight and penetrate the skin to behave as haptens. After conjugation with a skin protein the complete antigen is taken up by Langerhans' cells, which migrate to regional lymph nodes and act as antigen-presenting cells in a T lymphocyte-mediated (type IV) immune response. All subsequent contact with the sensitizer results in an acute eczematous reaction. This type of response is uncommon in children, although contact urticaria is frequent in atopies (see below). Contact allergic dermatitis usually begins at the site of contact, but the clinical picture can at times be deceptive, e.g. nail varnish resin, which causes no trouble locally but gives rise to eczematous patches when it is transferred to the face and neck by the fingers. Many factors determine the acquisition of contact allergy. Some substances, e.g. epoxy resin and poison ivy, are inherently more likely to sensitize. Occlusion (e.g. a hand in a glove) and skin damage (e.g. from irritants) enhance penetration. It is possible that some individuals are more likely than others to be sensitized. The thick horny layer of the palms renders them less likely than the dorsa of the hands to be affected. Once initiated, repeated exposure tends to produce increasingly severe and widespread dermatitis, and sites far removed, such as the eyelids, can be involved. The allergic state usually persists for life. Some common sensitizers are listed in Table 10.14. Investigation: patch testing The allergic state is reproduced in miniature by the application of a non-irritating dose of a suspected chemical beneath an occluding disc or patch placed on the back for 48 hours. Because some allergens are ubiquitous, a

CASE STUDY 10.2 A 25-YEAR-OLD NURSE WITH AN ITCHY RASH ON HER HANDS A 25-year-old nurse developed an itchy rash that particularly involved her hands. She had been working on the labour ward for the previous 3 months. She reported to the hospital occupational health department one morning because of pain and swelling in her right hand. She had been feeling unwell, had been shivery, and had had aching muscles for the previous few hours. She had a past history of eczema in childhood and urticaria when given penicillin. She had bouts of sneezing when near furry animals. In her family history there was a sister with asthma. On examination she appeared unwell, had a temperature of 38.5°C, an erythematous, swollen, tender right hand, and both hands showed areas of erythematous thickening, fissuring and yellow crusts (Case Fig. 10.2.1). There were erythematous and excoriated patches and plaques on the chest and, back, and behind the knees. Questions 1. What is the most likely reason for the febrile illness and swollen right hand? 2. What investigations would be appropriate in this case? 3. What would be your immediate management? She has cellulitis secondary to a breach in the skin owing to hand eczema. The most likely cause is Bhaemolytic streptococcus group A, but Staphylococcus aureus, both organisms, or less probably some other bacteria, e.g. from her workplace, are also possible causes. Bacteriological confirmation is valuable, especially since she is systemically unwell; works in a hospital environment; and has a history very suggestive of allergy to penicillins. Useful investigations at the outset are swabs from crusted lesions, blood cultures, a full blood

count and serum antistreptolysin O titre (ASOT) and antideoxyribonuclease B (ADB). It was decided to admit her to hospital for parenteral therapy with i.v. cefuroxime. She tolerated this well and made a full recovery over the next 5 days. Investigations gave the following results: Hb 13.5 g/dL, WBC 23 x 10"/L, neutrophils 95%, ASOT 300, ADB >1200. Cultures from the skin showed Staph. aureus, sensitive to flucloxacillin. Blood cultures were negative. Note that she had a neutrophil leukocytosis, and of the two serological tests for evidence of acute streptococcal infection only the ADB was unequivocally positive. In general, ASOT is more commonly positive with pharyngeal infections. She was discharged and returned to work 1 week later. Shortly thereafter, she found that her hands became red and itchy again, but this time the eruption was over both hands and extended on to the wrists. She saw her GP, who prescribed betamethasone valerate cream (a potent topical corticosteroid) to use twice daily and an emollient for the dryness, but her hands did not improve. A referral

10

letter was sent to the local dermatologist but before she could be seen she had an episode of acute shortness of breath and collapsed while eating a salad in a restaurant. She was taken to the nearest A & E department and was found to be hypotensive, dyspnoeic, with grossly swollen lips and tongue, and with an extensive urticarial eruption. Anaphylaxis was diagnosed; she was treated with parenteral adrenaline (epinephrine), chlorphenamine (chlorpheniramine) and hydrocortisone, and made a rapid recovery.

Questions 4. What is the likely mechanism for the acute anaphylactic reaction? 5. How can it be investigated? Release of inflammatory mediators, especially histamine, from mast cells and basophils is the most common cause of acute anaphylaxis. In this case the release was probably IgE mediated. Other mechanisms include the direct action of some pharmaceutical agents (e.g. opiates) and circulating immune complexes.

CASE FIG. 10.2.1 Lymphoedema of the hands following recurrent eczema and infection

401

CASE STUDY 10.2 CONTINUED Investigations for IgE-mediated reactions include prick tests and in vitro methods such as the RAST (radioallergosorbent test). When the patient was evaluated by the dermatologist, the history revealed that she had had an acute swelling of the lips when blowing up balloons a few weeks earlier, and had also noted that her hands were becoming red and itchy immediately after she put rubber gloves on. Serum was sent for a RAST to latex proteins, and was strongly positive. As well as being relevant to her immediate reactions to rubber gloves and balloons, her hypersensitivity to latex proteins probably explains her episode of anaphylaxis in the restaurant: she had eaten avocado in the salad, and this is one of a number of plants which have proteins that cross-react with latex protein. Others include banana, chestnut and kiwi. She was also patch tested and had a positive reaction to sorbic acid (the preservative in her emollient cream). Discussion This case illustrates a number of problems that can occur in the atopic individual. Having had self-limiting eczema as an infant, the patient developed it again as an adult when her hands were subjected to repeated hand-washing with detergents and soaps in her work as a theatre nurse.

The skin in someone with atopic eczema commonly becomes colonized with Staph. aureus and sometimes Strep. pyogenes group A, and this led on to acute soft tissue infection (cellulitis). It is common for streptococcus not to be cultured from either the surface or the blood, and diagnosis only to be confirmed serologically. The interpretation here is that the cellulitis was due to B haemolytic streptococcus; the Staph. aureus was a superficial infection only. In a patient known to be allergic to penicillins there is some risk of cross-hypersensitivity with other antibiotics containing the (3lactam ring in their structure, but the risk was thought to be low because she had had urticaria without systemic symptoms many years ago. There is also a possibility that the causative organism would not be sensitive to a macrolide (e.g. erythromycin), which would otherwise be a reasonable choice. Hence the decision to admit her and give cefuroxime, a cephalosporin, albeit with a small risk of an allergic reaction. If she had been working on a medical ward there could well have been a significant risk of her being infected with a multiresistant Staph. aureus (MRSA), in which case i.v. vancomycin would have been the drug of choice. In a patient with continuing hand eczema, patch testing was the

selection of common sensitizers is tested as well as any chosen on the basis of the patient's history. A positive reaction does not prove the cause of a patient's rash, but does help confirm suspicions and suggest unsuspected materials. Positive reactions may be of past relevance only. Light-induced eczema Eczema on light-exposed skin may be due to:

1

402

Fig. 10.15

appropriate step to take to evaluate whether contact allergy (type IV hypersensitivity) was playing a part in addition to atopy and irritant exposure. She was tested for a large number of potential allergens, including those in her treatments. When it was found that she was allergic to the preservative in her emollient, this was changed to a soft white paraffin product with no preservatives, and all the areas where she had been applying the bland cream to dry skin improved. Type I allergy to the low molecular weight proteins in latex has become increasingly common over the past decade, especially in atopies, and is a significant hazard for nurses and doctors because of their frequent use of rubber gloves. Direct exposure of the allergenic material to internal surfaces (e.g. introduction of a rubber catheter) can be fatal. In this patient, whereas contact with rubber produced urticarial reactions, limited to the skin (her hands and lips), the ingestion of avocado led to acute anaphylaxis. With avoidance of allergens, use of a soap substitute, soft white paraffin as emollient and a mild corticosteroid ointment, the patient's hands recovered and she was able to return to work. She now uses vinyl gloves, rather than rubber ones, for hand protection.

• Photocontact allergy, e.g. sunscreen medicaments. The essential role of UV radiation is shown by patch testing with and without UVR • Photosensitive drug eruption, e.g. chlorpromazine, sulphonamide, thiazide • Exacerbation of existing eczema, e.g. atopic dermatitis. Phototoxic reactions, which resemble sunburn rather than eczema, can occur following contact with tar and its derivatives, some dyes and plants, and with drugs (p. 418). Sensitivity to airborne allergens can closely mimic light reactions, but areas normally spared from sun exposure, such as the skin beneath the ears, under the chin and

10

TABLE 10.14 Some common sensitizers Chemical

Source

Nickel Chromate Epoxy resin Paraphenylenediamine Rubber antioxidants Preservatives Lanolin Perfumes Topical drugs

Jewellery, watch bands, silver coins Wet cement, many engineering processes, leather Two-part glues Hair dyes Gloves, shoes Cosmetics and medicaments Cosmetics and medicaments Cosmetics and medicaments Neomycin, antihistamines, sulphonamides, some local anaesthetics Sticking plasters, newsprint

Colophony

around the eyes, are involved. Dermatomyositis may occasionally mimic a light-induced eruption.

ENDOGENOUS ECZEMA Atopic eczema/dermatitis Atopic eczema/dermatitis (Fig. 10.23) is closely associated with asthma, hay fever, urticaria and dry skin. There is usually a family history of one or more of these. The separate condition ichthyosis vulgaris (p. 405) is probably more common in atopic individuals than in the rest of the population. The skin readily becomes itchy in response to a number of allergic and non-allergic stimuli, and the pruritus may be disproportionate to the physical signs. High levels of circulating IgE antibodies are very common, as are multiple positive prick tests. The disorder is very common, most cases beginning in infancy, and at least 10% of the population are affected to some degree. There is often a fluctuating course, but most cases clear up by adulthood, the majority within the first 5 years of life. Aetiology There is much evidence for a genetic basis for the atopic state, although how this is translated into the various clinical manifestations is unknown. A number of immunological abnormalities have been described, including the overproduction of IgE, reduced numbers of suppressor/cytotoxic T lymphocytes, and cutaneous basophil hypersensitivity, e.g. to house dust mite antigen. Environmental factors are important determinants of fluctuations. These include ingested proteins (e.g. dairy products and eggs), and exposure to house dust mites, pollens and pet hair. Wool next to the skin frequently causes irritation. Extremes of climate often worsen the disease. Beyond a certain level of colonization, Staph. aureus may produce flares of erythema and pruritus.

FIG. 10.23 Atopic dermatitis in a young child: erythema, weeping and crusting

Clinical features Onset is often in the first 3 months of life. Itching is the major symptom and at times there may be little rash. In infancy the skin is often patchily dry and red, with episodes of vesiculation and crusting. By 1-2 years the elbow and knee flexures, wrists, hands and tops of the feet are typically involved. With prolonged rubbing and scratching, papules and then leathery plaques (lichenification) are seen. 1 At any stage secondary infection may occur, with superimposition of impetigo and follicular pustules. Urticaria from contact, e.g. with a pet or foods, is also common. Complications Eczema may be complicated by erythroderma (p. 409). Severely affected children are often growth retarded. Herpes simplex infection can be widespread (eczema herpeticum) and, particularly in the very young, may be disseminated to internal organs, with potentially fatal outcome. If eczema herpeticum is suspected, treatment with aciclovir (i.v. if the patient is unwell) should be commenced immediately. When eczema begins in infancy or early childhood the prognosis for recovery is generally good. Adults can be affected, especially when the skin is exposed to irritants, e.g. in nursing and hairdressing. Management The mainstays of management are the avoidance of irri-

403

SUMMARY 1 Classification of eczema Exogenous Irritant contact dermatitis Allegic contact dermatitis Light-induced dermatitis

Endogenous Atopic eczema/dermatitis Discoid (nummular) eczema Seborrhoeic eczema/dermatitis Pompholyx Asteatotic eczema Gravitational eczema

tants such as bubble bath, mild topical steroids, plentiful use of emollients when the skin is dry, antibiotics when necessary, tar preparations for the more chronic and less inflamed lesions, and a sedative antihistamine at night. Cotton clothes and the use of a soap substitute, e.g. emulsifying ointment or aqueous cream BP, are helpful. Current evidence suggests that children of atopic families should be breastfed, at least up to 3 months if possible. Subsequent supervised dietary manipulation may be helpful.

Discoid eczema Discoid (nummular) eczema consists of discrete rounded patches of eczema, usually symmetrically distributed, for which there is no apparent cause. The aetiology is unknown and treatment is symptomatic. A group 1 or 2 potency topical steroid is often needed, and infection, if present, should be treated with an antistaphylococcal antibiotic.

Seborrhoeic eczema/dermatitis Seborrhoeic eczema/dermatitis (Fig. 10.24) is distinctive in both appearance and location. The cause is unknown, but Pityrosporum yeasts play a part. There is usually no correlation with sebaceous activity, although it is a disorder of areas where sebaceous glands are plentiful. Severe seborrhoeic dermatitis is an early feature in about one-third of AIDS patients. There is less itching than with other eczemas. The lesions are yellowish-red, scaly away from flexures, and involve some or all of the following sites: scalp, eyelids, eyebrows, sides of nose, ears, axillae and groins, beneath the breasts, front of the chest, and upper back. The differential diagnosis will depend on the sites affected. In the scalp, psoriasis can be very similar. In the flexures, psoriasis, contact allergic dermatitis, dermatophyte infections and candidiasis may need to be excluded; the truncal form can resemble pityriasis rosea and pityriasis versicolor (the latter diagnosed from a scraping). Methyldopa can produce a seborrhoeic dermatitis-like drug eruption.

1 Fig.10.16 404

2

MCQ 10.7

FIG. 10.24 Nasolabial erythema and scaling of seborrhoeic dermatitis

The addition of sulphur, a mild tar, salicylic acid or an imidazole to a suitable topical steroid is often helpful. Shampoos containing ketoconazole, selenium sulphide or zinc pyrithione are useful for scalp involvement.

Pompholyx Pompholyx is an eruption of vesicles on the sides of fingers, palms and soles. Itching may be severe. The condition is usually episodic. Causes of pompholyx include the onset of warm weather; dermatitis or fungal infection of the feet (producing pompholyx of hands); ingestion of a contact allergen, e.g. nickel; and stress. Often no cause can be found. Treatment is symptomatic. If weeping is present, potassium permanganate soaks are helpful, followed by a group 1 or 2 potency topical steroid. Secondary infection is common and should be treated with a systemic antibiotic.

Asteatotic eczema Asteatotic eczema is a mildly inflammatory condition that occurs mainly in the elderly on the shins, as erythema with a crazy paving-like pattern of superficial fissuring and scaling. It may be a result of reduced epidermal lipids. Hypothyroidism, uraemia, dehydration, low humidity and overzealous cleansing may contribute. It often responds to less bathing and to emollients. A mild topical steroid ointment can be added if necessary.

Gravitational eczema 1

10

Eczema often occurs on the lower legs as a response to chronic venous hypertension, in association with haemosiderin staining, patchy atrophy, sclerosis, oedema, and perhaps ulceration. It is particularly common near the medial malleolus in patients with previous deep vein thrombosis, but may also be seen in relation to varicose veins and incompetent perforating veins (p. 429). Patches indistinguishable from those of discoid eczema may spread elsewhere. Important in management are improvement of the venous hypertension, e.g. with support hosiery, and the avoidance of potent topical steroids and potential sensitizers, such as neomycin, lanolin and preservatives. Symptomatic treatment of eczema Any causative and aggravating factors should be removed if at all possible. Acute weeping eczema can be soothed with wet compresses, e.g. aluminium acetate lotion BNF or 1:10000 potassium permanganate solution, followed by a topical steroid cream. In some cases, particularly contact allergy, when a cause has been recognized and dealt with, a short course of systemic steroids is justified. For less acute eczema, a topical steroid of the lowest effective potency is needed, often in a greasy base. If the surface is dry, additional applications of an emollient are helpful and reduce the likelihood of steroid overuse. For chronic lichenification, tar preparations (e.g. as impregnated bandages) are valuable. Secondary infection is usually best treated with a systemic antibiotic, e.g. flucloxacillin or erythromycin. 2

FURTHER READING ON ECZEMA AND DERMATITIS Brehler R et al 1997 Recent trends in the treatment of eczema. J Am Acad Dermatol 36:983-994 Charman C 1999 Clinical evidence: atopic eczema. Br Med J 318:1600-1604 Friedmann P S 1998 Allergy and the skin II - contact and atopic eczema. Br Med J 316:1226-1229 McHenry P M et al 1995 Management of atopic eczema. Joint Workshop of the British Association of Dermatologists and the Research Unit of the Royal College of Physicians. Br Med J 310:843-847

ICHTHYOSIS The term ichthyosis is used for a number of conditions in which there is either a disorder of keratinization or shedding of the horny layer, producing persistent, generalized scaly skin (Fig. 10.25). Most types of ichthyosis are inherited and present in infancy or childhood. Onset in adulthood can be associated with an underlying neoplasm (p. 423). Ichthyosis vulgaris is the only common type, occurring in about 1 per 300 of the population. It is inherited as an auto-

FIG. 10.25 Rough skin and scaling of ichthyosis

somal dominant trait and has distinctive histology: absence of the granular cell layer of the epidermis. Individuals have rough skin with white scales, maximal on the extensor surfaces and sparing the antecubital and popliteal fossae. There are often rough follicular papules (keratosis pilaris) on the upper arms and thighs. The physical signs are worse in winter. Avoidance of soap and regular use of emollients are helpful measures. Sex-linked recessive ichthyosis occurs in about 1 per 6000. Female carriers may show mild changes, but the fully developed condition, as seen in males, is more severe than ichthyosis vulgaris, with large brownish scales; the flexures are involved. The condition is characterized by steroid sulphatase deficiency. Acquired ichthyosis has a similar clinical pattern to ichthyosis vulgaris. As well as being associated with neoplasia, particularly Hodgkin's disease, it can occur in gross nutritional deficiency.

REACTION PATTERNS Despite the large number of named diseases, as an organ the skin has a limited range of reactions to noxious stimuli such as circulating immune complexes and infections. Some of these reaction patterns have sufficiently few common causes that their recognition can be helpful in arriving at a diagnosis. This may be particularly true when the patient has other symptoms, e.g. cough or malaise, for which there is no obvious cause.

405

SUMMARY 2 Reaction patterns • • • • • •

Urticaria and angioedema Erythema multiforme Erythema nodosum Pyoderma gangrenosum Cutaneous vasculitis Erythroderma

A single aetiological factor may produce one or more of the reaction patterns described below. For example, the hepatitis B virus may lead to urticaria if the effect on dermal blood vessels is fully reversible, but purpuric vasculitis if the damage is more profound. FIG. 10.26 Urticaria: arcuate weals

URTICARIA AND ANGIOEDEMA The characteristic lesion in urticaria is a weal (Fig. 10.26). This is caused by oedema in the dermis, so weals are raised and often pallid but surrounded by a zone of erythema. Urticaria is evanescent, usually lasting from minutes to a few hours (rarely more than 24 hours) and fades without trace. Itching may be severe. Lesions initially like urticaria may occur in vasculitis, arthropod bites and stings, and lupus erythematosus, but in these circumstances they are more long-lasting. If there is doubt, a skin biopsy can be helpful in distinguishing these conditions. Urticaria lasting more than 3 months is regarded as chronic, and in most cases no cause is found. Angioedema occurs in the subcutaneous tissues, particularly where they are rather loosely organized, such as in and around the mouth, around the eyes and on the scrotum. 1 The lesions are larger, less well defined and not necessarily itchy. Common causes of urticaria and angioedema are listed in Table 10.15. Many cases of chronic idiopathic urticaria are now recognized to be due to IgG autoantibodies directed against the high-affinity IgE receptor and, less commonly, IgE itself. The basis for urticaria and angioedema is vasodilatation and extravasation of fluid owing to release of mediators from mast cells and/or basophils. In some circumstances the mediators responsible can cause hypotension and bronchospasm, constituting anaphylactic shock (p. 97). Mechanisms include IgE and immune complex-mediated reactions, and direct effects of chemicals, e.g. toxins and drugs. Symptomatic treatment is with H1 antihistamines. Hereditary angioedema is a rare but potentially lifethreatening condition, in which non-pruritic swellings

TABLE 10.15 Common causes of urticaria and angioedema Drugs Penicillin, aspirin, X-ray contrast medium Foods Azo dyes, benzoate preservatives (chronic urticaria) Arthropod reactions Transfusion of blood products and foreign proteins

Plants Stinging nettles Systemic disease Hepatitis B, other virus infections systemic lupus erythematosus Contact urticaria Some foods and pet saliva on atopies Latex Inhalants Grass pollens, house dust

Physical stimuli Light pressure (dermographism), cold, light, increase in body temperature (cholinergic urticaria)

occur in the gastrointestinal and respiratory tracts as well as the skin and adjacent mucous membranes. A careful family history may suggest the diagnosis. There is either deficiency or an abnormality of the inhibitor of activated complement component Cl. The best laboratory investigation is the measurement of Cl esterase inhibitor by antigenie activity or functional assay. Also useful is the demonstration of low C4 and C2, the substrates of activated and unopposed Cl. Acute episodes may respond to Cl esterase inhibitor concentrate, and danazol can be used to suppress attacks. © FURTHER READING ON URTICARIA

406

1 Fig. 10.17 4 Fig. 10.19

2

MCQ 10.8

5

MCQ 10.9

0 Fig. 10.18

Kaplan A P 2002 Chronic ruticaria and angioedema New Engl J Med 346:175-179. Sabroe R A, Greaves M W 1997 The pathogenesis of chronic idiopathic urticaria. Arch Dermatol 133:1003-1008.

10

FIG. 10.27 Erythema multiforme: target lesions with central blisters

ERYTHEMA MULTIFORME AND STEVENS-JOHNSON SYNDROME The causes of erythema multiforme include: • • • • •

Herpes simplex Other infections, e.g. mycoplasma Drugs, e.g. sulphonamides Pregnancy Collagen vascular disease, e.g. SLE.

The distinctive features of erythema multiforme are the target lesion (Fig. 10.27), the distribution of the rash and the histology. It is an acute eruption that usually lasts about 3 weeks, maximally involving the hands and feet, forearms, 3 elbows and knees. Among a variety of erythematous lesions there are usually some target lesions. These have a purplish or blistering centre inside one or more red rings. The mucosae of the mouth, 4 genitalia and eyes may be involved (Stevens-Johnson syndrome, Fig. 10.28) with blistering and haemorrhage. Sometimes internal organs, e.g. the kidney and lungs, are affected. There is an interval of about a week between the stimulus and the eruption, but this is shorter with recurrences. The pathogenesis is not fully understood. When erythema multiforme is recurrent, the most common cause is herpes simplex. The condition is self-limiting but supportive care is needed, especially if there is eye and oral involvement. This will include oral toilet, analgesia, antibiotics, treatment of ocular complications and, in cases with widespread blistering, management as for burns on an intensive care unit.

FIG. 10.28 Stevens-Johnson syndrome Severe muco-cutaneous erythema multiforme due to mycoplasma infection. Note the conjunctivitis, haemorrhagic crusting on the lips and target lesions.

nodosum is often accompanied by fever, malaise and painful joints. Lesions disappear over 3-6 weeks, with the same colour changes as in a resolving bruise. It is mainly a condition of young adults. Most cases are probably due to circulating immune complexes lodging in or near the venules in the deep dermis and subcutaneous fat. Causes include: • • • • •

Sarcoidosis Streptococcal infection Ulcerative colitis and Crohn's disease Tuberculosis Other infections, e.g. yersinia, deep mycoses (where e.g. blastomycosis occurs commonly), Chlamydiapsittaci and lymphogranuloma venereum • Drugs (uncommon), e.g. oral contraceptives, sulphonamides • Leukaemia and Hodgkin's disease. 0 Similar but smaller nodules are seen in Behcet's disease (p. 1153). Erythema nodosum leprosum is clinically and histologically quite different (p. 327). The lesions of nodular vasculitis are more persistent and occur more on the calves. Bed rest is advisable at first, and a non-steroidal antiinflammatory drug may be helpful. Further management will depend on the cause.

ERYTHEMA NODOSUM The plaques and nodules of erythema nodosum (Fig. 10.29) are red, hot and tender. They typically occur over the shins but may be more widespread on the legs, on the outer aspect of the arms and sides of the neck. Erythema

PYODERMA GANGRENOSUM Pyoderma gangrenosum (Fig. 10.30) is an uncommon, noninfective ulcerating condition frequently associated with an

407

FIG. 10.30 Pyoderma gangrenosum: ulcer with purplish undermined edge

small lesions topical and intralesional corticosteroids can be effective; minocycline can be a valuable agent in milder cases.

CUTANEOUS VASCULITIS FIG. 10.29 Erythema nodosum Tender erythematous nodules over the shins in a young woman due to streptococcal pharyngitis.

underlying systemic disease. 1 The lower limbs and face are the most common sites, but any part of the body may be affected. The necrotic area often begins as a sterile pustule which breaks down and spreads, sometimes to form a huge ulcer, with a characteristic ragged, purplish overhanging edge. Lesions may be multiple. The pathogenesis is unclear, but many patients have evidence of a depressed immune system. Causes include: • • • • •

Ulcerative colitis Crohn's disease Rheumatoid arthritis Other forms of arthritis Leukaemia, lymphomas, monoclonal gammopathies and myeloma.

The differential diagnosis includes infective causes of ulceration (e.g. postoperative gangrene, p. 386) and vasculitic disorders (e.g. Wegener's granulomatosis). Cultures for aerobic and anaerobic bacteria and other organisms, and a biopsy, should be taken before treatment. Treatment is with high-dose corticosteroids, although for

1

408

4

Case 10.1 MCQ 10.10

2 5

Fig. 10.20 Fig. 10.22

0 Fig. 10.21

Palpable purpura is the hallmark of cutaneous vasculitis (Fig. 10.31) 2 but, depending on the type and degree of vessel wall damage, other lesions, not all of which are purpuric, may occur; these include livedo reticularis, weals, papules, pustules, infarcts and ulcers. Aetiological factors of cutaneous vasculitis include: • Bacteria, e.g. streptococci, gonococci, Mycobacterium tuberculosis (erythema induratum) and M. leprae (erythema nodosum leprosum) • Viruses, e.g. hepatitis B • Neoplasia, e.g. lymphoma • Drugs, e.g. sulphonamides,thiazides,captopril (see p. 418) • Food additives, e.g. tartrazine • Autoimmune diseases, e.g. lupus erythematosus, polyarteritis nodosa, scleroderma. The initial event is damage to the endothelium of the vessel. Both circulating and resident extravascular inflammatory cells are then activated to release mediators, some of which have destructive effects whereas others contribute to repair. The net result depends on many factors; these include the nature of the initial insult, the type and location of the vessels affected, and the adequacy of the collateral blood supply. One of the most common histological appearances is the accumulation of neutrophils, their breakdown products, and nuclear debris around damaged venules, an appearance termed leukocytoclastic vasculitis. Sometimes vasculitis occurs because there is inadequate clearance of a potential cause, such as circulating immune complexes, from the circulation, or there is a deficiency in repair mechanisms, e.g. removal of fibrin is

TABLE 10.16 Some multisystem vasculitides that may involve the skin Clinical patterns

Other organs commonly involved

Henoch-Schonlein purpura Polyarteritis nodosa

Joints, gut, kidney Kidneys, gut, cardiovascular system, nervous system, muscles and joints Eyes Kidneys and respiratory tract Gut and brain

Temporal arteritis Wegener's granulomatosis Malignant atrophic papulosis (Degos' disease) Lupus erythematosus

10

Widespread

are characterized by a leukocytoclastic vasculitis, and in many cases there is deposition of IgA around venules. There is no specific treatment, but some authorities use high-dose corticosteroids if there is severe renal disease. O

FIG. 10.31 Vasculitis: purpuric papules and nodules

too slow to maintain patency of affected vessels. A defect in the reticuloendothelial system may account for the occurrence of vasculitis in lymphomas, other malignancies and sarcoidosis. Sometimes the localization of vasculitic lesions can be explained by circumstances that slow blood flow, e.g. in the lower limbs and where skin is cooled by overlying fat. As well as attempting to find the cause, it is clinically important to know whether other organs are being affected, especially the brain, heart, kidneys, lungs and gut. There is no entirely satisfactory classification of multisystem vasculitis (see Ch. 22, p. 1181 for a further account of the vasculitides). Most multisystem vasculitides can involve the skin (Table 10.16).

Clinical features Henoch-Schonlein purpura Henoch-Schonlein purpura comprises arthralgia, abdominal pain and vasculitic rash, often with renal involvement (see p. 1076). Children are predominantly affected. A streptococcal sore throat and upper respiratory virus infections are the most common recognizable causes. The rash is mainly on the buttocks and extensor surfaces of the limbs. ©As well as purpura there are usually erythematous macules and papules, and urticarial weals. The lesions

Polyarteritis nodosa Polyarteritis nodosa, or PAN (see pp. 1076,1181), is a multisystem disorder in which there is necrotizing vasculitis of small and medium-sized arteries. 0 The distinctive feature in the skin is the occurrence of nodules along the course of subcutaneous arteries, best felt on the lower limb. In addition to nodules there are often purpuric papules, weals, or plaques of gangrene. Livedo reticularis, a net-like arrangement of bluish venules, is common. Malignant atrophic papulosis Malignant atrophic papulosis (Degos' disease) is a rare but very distinctive condition in which insignificant-looking red papules become slowly necrotic with a greyish-white central scale, and heal with porcelain-like white scars. Similar lesions occur internally, particularly in the intestine, and much of the high mortality is due to perforation or haemorrhage. The brain and kidney may also be affected. Rheumatoid disease Small purple or black spots around the nailfolds are very characteristic of rheumatoid disease (p. 1138), although they are also seen in other vasculitic diseases.

ERYTHRODERMA Erythroderma, or exfoliative dermatitis, is redness of all or nearly all the skin. There is usually some degree of scaling. Most cases of erythroderma evolve from more limited eczema or psoriasis. The principal causes include eczema; psoriasis; drugs, especially gold, allopurinol and sulphonamides (see p. 417); and lymphoma and leukaemia. The cause can usually be determined from the history and a skin biopsy, but in some cases no aetiology is found.

409

SUMMARY 3 Causes of erythroderma • • • • •

Eczema/dermatitis Psoriasis Drugs, e.g. gold, allopurinol Lymphoreticular neoplasia, e.g. mycosis fungoides Miscellaneous

The skin is red, hot to the touch and often oedematous; itching can be severe and scaling is variable. Patients feel uncomfortably cold and shiver. Erythroderma can have serious metabolic consequences, particularly in the elderly. The greatly increased blood flow through the skin, with loss of normal capacity for vasoconstriction, leads to hypothermia in a cold environment and can cause high-output cardiac failure. Blood flow to other organs can be greatly diminished. There is increased fluid loss from the skin surface, with compensatory thirst and oliguria. Anaemia is common, and in part is due to a state of malabsorption secondary to the skin disease (dermatogenic enteropathy). Serum albumin falls as a result of loss from shed skin scales and reduced synthesis. The skin is more prone to infection, and thrombophlebitis is common. Lymphadenopathy is usual. Systemic corticosteroids are needed for severe cases. Maintenance of fluid balance and body temperature is important and use of appropriate emollients can be helpful (p. 383). More specific treatment will depend on the underlying cause.

BLISTERING DISEASES Blisters are focal collections of free fluid in the skin. By convention small ones are called vesicles and larger ones bullae. They are the result of a defect or disturbance in the normal mechanisms that hold the components of the skin together. Extensive blistering produces a serious loss of normal skin function, and before effective treatment was available some of the bullous diseases had a high mortality. Many different pathological processes can result in the clinical appearance of a blister. The diagnosis may be obvious from the history, such as in an insect bite reaction, but often there is a differential diagnosis and investigations are necessary. Light microscopy of an early blister, preferably one less than 24 hours old (before re-epithelialization has begun), will provide useful evidence of the level of the split in the skin and sometimes details of the pathogenetic mechanism, e.g. spongiosis (Fig. 10.32). Electron

1

410

Fig. 10.23

2

MCQ 10.11

3

Fig. 10.24

A Sub-corneal: impetigo B Mid-epidermal: viral infections (e.g. herpes simplex, varicella zoster), eczema, friction blister C Supra-basal: pemphigus vulgaris D Basal lamina: pemphigoid, dermatitis herpetiformis E Sub-basal: porphyria F Variable: erythema multiforme, epidermolysis bullosa, blistering drug reactions G Variable: thermal burns (depends on severity of burn) FIG. 10.32 Examples of blistering diseases at different levels in the skin

microscopy allows a much more detailed analysis of blister formation and is particularly important in the accurate diagnosis of epidermolysis bullosa, a heterogeneous group of inherited blistering disorders. Immunological phenomena are central to some of the bullous diseases, particularly pemphigus, pemphigoid and dermatitis herpetiformis, and these are best demonstrated on a sample of unfixed skin from just beyond the edge of a blister (peribullous skin). Other tests that may be indicated include serum for antibodies (see below), bacteriological and viral studies on blister fluid, and, when appropriate, porphyrin estimations and patch testing.

Pemphigus vulgaris Pemphigus vulgaris (Fig. 10.33) is an uncommon, chronic intraepidermal blistering disease of unknown aetiology, characterized by the presence in affected skin and mucosa of an IgG antibody which localizes to the cell wall of keratinocytes (Fig. 10.34). A split forms within the epidermis. Aetiology and pathogenesis The disease is most common in middle age and in Ashkenazi Jews. The characteristic antibody is usually found in the bloodstream, and antibody titres correlate approximately with disease activity. Relapse is sometimes shown to be preceded by an increased titre. The pemphigus antibody probably brings about separation of keratinocytes by

10

FIG. 10.33 Erosions due to pemphigus vulgaris

activating proteinases. Occasionally pemphigus is associated with other immunological diseases, e.g. thymoma, myasthenia gravis and SLE. Clinical features Pemphigus often begins in the mouth and sometimes other mucous membranes, with non-healing erosions. Blisters appear sooner or later on previously normal-looking skin, and when the disease is active sideways pressure with a finger on unblistered skin can produce new blisters (Nikolsky sign). Increasing numbers of flaccid blisters appear and, when they burst, leave erosions. The resultant problems of infections and fluid loss are similar to the consequences of extensive burns. 1 2 Diagnosis The diagnosis is made from characteristic histology, showing a suprabasal split and free-floating keratinocytes (acantholysis). 3 Immunohistochemical techniques show that the keratinocytes are coated with IgG. In all but localized or treated disease the pemphigus antibody can be demonstrated in the serum. Treatment Treatment is with high doses of corticosteroids (about 80mg prednisolone daily), together with an immunosuppressant such as azathioprine, with subsequent gradual reduction of the steroid. Gold salts have also been used with success, but toxicity is a problem. Even with treatment pemphigus remains a serious disorder, up to 25% of patients dying from the disease or the consequences of treatment. Most patients need maintenance therapy for lengthy periods, although in some a true remission occurs.

FIG. 10.34 The lesions of pemphigus vulgaris A, pemphigoid B and dermatitis herpetiformis C (IF = immunofluorescence).

Bullous pemphigoid Bullous pemphigoid (Fig. 10.28) is a fairly common subepidermal blistering disease characterized by an antibody that localizes to the lamina lucida of the basement membrane (Fig. 10.35). Pathology In pemphigoid the complement system is activated, and C3a and C5a attract and activate eosinophils. Eosinophil products and mast cell mediators contribute to the split that occurs in the basement membrane and the clinically evident inflammatory reaction. Clinical features Pemphigoid is mainly a disease of the elderly. It is usually preceded by an erythematous itchy rash, which can be mistaken for urticaria or eczema. It may be localized for some weeks before becoming generalized. The blisters are tense, clear or bloodstained, and often remain intact for several days. Unlike pemphigus, pemphigoid does not usually involve the mucous membranes. When blisters do

411

RECENT ADVANCES IN UNDERSTANDING BLISTERS AT THE MOLECULAR LEVEL The structural integrity of cells is greatly dependent on the three-dimensional network of intermediate filaments within them, and defects in these filaments can result in mechanical weakness. The keratins are a family of intermediate filament proteins maximally expressed in the keratinocytes, and contribute greatly to the strength of the skin. Some forms of epidermolysis bullosa, a group of mostly inherited blistering diseases, have now been attributed to mutations in keratin genes, and the split in the epidermis occurs at the site where the particular keratin is maximally produced. The molecular genetic basis for the most severe types of epidermolysis bullosa, EB letalis and recessive dystrophic EB, has also been discovered recently. EBL is usually due to a defect in the basement membrane protein laminin 5, and RDEB is due to mutations in the gene coding for collagen type 7. In some families it is now possible to make a prenatal diagnosis from a firsttrimester chorionic villus sample. The molecular basis for the immunobullous diseases is now better understood. In pemphigus vulgaris the antigen is desmoglein 3, a transmembrane component of the specialized keratinocyte attachment zones, the desmosomes. In pemphigus foliaceus the antigen is desmoglein 1. In bullous pemphigoid the target antigen is either a 230 or 180 kDa protein in the basement membrane zone.

FURTHER READING Huilgol S C, Bhogal B S, Black M M 1995 Immunofluorescence of the immunobullous disorders. Eur J Dermatol 5:186-195. Mellerio J E 1999 Molecular pathology of the cutaneous basement membrane zone. Clin Exp Dermatol 24:25-32.

break, the skin usually heals and the prognosis is better than that of pemphigus. There is a localized variant of pemphigoid, involving the mucous membranes, in which scarring occurs. Treatment Treatment of pemphigoid is with prednisolone and azathioprine to facilitate steroid withdrawal.

Dermatitis herpetiformis Dermatitis herpetiformis is an uncommon, chronic, intensely itchy disorder in which subepidermal blistering usually occurs.

1 412

MCQ 10.12

2

Fig. 10.25

FIG. 10.35 The blisters of bullous pemphigoid are tense, clear or bloodstained, and often remain intact for several days

Aetiology and pathogenesis There is a close association with HLA-B8 and DRW3, gluten-sensitive enteropathy, and deposition of IgA in the dermal papillae throughout the skin. The enteropathy may not be symptomatic, but it is likely that absorbed products derived from gluten are important in the pathogenesis of the skin lesions. Abnormal immunity in patients with dermatitis herpetiformis is suggested by an increased incidence of organ-specific autoimmune disease, and a few patients (as with coeliac disease) develop small bowel lymphoma. The skin lesions are a result of activation of dense clusters of neutrophils in the dermal papillae, producing damage and then fluid accumulation. Clinical features Dermatitis herpetiformis usually begins in young adults, but can begin at any age and can persist indefinitely. The initial lesions are very itchy grouped weals or papules, on which arise small blisters (Fig. 10.36). Extensor surfaces are the characteristic sites, especially the elbows, knees, shoulders, buttocks and scalp. Oral lesions are sometimes seen. 1 Diagnosis The diagnosis is best made from finding typical histology in an early lesion and IgA deposition in unaffected skin. (In the lesions the IgA can disappear as a result of the inflammatory infiltrate.) A jejunal biopsy usually shows

10

FIG. 10.37 Pityriasis rosea Pink plaques with characteristic scaling centrally.

suggests an infective cause, and a viral aetiology is likely but unproven.

FIG. 10.36 Dermatitis herpetiformis: itchy papules and vesicles

subtotal villous atrophy. Potassium iodide, both systemically and topically, provokes dermatitis herpetiformis and has occasionally been used as a diagnostic test. Treatment Dapsone has a dramatic beneficial effect in this condition, probably by modulating the neutrophil myeloperoxidase enzyme system, and/or the alternative pathway of complement activation. Haemolytic anaemia can occur, in which case sulphamethoxypyridazine is an alternative. A glutenfree diet is helpful in most patients and may reduce the likelihood of intestinal lymphoma.

Clinical features The first lesion to appear, known as the herald patch, is red, circumscribed, slightly scaly and larger than those that will follow about 10 days later. The generalized eruption is maximal on the trunk and proximal aspects of the limbs, and usually includes distinctive ovoid patches (Fig. 10.37) whose long axes are parallel to the major skin creases. 2 This produces a pattern somewhat like a Christmas tree on the back. The pink borders of these lesions are separated from the slightly brown centres by collarettes of scale. There is usually an admixture of smaller pink papules. Itching may occur, and occasionally there is malaise and lymphadenopathy. The rash usually lasts 4-6 weeks. Recurrences are very unusual. Differential diagnosis There are no diagnostic tests. Conditions that may resemble pityriasis rosea include drug eruption, seborrhoeic dermatitis, guttate psoriasis, tinea corporis and secondary syphilis. Treatment A mild topical steroid may be helpful.

FURTHER READING ON BULLOUS DISEASES

Lichen planus

Nausari H C, Anhalt G J 1999 Pemphigus and bullous pemphigoid. Lancet 354:667-671.

Lichen planus (Fig. 10.38) is a fairly common disease that may affect the skin, hair, nails and mucous membranes. The histology is distinctive and shows damage to the basal epidermal cells, with a dense band of lymphocytic infiltrate immediately beneath. There are close similarities to the skin changes seen in graft-versus-host disease and some drug reactions (p. 418).

MISCELLANEOUS DISORDERS OF UNKNOWN CAUSE Pityriasis rosea Pityriasis rosea, a common self-limiting disorder, occurs mainly in children and young adults. Clustering of cases

Clinical features The skin lesions are usually very itchy, and typically form purplish, polygonal, shiny, flat-topped papules. They are usually found on the flexor aspect of the wrists, the lower

413

FIG. 10.38 The lesions of lichen planus typically form purplish, polygonal, shiny, flat-topped papules

back and the shins. Individual lesions often show distinctive fine white lacy patterning (Wickham's striae). Larger, more warty plaques are sometimes found on the lower legs. Induction of new lesions by injury to the skin (Koebner's phenomenon) is often seen. In the mouth lichen planus is seen mainly as a network of white lines on the buccal mucosa and white patches on the tongue, 1 but occasionally can cause chronic ulceration. The scalp may be involved, producing patches of inflammation from which hair fall is usually permanent. The nails can also be affected, usually with longitudinal areas of thinning; in some cases nails also can be permanently lost. Lichen planus often lasts a year or more, and to some extent it is steroid responsive. Postinflammatory hyperpigmentation is common. Treatment Topical steroids may help the pruritus. For widespread symptomatic lichen planus a course of prednisolone, 2030 mg daily for 4-6 weeks, then tapering, can shorten the course of the disease.

Granuloma annulare There are two forms of granuloma annulare, localized (Fig. 10.39) and generalized; as a rule both are asymptomatic. Histologically there is palisading granuloma formation around focal areas of dermal necrosis. Localized granuloma annulare presents as rings of smooth firm papules. The backs of the hands and the tops of the feet are common sites. The commonest misdiagnoses are warts and fungal infection, in which epidermal changes are prominent, in contrast to the smooth surface of granuloma annulare.

414

1

Fig. 10.26

2

Fig. 10.28

2

Fig. 10.27

3

MCQ 10.13

FIG. 10.39 Granuloma annulare, localized form Note the ring-shaped arrangement of dermal papules.

In generalized granuloma annulare the papules are much more widespread and not necessarily in ring-shaped configurations. 2 This is often associated with diabetes mellitus. Diagnosis is confirmed by biopsy. Treatment A potent corticosteroid, topically or by injection, can be used for symptomatic lesions, but granuloma annulare is self-limiting and most cases need no treatment.

Necrobiosis lipoidica This condition (Fig. 10.40) occurs in about 3 per 1000 diabetics, and more than half of patients with necrobiosis lipoidica are overtly diabetic (see Ch. 19). The pathology is similar to that of granuloma annulare but with more pronounced dermal necrosis. The most common site is the shin. The lesion begins as a purplish plaque which gradually enlarges and becomes atrophic, with yellowish areas centrally. Ulceration is quite common. The condition is chronic, and when there is associated diabetes careful control of blood glucose has no effect on the progress of the skin lesions. Treatment The active advancing edge may be responsive to topical or intralesional corticosteroid. Older lesions can sometimes be disguised by cosmetic camouflage.

Sarcoidosis The skin can be invaded in sarcoidosis (Fig. 10.41). Sarcoidosis is discussed on page 682.

10

FIG. 10.40 Necrobiosis lipoidica Subcutaneous vein visible due to central atrophy, bordered by dermal inflammation.

Lichen sclerosus Lichen sclerosus is a distinctive inflammatory condition in which lesions are usually white. The commonest site is the anogenital skin. The cause is unknown. Clinical features Most patients are female, and present either in middle or later life or, less commonly, before puberty. Whereas patients in the older age group tend to have a more protracted course, in prepubertal girls the disease may resolve spontaneously at or around puberty. Usually the vulva is affected, often together with the perianal skin, giving a figure-of-eight appearance. The initial lesion is a white papule; lesions tend to coalesce, and involved skin often becomes wrinkled and atrophic. In the vulva, purpura is common and resorption of the labia may occur. Pruritus is very common. Scratching will often produce excoriations and erythema, which can mask the characteristic whiteness. Dysuria, dyspareunia and, in children, constipation are complications. In the male, involvement of the foreskin and glans can lead to balanitis, phimosis and meatal stricture. Squamous carcinoma is a risk in both sexes. Extragenital lesions occur in a minority. There is an increased likelihood of organ-specific autoimmune diseases. Differential diagnosis Vitiligo can involve the genital skin but does not cause pruritus, and apart from the pallor the skin is otherwise normal. When the pallor is not evident, eczematous conditions, intraepidermal neoplasia and lichen planus may need to be excluded. A biopsy is usually diagnostic.

FIG. 10.41 Sarcoidosis Note the shiny brownish-red indurated swelling of the ala and the two hyperpigmented dermal nodules on the left cheek.

Treatment A superpotent topical corticosteroid twice daily for several weeks initially, and then for shorter periods as necessary, can reverse clinical and histological manifestations in the genital area. Circumcision is often indicated for phimosis. Long-term follow-up is worthwhile in view of the risk of carcinoma. ©

SKIN CHANGES (N DIABETES There are a number of cutaneous problems that occur in diabetics, one of the most common being recurrent infections. Mucosal candidal infections may occur, 4 as may recurrent staphylococcal boils (p. 387). Indeed, in developed societies, the carbuncle (p. 387) is only seen with any frequency in diabetics. Injection sites may become infected, and occasionally deep abscesses may form. Patients with sensory neuropathy may fail to notice repetitive trauma (usually from footwear). This predisposes to ulceration, commonly on the sole (p. 1016). Necrobiosis lipoidica (Fig. 10.33) and granuloma annulare (Fig. 10.32) have histological features in common, and both occur in diabetics (see above). The term 'diabetic dermopathy' refers to small, duskybrown, scar-like lesions that often appear on the shins in diabetics. The patients usually have microangiopathy.

415

TABLE 10.17 Pathogenesis of drug eruptions - some examples

Mechanism Direct effects Immunological IgE-mediated Immune complex Effector pathway stimulation Side-effect Overdosage Cumulative toxicity FIG. 10.42 Xanthelasma Creamy yellow dermal lipid deposits in a diabetic with hyperlipidaemia.

Indirect effects Drug interaction Metabolic effects Ecological disturbance

'Diabetic sclerosis' or 'cheiroarthropathy' occurs in at least 30% of juvenile-onset insulin-dependent diabetics. The fingers and toes become stiffened and sclerodermatous. The patient is unable to straighten the fingers completely, and cannot place them flat on a table. A rarer condition is scleroedema, in which the skin of the neck and upper back suddenly becomes stiff and thickened. These changes gradually spread on to the face, trunk and upper arms. Rarely, diabetics develop distinctive non-inflammatory blisters on the extremities. Hyperlipidaemia is a common complication of diabetes, and lipid deposits in the skin may occur in the form of eruptive xanthomata, most commonly on the buttocks and extensor surfaces. Palpebral xanthelasmata may also be seen in diabetics (Fig. 10.42). Injection-site lipoatrophy and lipohypertrophy are discussed on page 992. Rarely, glucose intolerance and diabetes are part of a constellation of changes, including generalized fat reduction (lipodystrophy) and acanthosis nigricans. Such patients often have marked insulin resistance. There is an association between vitiligo and maturity-onset as well as insulin-dependent diabetes.

DRUG ERUPTIONS

416

Drug eruptions are a frequent occurrence in hospital practice and can produce skin changes in many ways (Table 10.17). Some of these, such as the predisposition to candidal infection by a broad-spectrum antibiotic, are not direct effects of drugs on the skin, but are none the less a consequence of drug exposure. Often the exact mechanism of a drug eruption is unknown. Some drugs - such as antibiotics, thiazides and sulphonamides, gold, allopurinol, phenylbutazone and other non-steroidal anti-inflammatories - are relatively

Exacerbation of existing disease

Drug

Clinical pattern

Penicillin Sulphonamides Acetyl salicylic acid

Urticaria Vasculitis Urticaria

Cytotoxic drugs Anticoagulants Corticosteroids

Alopecia Purpura Striae

Aspirin-displacing warfarin Isoniazid Broad-spectrum antibiotics Hydralazine

Purpura

Lithium

Pellagra Candidiasis Lupus erythematosus Psoriasis

SUMMARY 4 Skin changes in diabetes mellitus • Cutaneous infections

• • • • • • • • •

Neuropathic ulcers Necrobiosis lipoidica (diabeticorum) Disseminated granuloma annulare Diabetic 'dermopathy' Diabetic sclerosis (cheiroarthropathy) Scleroedema adultorum (of Buschke) Bullae (bullosis diabeticorum) Xanthomata and xanthelasmata Effects of insulin injections

- candidiasis - staphylococcal infections - injection site infection

lipoatrophy lipohypertrophy (Insulin tumours') infections and abscesses

• Generalized lipodystrophy (lipoatrophy not due to insulin injection)* • Acanthosis nigricans* • Vitiligo * When these cutaneous signs accompany diabetes, there is usually gross insulin resistance

common causes of rashes, and others very rare, but in all patients who develop a rash the drug history is of great importance. Not only prescribed drugs, but also medicines and tablets bought over the counter, must be considered.

TYPES OF DRUG ERUPTION Urticaria The mechanisms whereby a drug can cause urticaria include IgE-mediated immediate hypersensitivity, immune complex-mediated generation of activated complement components (serum sickness), a direct action on mast cells causing them to release histamine, and modulation of arachidonic acid metabolism. The drugs commonly implicated in urticaria are: • Penicillin and related antibiotics • X-ray contrast media • Enzymes • Blood products

• Opiates • Non-steroidal antiinflammatory drugs • Pollen vaccines • Iodides.

Angioedema and anaphylaxis may accompany urticaria, particularly when the drug has been injected, and can be life-threatening. Urticaria usually begins within minutes or hours of the drug being given, but when it is immune complex-related it occurs several days after the challenge and is often associated with fever, lymphadenopathy, joint symptoms and haematuria as a result of renal damage.

Ampicillin, amoxicillin and derivatives, e.g. talampicillin, are common causes, especially when the patient has infectious mononucleosis or lymphatic leukaemia, or is also taking allopurinol, and can begin up to a few days after the antibiotic has been stopped. The common causes of morbilliform (exanthematic) drug eruptions are: • • • • • •

Allopurinol Antituberculous drugs Captopril Carbamazepine Gold salts H2 antihistamines

• • • • • •

10

Penicillamine Penicillins Phenothiazines Phenylbutazone Sulphonamides Thiazides.

Erythroderma When caused by a drug, erythroderma, or exfoliative dermatitis (p. 409), tends to begin several weeks after the drug has been started. Important causes of drug-induced erythroderma are: • • • • • •

Allopurinol Carbamazepine Phenytoin Isoniazid Lithium Gold salts

• • • • • •

Chloroquine Barbiturates p-aminosalicylic acid Captopril Sulphonamides Methyldopa.

Morbilliform eruption A widespread, symmetrical, blotchy, maculopapular erythematous rash (Fig. 10.43) is probably the commonest drug eruption. There is often a mild fever, but serious consequences are uncommon. The mechanism is usually obscure but does involve delayed hypersensitivity in some cases, especially when the onset is within a few days of the drug being started. Morbilliform eruptions usually begin within a week of the onset of the drug, and may progress to erythroderma if the drug is continued.

Erythema multiforme Erythema multiforme (p. 407) is a reaction distinguished by target lesions. Most cases are not due to drugs. The commonest drugs to cause erythema multiforme are: • • • •

Sulphonamides Phenytoin Phenylbutazone Barbiturates

• • • •

Penicillins Carbamazepine Rifampicin Gold salts.

Toxic epidermal necrolysis

FIG. 10.43 Morbilliform drug eruption Maculopapular erythema brought on by co-trimoxazole.

Toxic epidermal necrolysis is a rare drug reaction which has a high mortality. There is often a brief prodrome of malaise and fever, followed by widespread, tender erythematous areas which then blister. Large sheets of epidermis readily rub off with light pressure, to leave painful denuded dermis. Mucous membranes as well as skin may be involved. Fluid imbalance, septicaemia and pneumonia are the most common problems. Differentiation from staphylococcal scalded skin syndrome (p. 385) can be made histologically on a blister roof because in toxic epidermal necrolysis the roof consists of the whole epidermis. Butazones, Sulphonamides, allopurinol, gold salts and phenytoin are examples of drugs that can cause the syndrome, which can also be due to infections, graft-versus-host disease and lymphoma. Management is similar to that of widespread burns, with treatment of fluid and protein loss and of infection.

417

Allopurinol Thiazides Phenytoin Thiouracil Non-steroidal antiinflammatory drugs

Photosensitivity The most common photosensitivity drug reaction resembles sunburn and is usually phototoxic, i.e. does not involve immunological mechanisms. Occasionally drugs produce photoallergic reactions which may look like eczema on light-exposed skin. Drugs which can induce phototoxic and photo-allergic reactions include: Phenothiazines Thiazides Sulphonamides Tetracyclines Sulphonylureas

• • • • •

Nalidixic acid (bullous) Amiodarone Azapropazone Protriptyline Psoralens.

Drugs can also induce photosensitive diseases, e.g. procainamide can induce lupus erythematosus, and isoniazid pellagra.

Fixed drug eruption The characteristic feature of a fixed drug eruption is that inflammation occurs in exactly the same place or few places each time the drug is taken. The reaction is usually a round red patch, which may blister (Fig. 10.44), and after the inflammation has subsided there is often prolonged hyperpigmentation. The causes change with variations in drug use, but a recent survey in the UK identified several common offenders - paracetamol, non-steroidal anti-inflammatory drugs, fluconazole, terbinafine, sulfasalazine, tetracylines, trimethoprin, diltiazem and proton pump inhibitors.

• • • •

Sulphonamides Hydralazine Quinidine Captopril.

Erythema nodosum Drugs are rarely responsible for erythema nodosum, and other causes should be sought (p. 407). Oral contraceptives and Sulphonamides are the drugs most likely to be involved.

Pigmentation changes induced by drugs Colour change (Table 10.18) can be produced by deposition of the drug in the skin and mucous membranes, stimulation of melanin production and alteration of the distribution of pigment so as to make it more apparent, as in postinflammatory hyperpigmentation (p. 427). The hyperpigmentation after fixed drug eruption may be the presenting feature. In some situations the pigmentation is only evident in light-exposed skin. In many examples of drug-related pigment change the mechanism is unknown.

Lichenoid drug eruption

Vdsculitis

The rash in lichenoid drug eruption is similar to that of lichen planus (p. 413). It usually begins weeks to months after the drug is started. As in idiopathic lichen planus, hyperpigmentation is common. Important causes of lichenoid drug eruptions are:

Drug-induced vasculitis (p. 408), usually in the form of palpable purpura, and often with urticarial and blistering lesions, can be accompanied by similar lesions in other organs and may be a serious illness. The commonest drugs to cause allergic vasculitis are:

• • • • •

Gold salts Antimalarials Thiazides Phenothiazines Methyldopa

• • • •

Furosemide (frusemide) p-Blockers Penicillamine Chlorpropamide.

Drug-induced lupus erythematosus

418

FIG. 10.44 Fixed drug eruption: dusky erythema with central blister

Drug-induced lupus erythematosus has some differences from the naturally occurring condition (p. 1172). It is likely that affected individuals have a genetic predisposition. This has been best documented in association with hydralazine and procainamide. With hydralazine the disorder is broadly dose-related, being seen in 5% of those on 100 mg daily, in 10% of those on 200 mg daily, but not in patients on 50 mg daily. Other common causes include penicillamine, phenytoin, methyldopa, B-blockers, sulfasalazine and oral contraceptives. In most cases druginduced lupus erythematosus occurs after the drug has been taken for at least a few months. On discontinuation, the disease resolves in most cases, but can persist for years. The condition is relatively less common in black people and more common in the elderly; renal and central nervous system disease are less likely to occur than in idiopathic

TABLE 10.18 Pigmentation caused by drugs

TABLE 10.19 Eczertia induced by topical drugs

Patterns

Eliciting drug

Source of topical exposure

Antibiotics, e.g. penicillin, streptomycin Antihistamines Aminophylline

Contents of medical and veterinary ampoules

Drug

Brown Similar to Addison's disease Similar to chloasma Light-exposed areas

Generalized Generalized, with patchy pigmentation Blue-grey Light-exposed areas

Generalized, maximal on light-exposed areas

Shins, nails and palate (and sometimes light-exposed areas) Yellow Generalized

Red Generalized

ACTH Phenytoin Oral contraceptives Phenothiazines Psoralens Cytotoxics Cytotoxics e.g. busulfan Pyrimethamine Arsenic

Minocycline Amiodarone Phenothiazines Gold Silver Bismuth Antimalarials

Mepacrine

Clofazimine Methysergide

lupus erythematosus. Antihistone antibodies are characteristic, and antinative DNA antibodies are not found. Complement levels remain normal.

Purpura When caused by a drug, purpura is usually a result of thrombocytopenia, vascular damage or both. Mechanisms include both toxic and allergic effects.

Pruritus Itching without a rash is considered on page 433. Drugs can cause generalized pruritus and the most common causes are: • Opiates • CNS stimulants • Antidepressants

• Oral contraceptives • Hepatotoxic drugs • Chloroquine.

For many drugs the mechanism is unknown, but some produce intrahepatic cholestasis.

Acne and acneform eruption In true acne there are comedones, papules and pustules in

10

In antipruritic creams, e.g. mepyramine Cross-reaction with ethylenediamine in Tri-adcortyl cream

a characteristic distribution (p. 431). Drugs exerting an androgenic stimulus, such as testosterone and some progestogens, can exacerbate or induce acne. An acne-like eruption, in which there are follicular lesions but in which the morphology or the distribution differs from true acne, is seen with a number of drugs, including corticosteroids and ACTH, halogens, antiepileptic drugs, isoniazid and lithium salts.

Alopecia Partial or complete hair loss is very common and appears early with cytotoxic drugs, but can occur, rarely and late, with other drugs, e.g.: • Cytotoxic agents (especially cyclophosphamide and anthracyclines) • Retinoids

• • • • •

Anticoagulants Oral contraceptives Antithyroid drugs Phenytoin Valproate.

Hypertrichosis Corticosteroids and androgenic steroids can produce hirsutism, but some drugs can induce increased hair growth on areas not dependent on a sex hormone stimulus. Minoxidil, ciclosporin, diazoxide, phenytoin and penicillamine are the main causes.

Eczema If sensitized via the skin to a 'drug' and then 'challenged' by an internal route, the patient can develop an often widespread eczematous eruption. Examples are given in Table 10.19.

Vesicular and bullous drug eruptions Blistering can occur in several different types of drug eruption already described, e.g. fixed eruption, erythema multiforme, photosensitivity and vasculitis. Bullae are common at sites subjected to prolonged pressure in drug-induced coma. Porphyria (p. 1026) can be precipitated by drugs, as occasionally can pemphigus (e.g. rifampicin) and acquired epidermolysis bullosa (e.g. furosemide [frusemide]).

419

INVESTIGATION AND TREATMENT OF DRUG REACTION The evaluation of a possible drug reaction should include the following: • A careful assessment of the risk to the patient, from the rash itself and from the involvement of other organs, e.g. kidneys and brain in a vasculitis; an assessment of the airway and maintenance of adequate circulation in urticaria/anaphylaxis. • A full drug history, including all current and recently completed treatments (tablets, mixtures, injections, suppositories etc.) and their timing in relation to the onset of the rash. Some drugs can cause a rash several days after they were last given, e.g. ampicillin, gold and depot preparations. • Past drug history and associated rashes. • Other possible explanations for the rash, e.g. a virus infection. • Consideration of which drugs can be stopped. Nonessential drugs should be stopped and essential drugs changed to structurally different ones if possible. • Laboratory investigations. Except for the investigation of drug-induced blood dyscrasias, these are disappointing. Blood eosinophilia, if present, is supportive of drug eruption. • A skin biopsy may help in some cases. • Treatment. The most immediately serious drug reaction is anaphylaxis. Emergency treatment is with 1:1000 adrenaline, 1 mL i.m., or slowly i.v. if the patient is moribund. Parenteral antihistamine and hydrocortisone are also given, and if necessary the patient is intubated and managed on ICU.

and vasculitis will require treatment with systemic corticosteroids. FURTHER READING ON DRUG ERUPTIONS Breathnach S M, Hintner H 1992 Adverse drug reactions and the skin. Oxford: Blackwell Scientific.

INTERNAL MALIGNANCY AND THE SKIN Skin abnormalities may be the first indicator of a systemic malignancy (Table 10.20). The cause of many other skin disorders associated with malignancy is unknown. Some are: • Generalized pruritus without a primary rash (p. 433). • Dermatomyositis, adult onset 1(p- 1180). • Acanthosis nigricans (Fig. 10.45). The skin becomes hyperpigmented and warty but feels soft. Body folds, e.g. axillae, are affected. There are other causes, such as obesity, some drugs and some inherited syndromes, but onset in a lean adult is highly likely to be associated with an adenocarcinoma or lymphoma. • Acquired hypertrichosis lanuginosa. This is a rare condition but one that has a strong association with malignancy, especially Hodgkin's disease. The patient

Depending on severity, some patients with erythroderma, bullous erythema multiforme, toxic epidermal necrolysis TABLE 10.20 Skin manifestations of malignancy Direct involvement e.g. Direct spread from breast carcinoma; plaques and nodules from lymphoma Genetic predisposition e.g. Multiple cysts and benign skin tumours occurring with colonic carcinoma with autosomal dominant inheritance (Gardner's syndrome) Carcinogen (also causing skin disease) e.g. Vinyl chloride skin eruption and angiosarcoma of the liver following exposure to vinyl chloride monomer (used in plastic manufacture) Metabolic products of tumour e.g. Flushing due to malignant carcinoid; migratory necrolytic erythema with glucagonoma - partly due to essential amino acid deficiencies induced by metabolic abnormality

420

1

Fig. 10.29

2

Fig. 10.30

3

Fig. 10.31

4

Figs 10.32,10.33

FIG. 10.45 Hyperpigmentation and velvety thickening due to acanthosis nigricans

•

• •

•

develops blond downy hair over the whole face and then elsewhere. Erythema gyratum repens. Here the skin develops irregular wavy bands of erythema, the overall appearance resembling wood grain. Nearly all cases have an associated malignancy. Acquired ichthyosis. The unexplained occurrence of dry, scaly but non-inflamed skin is often associated with lymphoma (p. 405). Pachydermoperiostosis with finger clubbing. The skin becomes thickened, with extra folds; the hands, elbows, knees and tongue enlarge and there may be painful periosteal new bone formation (see hypertrophic pulmonary osteoarthropathy, p. 619). These changes are associated with bronchial carcinoma. Pancreatic panniculitis. Tender red nodules of distinctive histology, owing to the effect on subcutaneous fat of circulating lipases from the pancreas, indicate either malignancy or inflammation in that organ. There is often arthritis and eosinophilia.

NAEVI AND TUMOURS OF THE SKIN Most primary tumours are either clearly benign or malignant, but occasionally some benign lesions evolve into malignant patterns of behaviour and are termed premalignant. Some of the known causes are listed in Table 10.22 (p. 424). A naevus is a circumscribed developmental defect, and the term has a similar meaning to hamartoma. Naevi are often, but not necessarily, evident at birth. There may be more than one cell type involved in the process. Each of the cell types in the skin can proliferate to produce a tumour, and the exact diagnosis usually rests with the pathologist. Many benign tumours and naevi are of little more than cosmetic importance, but in some instances they are pointers towards a genetic syndrome or multisystem disorder. Malignant skin tumours can be fatal, and often come to attention during a routine medical examination. Malignant metastases in the skin can occasionally be the presenting feature of a cancer elsewhere, and the histology can provide a valuable indication of the organ of origin.

The blue naevus is a slate-coloured lesion composed of a localized proliferation of melanocytes in the dermis which never reached the epidermis during development. Malignant melanoma (p. 424) may arise from a seemingly benign melanocytic naevus, but this is rare except from giant congenital pigmented naevi.

10

Vascular naevi There are several types of vascular naevi, some of which resolve in infancy or childhood (e.g. the salmon patch and strawberry naevus). The port wine stain is an area of permanent vascular dilatation, often on the face, where it may produce a disfiguring purplish discoloration. When the area affected is in the distribution of the ophthalmic division of the trigeminal nerve, there may be an associated vascular anomaly in the brain or meninges which can cause epilepsy, hemiplegia or retardation (Sturge-Weber syndrome), and various ocular abnormalities, especially glaucoma, can occur. The skin change can be successfully treated with the dye laser. Epithelial naevi Developmental malformations involving epidermal keratinocytes affect at least 1 in 1000 live births. They are usually circumscribed verrucous plaques which are often linear on the limbs, but may assume quite bizarre, whorled configurations, especially on the trunk. Occasionally lesions may be widespread. Such multiple lesions may be associated with developmental defects in the CNS and skeleton.

Naevi and multisystem disease

Naevi

Tuberous sclerosis Tuberous sclerosis is an autosomal dominantly inherited disorder in which mental deficiency is associated with an abnormality of many other organs. The first skin lesions to appear, often in early infancy, are the 'ash leaf macules. These well circumscribed, often irregularly shaped pale patches are best seen with Wood's light. From later childhood onwards, one or more of the following may appear: red angiofibromatous papules on the face (misnamed adenoma sebaceum), 2 thickened plaques on the trunk (shagreen patches) and periungual fibromas 3 (smooth nodules beside the nails). The skin lesions can be important in recognizing carriers of the gene.

Pigmented naevi Pigmented (melanocytic) naevi, or common moles, are occasionally evident at birth, but more commonly appear in childhood or adolescence. The naevus is initially brown, with clusters of proliferating melanocytes at the base of the epidermis. Melanocytic naevi that develop after birth are usually symmetrical in shape and colour and rarely grow to more than 1cm in diameter. In time the cells migrate into the dermis, and the lesion often becomes a fleshcoloured nodule.

Von Recklinghausen's disease (neurofibromatosis) The many features of von Recklinghausen's disease, or neurofibromatosis (Fig. 10.46), an autosomal dominantly inherited condition, include acoustic neuroma, glioma, spinal deformities and phaeochromocytoma. The skin shows multiple pale brown patches (cafe au lait macules}, especially in the axillae (a pathognomonic sign), and soft nodules (neurofibromas) 4 composed of nerve sheath cells.

421

FIG. 10.46 Neurofibromatosis is associated with localized areas of hyperpigmentation

Benign tumours and cysts The growth of these is usually slow, does not destroy normal tissues and is eventually self-limiting. The lesion produced is generally symmetrical in shape. Seborrhoeic warts are one of the commonest skin abnormalities, occurring with increasing frequency with advancing years. The lesions vary in colour from pale brown to very dark brown/black, and usually look as though they have been 'stuck on'. The surface is irregular and rather greasy to the touch. The simplest treatment is freezing with liquid nitrogen, although this is only necessary if the lesion is causing symptoms. Dermatofibromas (histiocytomas) normally present as firm papules, 0.5-1 cm in diameter (although occasionally they are significantly larger). The lesion moves easily within the skin and is never fixed to deeper tissues. If the diagnosis is certain dermatofibromas can safely be left, but excision is the treatment of choice if this is deemed necessary. Pyogenic granulomas grow rapidly, often following tauma and often on a digit. They are more common in children and young adults, but may be seen at any age.

1

422

Fig. 10.34

The surface is red, friable, and bleeds easily, consisting as it does of vascular proliferative tissue. In most instances a thin epithelium eventually extends to cover the raw surface, and lesions may involute spontaneously. It is more practical, however, to remove the whole lesion by curettage or by excision; the tissue should always be sent for histopathology. Smaller tumours may respond to cryotherapy. Skin tags develop in clusters around the neck, axillae and groins, more commonly in the obese (although there appears to be an inherited tendency in some families). Occasionally there may be a background of mild acanthosis nigricans (see p. 420). Lesions catch on clothing and jewellery and can be removed easily with a pair of scissors or by diathermy. Lipomas may be solitary or multiple, a tendency to numerous lipomas being inherited as an autosomal dominant trait in some families. They feel soft, subcutaneous and, as they enlarge, lobulated. Some lesions are mildly painful when knocked. There is no satisfactory therapy other than formal excision. Kemtoacanthomas are rapidly growing and, by definition, self-limiting tumours with a histopathology that may be difficult to distinguish from a well-differentiated squamous cell carcinoma. They arise most commonly on the head and neck, the hand and forearm and on sun-damaged skin in general. The normal course of events is for the lesion to grow rapidly over 4-6 weeks, reaching a maximum of about 2-3 cm in diameter. Growth then ceases, and a period of quiescence is followed by equally rapid shrinkage and disappearance. There is often a small pit left in the skin. The lesions are highly characteristic in shape, annular with a central keratin plug, 1 but these feaures are shared by some invasive squamous cell carcinomas (see below) and complete removal is recommended unless the patient is very old and frail. Epithelioid cysts and pilar cysts are both frequently called 'sebaceous cysts' by the terminologically inexact. Both consist of spheres of epithelium producing keratinous material internally. This results in a smooth, round swelling which moves easily over deeper structures in most instances. Occasionally inflammation and subsequent scarring leads to a tethered feel. Epithelioid cysts may occur anywhere, but are most common on the face, neck and trunk. They frequently follow in the wake of severe acne and there may be a small punctum on the surface. Pilar cysts are essentially a malformation of hair root sheath and are normally found on the scalp. There is often a positive family history. Another common keratinizing cyst is the milium, which are usually multiple (milia). They may arise in scars or other areas of trauma, but are most commonly found on the face, either in clusters around the eyes or more widely scattered, when they may be associated with signs of chronic sun damage. Keloid. A keloid is an excessive growth of connective tissue after an injury (including surgery or infection). Unlike a hypertrophic scar, it tends to grow beyond the limits of the initial trauma and can occur spontaneously.

Keloids are very long lasting. They are firm, smooth, rounded, protuberant swellings, often itchy and sensitive. In whites they tend to be purplish, and in dark-skinned patients they are often hyperpigmented. They are common in black races and at certain sites, e.g. the sternum, shoulders, upper back and the beard area. Treatments include injections of corticosteroid, compression and applications of silicone sheeting.

sites. In some cases there is an association with internal malignancy. Bowen's disease presents as a slow-growing, well-defined, red scaly or crusted patch. It is often solitary, although it can be multiple. The major differential diagnosis is psoriasis. Small lesions can be treated as for actinic keratosis, but for larger lesions excision is preferable.

10

Malignant and premalignant tumours Precursors to squamous carcinoma Actinic keratosis Actinic (solar) keratosis is a lesion that arises on skin chronically damaged by the harmful effects of UV radiation, in particular UVB (p. 426). The face (Fig. 10.47), bald scalp, backs of hands and forearms are common sites. Actinic keratoses are often multiple; they are more likely in the blue-eyed, fair or red-haired, and rare in darkskinned people. The individual lesion usually presents as a papule or plaque of hard, often brownish keratin. Transformation to squamous carcinoma is uncommon. Actinic keratoses can be treated by a number of modalities, e.g. cryotherapy, curettage and 5-fluorouracil cream. Bowen's disease Bowen's disease can be a consequence of previous ingestion of inorganic arsenic, usually given as a tonic many years ago. It is more common on covered than on exposed

FIG. 10.47 Solar keratosis A focal area of hard keratin accumulation with a zone of underlying erythema on sun-damaged skin.

Squamous carcinoma More than in other skin malignancies, there is likely to be an identifiable cause (Table 10.21) for squamous carcinoma (Fig. 10.48). The tumour often arises from a premalignant condition and is distinguished clinically by induration of the base and surrounding normal tissues. With growth, ulceration usually occurs. Squamous carcinoma is more liable to metastasize to regional lymph nodes than is basal carcinoma, and distant metastases can occur. Diagnosis is by biopsy. Keratoacanthoma (see Table 10.21) can be difficult to distinguish both clinically and histologically from squamous cell carcinoma. Treatment will usually be by surgical excision, although in some circumstances radiotherapy may be preferable. Basal cell carcinoma Basal cell carcinoma (BCC), or rodent ulcer (Fig. 10.49), arises from cells that resemble the basal layer of the epidermis, and may have its origin from skin appendage epithelium. It is the commonest malignant skin tumour in white skin, and although sunlight exposure is clearly a major aetiological factor, unlike squamous cell carcinoma it is not often seen on the backs of the hands. The face is the commonest site, although when there has been exposure to a systemic carcinogen, such as inorganic arsenic, BCC can be multiple and widespread.

FIG. 10.48 Squamous carcinoma is characterized by induration of the base of the tumour and eventual ulceration

423

TABLE 10.21 Factors predisposing to primary skin malignancy • • • • • • • •

Ultraviolet radiation X-rays Chemicals, e.g. hydrocarbons in mineral oil, inorganic arsenic Immunosuppression Genetic, e.g. xeroderma pigmentosum Scars, e.g. old burn scar Long-standing skin disease, e.g. lupus vulgaris, chronic leg ulcer Some human papilloma viruses (rare)

TABLE 10.22 The rising tide of melanoma USA lifetime risk (to age 75) 1935

1:1500

1991

1:105

2000

1:75

Melanoma is: • the commonest cancer in young adults aged 25-29 • the commonest cancer in males aged 30-34 • the second commonest cancer (to breast) in females aged 30-34

FIG. 10.50 Superficial spreading melanoma usually grows with an irregular edge and variable pigmentation

FIG. 10.49 Basal cell carcinoma is often translucent, with overlying telangiectatic vessels

The tumour is usually slow-growing, evolving from a translucent papule. Various growth patterns may occur, e.g. superficial spreading, nodular, infiltrative and sclerosing. Most nodular and infiltrative tumours ulcerate. Except for the sclerosing variety, BCC tends to retain a translucent raised margin as it extends into surrounding tissues. Telangiectatic vessels are often visible on the surface and

424

Fig. 10.36

1

Fig. 10.35

2

4

Fig. 10.37

0 Fig. 10.38

7

Fig. 10.39

3

Case 10.2

6 MCQ 10.14

may bleed on contact. 1 In some tumours there is an admixture of melanocytes, and the resultant lesion can simulate malignant melanoma. Local destruction can be extreme, 2 and very rarely BCC metastasizes. Treatment is usually by excision, radiotherapy, curettage or cryotherapy. Malignant melanoma Malignant melanoma (Fig. 10.50) usually arises either from previously normal-looking skin, 3 4 or from a benign melanocytic naevus. It may also arise from a naevus in a nailbed, in a mucous membrane or from the choroid or the iris. At present there is a rapidly rising incidence among white people (Table 10.22) and, aetiologically, short bursts of sun exposure, as during a hot, sunny holiday, may be important. There is not the same association with chronic sun damage as with squamous cell carcinoma, and to a lesser extent BCC. Conditions that can give rise to melanoma are lentigo maligna (Hutchinson's freckle) and dysplastic naevus. Lentigo maligna evolves as a flat, brown, variably pigmented patch, which slowly spreads on markedly sun-damaged skin. 5 Dysplastic naevi are

unusual-looking, often larger than average, with a distinctive histology; they are sporadic or inherited as an autosomal dominant characteristic, and may be a precursor to melanoma. Melanoma is often suspected when an enlarging pigmented lesion has one or more of the following: irregular notched border; irregular pigmentation, often with red and white as well as brown/black areas; itching or prickling sensation; inflammatory halo; ulceration and bleeding. The malignant cells spread laterally, in and just beneath the epidermis, and inwards. The prognosis depends on the depth of invasion, being excellent for completely excised, very superficial tumours. Metastasis occurs both to regional lymph nodes and distantly. ©

plaques and tumours appear (Fig. 10.51). Itching usually occurs. Erythroderma (p. 409) can develop, and if there are circulating neoplastic T cells this constitutes Sezary's syndrome. Lymphadenopathy and visceral involvement occur late and indicate a poor prognosis.

10

Metastatic malignancy Nodules, often ulcerating, may occur with many visceral malignancies. Paget's disease of the nipple The epidermis of the nipple becomes invaded by malignant cells arising from an underlying intraduct carcinoma of the breast tissue. The lesion is red, crusted and well defined, like Bowen's disease.

Lymphoma and leukaemia in the skin B-cell lymphoma, Hodgkin's disease and leukaemia may occasionally metastasize to the skin, or rarely present with cutaneous nodules. The skin is usually involved rather late in the course of the disease. The diagnosis is made from biopsy and other characteristic features of these diseases. T lymphocytes have an affinity for the epidermis and its appendages, and therefore T-cell neoplasia often involves the skin from the outset and may appear to be localized to the skin in some cases. Mycosis fungoides is the best-characterized T-cell cutaneous neoplasm. There is often a long phase of poikilodermatous patches (poikiloderma is a combination of erythema, atrophy and reticulate pigmentation) before

Kaposi's sarcoma Kaposi's sarcoma (Fig. 10.52) probably has a viral aetiology - humanherpesvirus 8. The lesions of this multicentric vascular neoplasm are usually purplish patches, plaques or nodules. Leakage of blood readily produces purpura and brown staining of the skin. The lymphatics can be affected, producing lymphoedema. Initially the lesions may be insignificant-looking, flat, reddish-brown or purple patches. ©A distinctive feature of AIDS-related Kaposi's sarcoma (p. 449) is a tendency for ovoid lesions to orientate along major skin creases, which on the trunk resembles the arrangement seen in pityriasis rosea.

FIG. 10.51 Mycosis fungoides This patient has numerous infiltrated plaques of T cell lymphoma in the skin.

FIG. 10.52 Kaposi's sarcoma in an AIDS patient There are multiple purple plaques and nodules.

425

Treatment, when indicated, is radiotherapy for localized lesions and, in some cases, cytotoxic chemotherapy for extensive disease. Primary sarcomas Malignant tumours arising from the various mesenchymal elements in the skin are all rare and present as enlarging masses.

AIDS AND THE SKIN

TABLE 10.23 Skin and oral lesions in AIDS Seborrhoic dermatitis (early feature in 30%) Itchy folliculitis (Fig. 10.53) Kaposi's sarcoma Infections Molluscum contagiosum Perianal warts Superficial fungal infection Herpes zoster and simplex Atypical mycobacteria

Severe drug eruptions Exacerbation of psoriasis Vasculitis Diffuse alopecia Dry skin (late) Oral hairy leukoplakia (see Fig. 11.9) Oral candidiasis Oral Kaposi's sarcoma

Skin and oral lesions are common features in patients with symptomatic HIV infection (Table 10.23). TABLE 10.24 Electromagnetic spectrum

SUNLIGHT AND THE SKIN For descriptive purposes, the continuum of electromagnetic radiation emitted by the sun is divided into segments according to wavelength (Table 10.24). Of the energy reaching the earth's surface, 99% is ultraviolet through to middle infrared. The shorter wavelengths of ultraviolet are filtered out by ozone in the stratosphere and by the atmosphere. The response of skin to solar radiation depends on the wavelength and on natural defences, the most important being melanin deposition. Ultraviolet B (UVB) is responsible for acute sunburn and the chronic changes of sun exposure: atrophy, dryness, blotchy pigmentation and wrinkling, solar keratoses and many skin cancers. The consequences of failure to repair damage inflicted on cellular DNA by UVB are seen in the rare autosomal recessively inherited condition xeroderma pigmentosum, in which sun-exposed skin becomes prematurely aged and skin cancers can appear in childhood. UVA has a much less obvious effect on normal skin, but is often responsible for photosensitive drug reactions and polymorphic light eruption. A clinical clue that UVA is playing a part is the provocation of a sun-related skin disorder behind window glass, as this filters out UVB. The action spectrum, i.e. the determination of exactly which wavelengths are responsible for a photosensitivity disorder, is known for only a few diseases. Visible light does not produce reactions on normal skin, but is responsible for photosensitivity in some diseases, notably the porphyrias, where the maximal reactivity is around 400 nm. Although most of the consequences of solar radiation on the skin appear harmful, one benefit is the synthesis of vitamin D3. UV radiation is essential to break the steroid

Type

Wavelength

Gamma Vacuum UV Ultraviolet C (UVC) Ultraviolet B (UVB) Ultraviolet A (UVA) Visible light Near infrared Middle infrared Far infrared Microwaves and radiowaves

0.1-100A 10-200 nm 200-290 nm 290-320 nm 320-400 nm 400-760 nm 0.74-1.5 um 1.5-5.6 um 5.6-1000 um >1mm

B ring and convert 7-dihydrocholesterol to previtamin D3, which then isomerizes spontaneously to vitamin D3. The immune system in the skin is modified by UV exposure, which may be important in the development of skin cancer.

Polymorphic light eruption Polymorphic light eruption is a common disorder that occurs mainly in females. Usually in spring, a day or so after sun exposure, itchy red papules, plaques and sometimes vesicles appear on sun-exposed sites. The condition gradually settles with continued sun exposure.

Drug photosensitivity It is usually not known whether allergic or toxic mechanisms are involved in drug photosensitivity. The drugs commonly responsible are given on page 418.

Phytophotodermatitis 1 426

1

Fig. 10.40

4

MCQ 10.15

2

Fig. 10.41

0 Fig. 10.42

Psoralens, a group of chemicals found in many plants, can be activated by UVA to produce inflammation and hyperpigmentation; 8-methoxypsoralen with UVA is used

TABLE 10.25 Causes of widespread hyperpigmentation Condition

Cause

Hypoadrenalism Cushing's syndrome

ACTH excess

Renal failure

Vitamin B12 deficiency

MSH Carotenoids Melanin Haemosiderin Melanin

Any debilitating diseasf disease

Melanin

Drugs

Melanin and sometimes the drug Keratin

Haemochromatosis

FIG. 10.53 Itchy folliculitis Excoriated follicular papules in a patient with AIDS.

therapeutically as PUVA (p. 418). Contact with a plant containing a psoralen (e.g. common rue, giant hogweed), together with sun exposure, produces erythema, often blistering, and then considerable hyperpigmentation.

Photoallergic contact dermatitis A number of cosmetics and even sunscreens contain chemicals which are activated by UV and, in combination with skin constituents, generate photoallergy. Investigation is by patch testing using two sets of patches, one of which is exposed to UV after a period of contact with the skin. Photocontact allergy produces positive results only in the set that has been irradiated. In some patients the state of photosensitivity persists long after the allergen has been withdrawn.

Metabolic disorders with light sensitivity Skin sensitivity to sunlight may occur in porphyrias (except the acute intermittent variety 2)> pellagra, carcinoid syndrome 3 and Hartnup disease.

Disorders aggravated by sunlight Many skin diseases may be worsened or triggered by sunlight. Common examples are lupus erythematosus and recurrent facial herpes simplex. Careful use of sunscreens can be helpful. 4

Acanthosis nigricans

10

Pattern Maximal in flexures, creases of palms and soles, sites of friction and pressure, and buccal mucous membranes Diffuse, maximal on hand and face Exposed skin and flexures Widespread, accentuated over knuckles Hypomelanosis of hair Generalized or Addisonian See p. 419, Table 10.18 Flexures, sides of the neck

the melanocytes and distributed in cellular organelles called melanosomes via cell processes (dendrites) to the keratinocytes of the skin and hair. All races have the same number of melanocytes - about 1 per 10 basal keratinocytes - the differences in skin colour being due to the amount of melanin formation in the melanosomes (maximal in dark races) and the way these are dispersed (singly in black people; in membrane-bound packages in white people). Apart from genetic factors, melanin production is stimulated by UV radiation and certain hormones, e.g. melanocyte-stimulating hormones (MSH), the chemically related ACTH, and female sex hormones. Hyperpigmentation is a common sequel to many inflammatory skin diseases, especially those in which the lower epidermis is damaged. There is both stimulation of melanogenesis and passage of melanin into dermal macrophages. Sometimes hyperpigmentation is due to substances other than melanin, e.g. haemosiderin (brownish-red) and drugs, e.g. amiodarone (grey). Many conditions associated with localized areas of hyperpigmentation are described elsewhere. These include neurofibromatosis (Fig. 10.38), xeroderma pigmentosum and Peutz-Jeghers disease. In practice the commonest are freckles, benign melanocytic naevi and postinflammatory hyperpigmentation. The causes of generalized or extensive hyperpigmentation are given in Table 10.25.

PIGMENTATION LOSS OF PIGMENT Normal skin colour is determined by haemoglobin, both oxygenated (red) and reduced (blue),carotenoids (yellow) and melanin (brown). Melanin is a polymer, synthesized from tyrosine in

Complete or partial loss of melanin pigment can have many causes. It is helpful to consider localized and generalized loss of pigment separately (Table 10.26).

427

TABLE 10.26 Localized and generalized pigment loss Condition

Causes

Clinical features

Localized hypopigmentation Vitiligo Loss of melanocytes

Pityriasis versicolor

Toxic effect on melanocytes

Well-defined white patches, often with hyperpigmentation of the bordering skin Personal or family history of organ-specific autoimmune disease common Slightly scaly macules mainly on the trunk, paler after sun exposure

Postinflammatory

(i)

(ii)

Faulty melanin transfer to keratinocytes Loss of melanocytes

Tuberous sclerosis 1 Poorly developed melanosomes Leprosy Decreased melanocytic activity Generalized pigment loss Albinism Faulty production of melanin in skin, hair and eyes Phenylketonuria Competitive inhibition of tyrosinase by phenylalanine ACTH and MSH Hypopituitarism deficiency

Pale patches with or after inflammation, e.g. eczema e.g. scarring after lupus erythematosus Leaf-shaped, pale macules and fibromas Pale patches are anaesthetic FIG. 10.54 Vitiligo shows sharply outlined areas of pigment loss Pale skin and hair, pink irises Pale skin with mental deficiency Associated endocrine deficiencies

Vitiligo Vitiligo (Fig. 10.54) is an acquired disorder in which melanocytes are lost from the basal layer of the epidermis. It occurs in about 0.4% of the population. A genetic basis is likely, as about 40% have a family history. In the individuals and their families there is an increased frequency of organ-specific autoimmune diseases, e.g. hypo- and hyperthyroidism. The precise pathogenesis is unknown. Clinical features Vitiligo has often begun by the age of 20. It is characterized by sharply defined areas of pigment loss, sometimes with mild hyperpigmentation of the adjacent normal skin. 0 The texture is normal and there is no scaling. In paleskinned individuals their first awareness may be sunburn in the depigmented skin. Any site can be affected and symmetrical involvement is common. Occasionally vitiligo

428

1

Figs 10.30, 10.31

3

MCQ10.16

0 Fig. 10.43

follows a dermatomal pattern. The face, axillae, groins, backs of hands, knees and elbows, and the genitalia are common sites. The condition is usually progressive, but repigmentation can occur spontaneously. 3 Diagnosis Diagnosis is usually straightforward. By Wood's light (p. 382) pityriasis versicolor (which is scaly) fluoresces yellow, and vitiligo is more strikingly white than the other causes of hypopigmentation (Table 10.27). The pale macules of leprosy are hypoanaesthetic. Treatment No treatment is entirely satisfactory. Cosmetic camouflage is helpful in a few. Sunscreens should be used on exposed areas to prevent burning and chronic sun damage, and in the pale-skinned this measure can improve the appearance by reducing tanning of the surrounding normal skin. Other measures used include psoralens and UVA (see PUVA treatment of psoriasis, p. 399) and the short-term application of potent topical corticosteroids.

PRESSURE SORES A pressure sore (bed sore, decubitus ulcer) is the consequence of sustained pressure to the tissues and is usually seen over a bony prominence, such as the ischial tuberosity, greater trochanter and elbow. The normal response to sustained pressure is movement to relieve that pressure. If

TABLE 10,27 Pathogenesis of pressure sores

TARLE 10,28 Causes of leg ulceration

Major factors Prolonged immobility e.g. Paraplegia Arthritis Operations and intensive care Plaster casts Apathy Loss of sensory stimuli e.g. Coma Multiple sclerosis Vascular disease e.g. Atherosclerosis

Cause

Example

Venous hypertension Arterial diseases Vasculitis Pyoderma gangrenosum Trauma Infections

Venous thrombosis Atherosclerosis, diabetes mellitus Rheumatoid disease, Wegener's granulomatosis See p. 407 Mechanical injury, burns, bites Ecthyma, tuberculosis, syphilis, deep fungal infection Leprosy, diabetes mellitus Disseminated intravascular coagulation, platelet disorders, sickle cell anaemia and spherocytosis, polycythaemia Cryoglobulinaemia Radiodermatitis Cutaneous carcinomas, Kaposi's sarcoma

Contributory factors Anaemia Hypoproteinaemia Severe weight loss

Neuropathy Blood disorders Plasma protein disorders Scarring disorders Neoplasia

discomfort is not felt or movement is not possible, then ischaemia is likely to occur. Friction (lateral pressure) exacerbates the tissue damage. Because fat and muscle are more susceptible, subcutaneous necrosis may be much greater than the surface changes suggest. Factors responsible for pressure sores are listed in Table 10.27. The earliest clinical sign is redness on an area subjected to pressure, the erythema not fading within 30 minutes. Action at this stage may prevent further damage. Progressive damage is manifested by blistering, ulceration and the formation of slough.

Management For any patient at risk, prevention is the chief aim. Frequent changes of position - with care not to exert friction by dragging the skin, and careful choice of the surface on which the patient is lying or sitting - are crucial. For the established pressure sore, these measures are equally important. In addition, anaemia and malnutrition should be corrected, slough should be removed and the sore kept scrupulously clean. In some cases excision of the entire necrotic area, with closure or grafting, should be considered.

10

VENOUS LEG ULCERATION Venous leg ulceration is also known as stasis or gravitational ulceration. The skin changes are due to the consequences of raised pressure in the venous system.

Aetiology and pathogenesis In most cases venous return is compromised by previous deep vein thrombosis. In a few cases ulceration is associated with varicosity of the superficial veins, or with defective communicating veins. Pressure in the venous system is raised because the valves are absent or destroyed, and this increased pressure is transmitted back to the capillary bed in the skin. There is extravascular deposition of fibrin, followed by fibrosis, and these changes are thought to impair transfer of oxygen and nutrients to the skin. It is also likely that venous distension produces reflex arteriolar constriction. This phenomenon occurs in normal individuals when standing still, and in them is promptly relieved by exercise, but this relief does not occur in those with persistently raised venous pressure. Patients who develop leg ulcers may also have decreased ability to lyse the fibrin, and this may contribute to the pathogenesis.

LEG ULCERATION

Clinical features Ulceration of the lower leg can have many causes, some of which are shown in Table 10.28. In the developed world most cases of leg ulceration are wholly or partly due to disorders of the venous system, but arterial insufficiency often coexists and may be the major factor in many patients. In the tropics, infective causes are much more important. Particularly in the elderly there can be multiple aetiological factors.

Venous ulceration is most common in middle-aged and elderly women. 'Interestingly, leg ulcers occur frequently and early in Klinefelter's syndrome (p. 68). Venous ulcers are usually preceded by varying combinations of oedema, prominent venules around the ankle, brown discoloration due to extravasation of red blood cells, eczema (p. 405), fibrotic thickening of the dermis and subcutaneous fat, and white plaques stippled with telang-

429

tional deficiency, anaemia, diabetes, obesity, hypertension, cardiac and renal disease, myxoedema, and any disease causing immobility. It is most important to assess the arterial system in the legs, as the compression that is valuable for venous insufficiency can be harmful if there is a poor arterial inflow. A simple Doppler device should be used, and if the ratio of inflow pressure in the ankle or foot to that in the arm is less than 0.8, compression should not be used without a vascular surgical assessment. Raised venous pressure can be counteracted by elevation of the legs at night and for periods during the day, and by careful use of compression bandages combined with exercise. If the ulcer is due to incompetent superficial or communicating veins rather than to deep vein thrombosis, surgical treatment may be indicated. Numerous treatments are available for the ulcer itself, and in most circumstances materials that do not contain potential sensitizers and maintain a moist surface should be used. Desloughing is often best achieved with a hydrocolloid dressing. Cellulitis requires systemic antibiotics. When venous ulceration has healed, the patient should continue to exercise and maintain compression with suitable stockings indefinitely. FIG. 10.55 Venous hypertension Ankle showing changes of venous hypertension: dilated veins, haemosiderin staining and atrophie blanche.

iectases (atrophie blanche) (Fig. 10.55). These signs occur most commonly near the medial malleolus, 1 but may be seen also on the lateral aspect of the ankle and over communicating veins. Another characteristic site is over the dorsum of the foot near the base of the toes. Ulceration may occur spontaneously or after trauma, which is often trivial. 2 Unlike ulceration due to arterial disease or vasculitis, venous ulcers are often painless. The base of the ulcer usually shows red granulation tissue and the edge is oedematous. The ulcer may be complicated by infection; cellulitis, usually due to Staph. aureus and/or Strep. pyogenes, can develop rapidly. Another common complication is contact allergic dermatitis due to medicaments applied to the ulcer, e.g. antibiotics and antiseptics (p. 400). In time the fibrotic process causes lymphoedema, and loss of movement at the ankle joint often occurs. Rarely, squamous and basal cell carcinoma can develop in chronic venous leg ulcers.

Management Patients with venous leg ulcers often have other disorders, correction of which facilitates healing. These include nutri-

1

430

Fig. 10.44

2

Case 10.3

3

Fig. 10.45

ARTERIAL ULCERATION When the arteriolar supply to a region of skin is insufficient or interrupted necrosis occurs, with resultant ulceration. The more profound process of gangrene is discussed on page 603. In general, vascular occlusion can occur because of changes outside the vessel wall, within the wall itself (as in atherosclerosis and vasculitis), and as a result of changes in the blood (as in cryoglobulinaemia and platelet thrombi). The symptoms and signs will be determined by the size of the ischaemic area and the speed with which the process occurs. The most common cause of arterial ulceration is atheromatous disease of the aorta and its tributaries.

Clinical features of ulceration due to atherosclerosis Arterial ulcers tend to be painful, sometimes severely so. There may be coldness, and pallor exaggerated by elevation of the limb, but cyanosis and blotchy erythema are also seen. Chronic ischaemia results in dryness and atrophy of the skin, loss of hair, and thickened, distorted nails. Intermittent claudication may be present (p. 602). Common sites for arterial ulceration due to atherosclerosis are the front or lateral aspect of the ankle, 3 and the toes. The ulcer tends to be well demarcated, with a grey sloughcovered base, which may expose deeper structures such as tendons. The management of arterial ulceration is usually the province of the vascular surgeon.

10

ACNE VULGARIS Acne vulgaris primarily involves the pilosebaceous follicles, i.e. the sebaceous glands, ducts, and the distal part of the hair follicles into which they open. The lesions include keratinous plugs in the ducts (comedones), inflammatory papules, pustules, nodules, cysts and scars.

Pathogenesis The best-defined factors determining the occurrence of acne vulgaris are androgenic stimulation of the sebaceous glands and the commensal anaerobic bacterium Propionibacterium acnes, which heavily colonizes active pilosebaceous follicles. Although the onset of acne around puberty is explained by the increased output of androgens at that time, it is much less clear why the disease becomes quiescent without any measurable change in the hormonal milieu or numbers of propionibacteria. The basis for the disorder of keratinization in the sebaceous ducts which produces the comedones is uncertain. Inflammatory lesions are mainly derived from closed comedones (see below). Possible mediators of inflammation include free fatty acids, bacterial cell wall components and enzymes, and the patient's complement system. The pus in acne lesions is sterile and a consequence of inflammation, which is often so severe that scarring results.

Clinical features Acne can occur in infants, but is usually mild and due to the influence of transplacental hormonal stimulus. In older children acne often represents the beginnings of puberty, but may not occur until the mid-teens or beyond. The areas affected are those with maximal numbers of pilosebaceous follicles: the face, upper trunk and shoulders. Comedones are either open (blackheads) or closed (small whitish papules). The inflammatory lesions are erythematous, varying from papules through pustules to nodules and large collections of pus (wrongly) called cysts (Fig. 10.56). The more destructive lesions heal with scarring, which is usually pitted but is hypertrophic or keloidal in those so predisposed. The course of acne can be erratic and there are often premenstrual exacerbations. Lesions resembling acne can be provoked by halogenated hydrocarbons, mineral oils, tars, greasy cosmetics, and drugs (p. 419).

Management Acne can have serious psychological effects and should not be ignored on the basis that it will get better sooner or later. Mild cases often respond to topical agents alone, e.g. benzoyl peroxide 2.5-10% gel, or topical retinoic acid. The latter is more irritant, but often more effective when comedones are the predominant feature. If there is insufficient

FIG. 10.56 Acne Note the papules, pustules, cysts and nodules.

response after a few weeks' treatment, or if the acne is of moderate severity, an oral antibiotic should be used in addition (or instead, if topical therapy proves too irritant). Tetracyclines and erythromycin are equally effective, usually in a dose of 0.5 g twice daily for several months. Topical antibiotics, e.g. 1 % clindamycin, can also be effective. For females with moderate acne, hormonal modulation with the antiandrogen cyproterone acetate may be more effective. This drug must be given with an oestrogen in the form of a low-dose oral contraceptive pill. Patients with severe and destructive acne usually need treatment with oral isotretinoin. This vitamin A analogue is highly effective but has a number of adverse effects, notably teratogenicity, and in the UK is restricted to hospital use. Other measures sometimes used include intralesional corticosteroids for inflammatory nodules and cysts, and dermabrasion, a surgical technique to improve scarring.

HYPERHIDROSIS Excessive sweating can be a feature of several systemic illnesses, or an isolated complaint.

Generalized hyperhidrosis Sweating from much of the body surface is a feature of many febrile illnesses and may be the presenting symptom (e.g. of tuberculosis and malaria). Other diseases

431

associated with a similar pattern of widespread sweating include lymphoma, hyperthyroidism, hypoglycaemia, phaeochromocytoma, the menopause, diabetic autonomic neuropathy and acromegaly. Drugs that may cause sweating include alcohol (commonly) and fluoxetine (rarely).

Palmoplantar and axillary hyperhidrosis Emotional stress and mental activity typically produce excessive sweating to varying degrees from the palms, soles and armpits. Other areas that may be involved include the face and groin. This pattern of hyperhidrosis is common, not associated with underlying disease, but can cause significant disability.

Asymmetrical hyperhidrosis This can be caused by a variety of neurological lesions, from the brain downwards. A particular example is sweating on one side of the face associated with eating (gustatory hyperhidrosis). This often occurs after an injury that results in loss of the normal nerve supply to the parotid gland, the regenerating fibres instead supplying the sweat glands.

Treatment

and sinus tracts. The areas affected are the axillae, anogenital area and breasts. Tender inflammatory nodules, which discharge pus, are seen and blackheads are often present early on. Because bacteria play a perpetuating role in this disorder, courses of antibiotics (if possible governed by sensitivity testing) are helpful but not curative. Incision, marsupialization and even wide excision may be needed in chronic cases. Hormonal modulation, e.g. with antiandrogens in females, may have a role in treatment, and isotretinoin may help in some cases.

ROSACEA Rosacea is a common condition of the middle-aged and elderly, in which there are varying combinations of redness of the face with telangiectasia, papules and sterile pustules (Fig. 10.57). 1The facial erythema is often easily exacerbated by spicy foods, alcohol and emotional upset; chronic UV exposure may have an aetiological role. Unlike acne there are no comedones and no elevation of sebum excretion rate. Common associations are eye disorders (e.g. conjunctivitis and keratitis), facial lymphoedema and rhinophyma, a hypertrophy of the nasal skin and its appendages. Despite the cause remaining unknown, treatment with oral oxytetracycline (erythromycin is an

Axillary hyperhidrosis can sometimes be helped by 20% aluminium chloride in alcohol applied at bedtime. Palmar and plantar hyperhidrosis are usually better treated with iontophoresis, either using water alone or with the anticholinergic drug glycopyrronium bromide in solution. This treatment utilizes DC electricity to drive ions into the skin. Psychological methods and even psychotropic drugs are useful in some patients in whom anxiety is a prominent trigger. Transthoracic endoscopic sympathectomy can be very effective in selected patients with palmar hyperhidrosis. Recently, success has been claimed for localized injections of botulinum toxin.

HIDRADENITIS SUPPURATIVA Hidradenitis suppurativa is a chronic suppurative, inflammatory and scarring condition of blocked apocrine glands. There may be a genetic basis for this disorder, which has some similarity to acne vulgaris and may be associated with it. Following keratinous plugging of the apocrine glands there is bacterial infection, subsequent gland rupture, damage to surrounding tissues, and healing with scarring

1

432

Fig. 10.46

FIG. 10.57 Rosacea Rosacea is characterized by erythema, papules and pustules but there are no comedones. 1

TABLE 10.29 Systemic causes of generalized pruritus

TABLE 10.30 Causes of prurifus ani

Hepatic disease Obstructive biliary disease Pregnancy: oestrogen-induced cholestasis

Anorectal disease Fissure Haemorrhoids Rectal carcinoma Faecal soiling Skin disease Psoriasis Atopic dermatitis Irritant and allergic contact dermatitis (especially medicaments)

Endocrine disease Hyper- and hypothyroidism Diabetes mellitus (rare) Renal Chronic renal failure Parasitic Trichinosis Onchocerclasis Malignancy Hodgkin's disease Leukaemia and lymphoma Polycythaemia rubra vera Other neoplasms

Haematological Iron deficiency Polycythaemia rubra vera Drugs Opiates Subclinical drug sensitivity Psychiatric Neurosis Psychosis, e.g. delusion of parasitosis

alternative) is effectively suppressive in most cases. Topical metronidazole is also helpful. It is usually helpful to minimize the effects of sun exposure. Rosacea is slowly worsened by the use of topical steroids on the face, and a similar condition, perioral dermatitis, can be initiated by the more potent topical steroid preparations.

PRURITUS Pruritus, or itching, is the distinctive sensation whose outward manifestation is the act of scratching. Except for the role of histamine in some disorders, and bile salts in obstructive liver disease, we know disappointingly little about the mechanisms of this common symptom. When severe, pruritus can surpass pain in the distress it can cause. Many skin diseases can be itchy, and some, such as scabies, dermatitis herpetiformis, lichen planus and urticaria, severely so. Dry skin of whatever cause tends to be itchy, and this is a particularly common problem in the elderly and one of the reasons for pruritus in the atopic individual. Occasionally an external cause (such as fibreglass or scabies) can produce widespread pruritus with little to see. Itching can occur without any primary rash, the skin changes, if any, being those caused by rubbing and scratching. Generalized pruritus may be a pointer to systemic disease (Table 10.29). Wherever possible, the underlying cause of generalized pruritus should be treated. When symptomatic measures are called for the use of emollients can be helpful, probably because dry skin has a lower threshold for pruritus. Oral HI antagonists can provide some relief, particularly when sedative antihistamines are used. Uraemic pruritus

10 Infection Candidiasis Infestation Threadworms Psychogenic Anxiety Depression Unknown

TABLE 10.31 Causes of pruritus vulvae Diseases special to vulvaf skin Lichen sclerosus et atrophicus Leukoplakia Carcinoma Skin disease Psoriasis Atopic dermatitis Irritant and allergic contact dermatitis (especially medicaments)

Infection Candidiasis Trichomonas Infestation Pediculosis Psychogenic Anxiety Depression Unknown

can be helped by UVB. Pruritus due to cholestatic liver disease often responds to cholestyramine or colestipol. The itch in polycythaemia rubra vera may respond to antiserotonin agents, cyproheptadine or pizotifen. Localized pruritus without primary signs of skin disease is common in the perianal (Table 10.30) and vulval areas (Table 10.31). Sometimes there is a psychological basis for this.

ALOPECIA The evaluation of alopecia, or hair loss, takes account of the distribution, any abnormalities of the hair shafts, and the scalp.

Androgenic alopecia (common baldness) The prerequisites for this very common condition are an inherited tendency; some degree of ageing; and postpubertal androgen levels. The longer, coarser, pigmented terminal hairs on the scalp are progressively replaced by small, fine vellus hairs. In males there is recession of the anterior hairline and thinning over the crown, and ulti-

433

mately terminal hair may only be present at the back and sides of the scalp. In an endocrinologically normal female there is thinning over the top of the scalp, but baldness to the extent seen in males is uncommon, except in the elderly, and if present warrants careful evaluation. Treatment In a few, with a short history of hair loss, topical minoxidil can induce modest regrowth of hair, but even with sustained application the hair usually falls out eventually, and the treatment is expensive. Some resort to hair transplantation.

Diffuse alopecia Diffuse alopecia is a decrease in hair density over the whole scalp without changes in the skin or hair morphology. In some instances, e.g. a febrile illness, childbirth and severe emotional upset, the causative factor synchronizes a proportion of the follicles so that hairs are shed 2-3 months later, a phenomenon called telogen effluvium. Regrowth usually takes place in about 6 months. Normal hair growth requires a adequate nutrition, and diffuse alopecia accompanies malnutrition and wasting conditions. Iron deficiency, even in the absence of anaemia, has been associated with hair loss. Dry and progressively sparse hair is a feature of hypothyroidism. Many drugs (p. 419) can produce diffuse alopecia, sometimes (as with cytotoxic agents) within days of administration, or after several weeks (as with anticoagulants, antithyroid drugs and the retinoids). Stopping the oral contraceptive has been associated with mild hair loss, similar to the effect of childbirth.

Alopecia areata Alopecia areata (Fig. 10.58) is a common condition in which there is loss of hair in patches, typically with no skin changes. Hair on the face and body as well as the scalp may be affected. A distinctive feature is the presence of shortened hairs which taper markedly where they emerge from the scalp (exclamation-mark hairs). White hairs tend to be spared the process causing the loss. Alopecia areata is usually reversible, and when hair grows back it is often initially white. When the entire scalp becomes affected the condition is called alopecia totalis, and if all the body hair is lost as well the term alopecia universalis is applied. Regrowth usually occurs, but is less likely with these more severe forms. Nails may show a distinctive pattern of pitting and ridging. 1 There is sometimes an association with one or more organ-specific autoimmune diseases in the patient or family, such as vitiligo. The cause of alopecia areata is not known. Topical and intradermal corticosteroids can be helpful in selected cases. Treatment by induced contact allergic dermatitis, e.g. to diphencyprone, can be useful.

Inflammatory alopecia Fungal (p. 389) and bacterial infection can cause patchy hair loss with varying degrees of erythema, oedema and pustules. Lichen planus (p. 413) and lupus erythematosus (p. 1172) may affect the scalp, and characteristically produce atrophy. Occasionally metastatic neoplasia presents as inflammatory alopecia.

Traumatic alopecia Hair can be damaged by chemicals, e.g. those used in permanent waving, and physical trauma such as twisting between the fingers and pulling.

HIRSUTISM AND HYPERTRICHOSIS

FIG. 10.58 Alopecia areata A distinctive feature of alopecia areata is the presence of shortened 'exclamation mark' hairs which taper markedly where they emerge from the scalp.

Hair growth can be excessive by being coarser, longer and/or more profuse than is normal for age, sex and race at any particular body site. The term hirsutism is used to describe excessive hair growth in females on those areas where secondary sexual hair occurs in males, i.e. the upper lip, beard area, chest, around the nipples, upper back, lower abdomen etc. Hypertrichosis designates other patterns of excessive hair growth. Both hirsutism and hypertrichosis can have a familial basis and are commoner in certain races.

Hirsutism 1

434

Fig. 10.47

2

Fig. 10.48

When hirsutism occurs together with other symptoms and signs suggestive of an underlying endocrine cause, such as menstrual irregularities, infertility and virilization (clitoromegaly, deepening of the voice and male pattern baldness),

10

TABLE 10.32 Endocrine causes of hirsutism Ovarian Adrenal Pituitary Exogenous

Polycystic ovaries Virilizing tumours Congenital adrenal hyperplasia Adenoma and carcinoma Cushing's syndrome Acromegaly Corticosteroids, androgens, progestogens

appropriate investigations should be carried out. Table 10.32 lists some endocrine causes. In most cases hirsutism is an isolated finding, probably caused by an increased sensitivity of hair follicles to normal levels of circulating androgens, and in the absence of other symptoms investigation is unnecessary.

Hypertrichosis There is no known underlying cause for hypertrichosis. It can occur as an isolated event, e.g. over the lower spinal column in spina bifida occulta, in association with melanocytic naevi, and in areas of chronic inflammation; or as a more generalized phenomenon, e.g. in malnutrition, some porphyrias, and with certain drugs, notably ciclosporin and minoxidil.

Treatment When there is an underlying endocrine cause of hirsutism, treatment of this may help. For idiopathic hirsutism, treatment should be tailored to the needs of the individual. For mild cases counselling alone may be sufficient. Excessive hair can be removed by depilatory creams, waxing or shaving. Follicles can be destroyed by electrolysis or certain types of laser therapy. For some patients the antiandrogen cyproterone acetate (given together with ethinyl oestradiol to maintain a normal menstrual cycle) can gradually reverse hirsutism, but the effect only lasts as long as the treatment.

NAILS AND DISEASE The nail is a specialized sheet of keratin largely produced by an invagination of the epidermis called the matrix, which is located beneath the posterior nailfold and extends beneath the nailplate as the lunule (halfmoon). Disorders of the nails may result from local causes such as psoriasis, fungal infection and lesions beneath the nail (e.g. melanoma), or reflect a systemic disorder (see below).

Clubbing Normally the angle between the nail and the posterior nail-

FIG. 10.59 Koilonychia The nail plates are thinned and spoon-shaped depressions.

fold is less than 180°. In clubbing this angle increases, often becoming more than 180°. The causes are discussed on page 619.

Koilonychia Koilonychia is flattening or even depression of the normal slight convexity of the nail, and is usually due to iron deficiency (Fig. 10.59).

Beau's lines Beau's lines are transverse grooves in all the nails, reflecting a transient reduction in nail growth due to a severe illness. They may also occur after childbirth.

Yellow nail syndrome In yellow nail syndrome the nails become yellow, thickened and excessively curved, and virtually stop growing. 2 Cuticular attachment is lost and there is usually some separation of nail from nailbed. The nail changes are associated with lymphoedema, bronchiectasis and pleural effusions. It is more common to find yellow nails in circumstances such as psoriasis and dermatophyte fungal infection, these conditions being recognized by the associated skin changes and mycological investigations. Splinter haemorrhages Splinter haemorrhages are usually longitudinal or dotshaped areas of haemorrhage beneath the nails (Fig. 10.60). Most are the result of minor trauma, are painless, and appear beneath the distal third of the nail. Psoriasis and fungal infection of the nails may be associated with similar distal splinters. Painful and proximal splinter haemorrhages are much more likely to be associated with medical causes, of which the best established are subacute bacterial endocarditis, trichinosis, chronic mountain sickness and indwelling radial arterial catheters.

435

Dermatitis artefacta Dermatitis artefacta is mainly seen in females. The lesions are self-inflicted but this is denied, and they may indeed be produced subconsciously. They are often bizarre in appearance, resistant to therapy (except occlusive dressings, in which case lesions may appear at new sites), and often regarded with indifference. The patients tend to be immature or insecure, but the motivation behind dermatitis artefacta is often difficult to elucidate and the condition can be protracted. Self-inflicted damage is also seen in psychotics and the Lesch-Nyhan syndrome, a rare X-linked inherited disorder of purine metabolism associated with mental retardation and spasticity.

Delusions of infestation

FIG. 10.60 Splinter haemorrhages Elongated deposits of blood beneath the nail plate.

Paronychia Paronychia is a painful swelling of the posterior nailfold. When acute, staphylococcal infection is likely. A more gradual onset and chronic course is usually associated with Candida albicans. 1 Women are especially prone to chronic paronychia. Common predisposing factors are occupations in which the hands are frequently wet, diabetes mellitus and poor peripheral circulation.

This is the conviction that there are parasites on or in the skin, often supported by offerings of excoriated skin fragments. Such a delusion may be one element in a broader presentation of psychosis, including that due to organic disorders such as vitamin B12 deficiency, but can present in an isolated form (monosymptomatic delusional hypochondriasis). It is of course essential to exclude infestation such as scabies and pests in the home. The monodelusional state often responds well to the psychotropic drug pimozide.

Dermatological non-disease The term 'dermatological non-disease' describes a disproportionate anxiety about minor and often physiological skin changes. Such patients are difficult to manage, but can be seriously disturbed and at risk of suicide.

PSYCHOLOGICAL CAUSES OF SKIN

DISEASE

GENODERMATOSES

Many bona fide skin diseases, such as eczema and psoriasis, may worsen or be precipitated during periods of psychological stress. Similarly, anxiety and depression commonly accompany distressing skin disease, and can often be ameliorated by explanation, reassurance and encouragement. There is a direct psychological cause with dermatitis artefacta, many cases of delusions of infestation, and psychogenic pruritus. A diagnosis of psychogenic pruritus rests on exclusion of an organic cause as well as association with other psychoneurotic symptoms and signs.

1

436

Fig. 10.49

The skin is affected in a number of genetically determined syndromes, some of which are covered in this chapter or, because they most frequently present with the systemic manifestations, elsewhere in this book. However, there are some important disorders that often present with cutaneous features that should be discussed briefly here. Darier's disease is an autosomal dominant condition in which clusters of warty papules are seen on the face, neck and trunk, in association with abnormal nails (classically linear streaks and 'v'-notching) and palmar pits. The term palmoplantar keratoderma is used to describe diffuse or patchy thickening of the palms and soles. There are many variants, some of which are simply a nuisance and affect only the hands and feet. In some, however, there may be associated abnormalities, e.g. mental retardation and carcinoma of the oesophagus. Ehlers-Danlos syndrome is a name given to a group of disorders in which there are

10

TABLE 10.33 Syndromes associated with cancers Syndrome

Features

Cowden's syndrome Gardner's syndrome Rothmund-Thomson syndrome Xeroderma pigmentosum Xeroderma pigmentosum variant Bloom's syndrome Howel-Evans syndrome Muir-Torre syndrome Peutz-Jeghers syndrome

Multiple trichilemmomas; fibromas of the lips; acral keratoses; carcinoma of breast, pancreas, thyroid Multiple epidermoid cysts; intestinal polyposis Poikiloderma; photosensitivity; short stature; skin and soft tissue cancers (e.g. osteosarcoma) Early-onset photosensitivity; excessive freckling; multiple skin cancers; some variants also develop internal cancers and CNS defects Similar, but less severe changes; later onset Growth retardation; pigmentary anomalies; leukaemia Palmoplantar keratoderma with carcinoma of the oesophagus Sebaceous gland carcinomas; colonic carcinomas Intestinal polyps occasionally undergo malignant transformation

FIG. 10.61 Hereditary haemorrhagic telangiecstasia Papular telangiectases on the forehead. Unlike many spider naevi, these lesions do not fill from the centre when emptied by pressure.

abnormalities of connective tissue (see also p. 71). Some forms are mild, consisting largely of hyperextensible joints and rather lax, friable skin. Others, however, are associated with major vascular problems, including arterial rupture. Pseudoxanthoma elasticum (see also p. 71) is a condition in which the skin of the neck, the elbow folds and occasionally other sites, gradually becomes more and more papular. The final appearance is said to look like the skin of a plucked chicken. The condition may be entirely benign, but in some pedigrees the connective tissue defects lead to 'angioid' streaks in the retina, blood vessel rupture and mitral valve prolapse. Hereditary haemorrhagic telangiectasia presents with multiple telangiectases on the face (Fig. 10.61), especially the lips. There are often lesions in the respiratory and gastrointestinal tracts as well, which may bleed. The skin is affected in a number of inherited immunodeficiency syndromes, either with multiple infections (e.g. in agammaglobulinaemia and severe combined immune deficiency) or with more direct cutaneous changes (e.g. eczema in Wiskott-Aldrich syndrome or the telangiectases of ataxia telangiectatica). There are also a number of syndromes associated with cancers which have significant cutaneous features. Some of these are listed in Table 10.33.

437

This page intentionally left blank

11

the western world male homosexuals are more likely to contract hepatitis B, syphilis and HIV infection than are their heterosexual counterparts. However, the incidence of newly acquired syphilis and gonorrhoea in homosexual men fell as trends in sexual behaviour changed in the 1980s. The basis of any STD service is the control of sexually transmitted infections. There are basic requirements:

Genitpironary

Disease I G Williams

Sexually transmitted diseases 439

Acquired immune deficiency syndrome (AIDS) 439

HIV neurological disease 448 Sexually transmitted infections 456

SEXUALLY TRANSMITTED DISEASES Sexually transmitted diseases (STDs) are caused by infections transferred through sexual contact. Our appreciation of the range of agents capable of being transmitted sexually has increased.

Epidemiology The prevalence and incidence of sexually transmitted infections vary in different parts of the world. This is due to variations both in the diseases found and in the methods of reporting such infections. The type of sexual behaviour and the sexual orientation of the person involved also influence the transmission of the infectious agents. This is clearly demonstrated by the fall in incidence of syphilis and gonorrhoea after the introduction of penicillin (Figs 11.la, 11.1b), the rise in incidence in the 1960s due to an increase in the number of people's sexual partners, followed by a substantial fall in the 1980s, when the AIDS epidemic resulted in a change in sexual behaviour in homosexual men. Figure 11.2 shows the most commonly found STDs treated at genitourinary medicine clinics. This is an underestimate of the problem. Morbidity associated with STD is more common in women and the agents responsible are more difficult to identify. Reported STDs are observed mostly in the 20-24-yearold age group, followed by 25-29 and 15-19-year-olds. In

• The patient needs ready access to a specialist centre for accurate diagnosis and treatment of any infection. Attendance needs to be voluntary. Treating the index case only partly helps in controlling an infection, and therefore sexual partners need to be identified and treated. • Patients are much more likely to cooperate in these matters if they are assured of absolute confidentiality. The STD service is the most confidential are in the health service, and details are generally only released with the patient's consent. • Education about STDs, both for an individual and as part of general health education, is important in helping to control the spread of infection. • The clinics keep a record of the number of patients seen with each infection, and these figures form part of a detailed national epidemiology report.

ACQUIRED IMMUNE DEFICIENCY SYNDROME (AIDS) Definition The first recognized cases of the acquired immune deficiency syndrome (AIDS) occurred in the summer of 1981 in America. Reports began to appear of Pneumocystis carinii pneumonia and Kaposi's sarcoma in young men, who, it was subsequently realized, were both homosexual and immunocompromised. Even though the condition became known early on as AIDS, its cause and modes of transmission were not immediately obvious. The virus now known to cause AIDS, the human immunodeficiency virus (HIV), was discovered in 1983. The definition of AIDS has changed over the years as a result of increasing appreciation of the wide spectrum of clinical manifestations of infection with HIV. Currently AIDS is defined as an illness characterized by one or more indicator diseases (Table 11.1). In the absence of another cause of immune deficiency, and without laboratory evidence of HIV infection (if the patient has not been tested or the results are inconclusive), certain diseases, when definitely diagnosed, are indicative of AIDS. Regardless of the presence of other causes of immune deficiency, if there is laboratory evidence of HIV infection, other indicator diseases that require a definitive (or in some cases only a presumptive) diagnosis also constitute a diagnosis of AIDS.

439

TABLE 11.1 AIDS indicator diseases

FIG. 11.1 Incidence in England and Wales of A gonorrhoea 1918-1990, B infectious and early latent syphilis 1931-1998

Recurrent/multiple bacterial infections - child aged under 13 Candidiasis - pulmonary *Candidiasis - oesophageal Cervical carcinoma - invasive Coccidioidomycosis - disseminated Cryptococcosis - extrapulmonary Cryptosporidiosis - with diarrhoea persisting for more than one month *Cytomegaiovirus retinitis Cytomegalovirus disease - not in liver, spleen or nodes HIV encephalopathy Herpes simplex virus (HSV) infection - mucocutaneous ulceration lasting for more than one month or pulmonary, oesophageal infection Histoplasmosis - disseminated Isosporiasis - with diarrhoea persisting for more than one month *Kaposi's sarcoma tymphoid interstitial pneumonia - child aged under 13 Non-Hodgkin's lymphoma - Burkitt's or immunoblastic Primary cerebral lymphoma *Disseminated mycobacteriosis, e.g. Mycobacterium avium *Mycobacterial tuberculosis - extrapulmonary, pulmonary * Pneumocystis carinii pneumonia *Recurrent pneumonia within a 12-month period Progressive multifocal leucoencephalopathy Salmonella septicaemia - recurrent *Cerebral toxoplasmosis Wasting syndrome due to HIV * These diseases may be diagnosed presumptively if laboratory evidence of HIV exists.

Disease Genital warts Non-specific genital infection Chlamydia infection Genital herpes Gonorrhoea Trichomonas Syphilis FIG. 11.2 Number of cases of sexually transmitted disease seen in STD clinics in England and Wales, 1998 (excluding HIV)

440

In 1993 the Centers for Disease Control in the USA extended the definition of AIDS to include all persons who are severely immunosuppressed (a CD4 count <200 x 106/L), irrespective of the presence or absence of an indicator disease (Table 11.2). For surveillance purposes this definition has not been accepted within the UK and Europe, where AIDS continues to be a clinical diagnosis defined by one or more of the indicator diseases.

Epidemiology Transmission

HIV has been isolated from semen, cervical secretions, lymphocytes, cell-free plasma, cerebrospinal fluid, tears, saliva, urine and breast milk. This does not mean, however, that these fluids all transmit infection, as the concentration of virus in them varies considerably. Semen, blood, and possibly cervical secretions are particularly infectious. The commonest mode of transmission of the virus throughout the world is by sexual intercourse: whether this is anal or vaginal is unimportant. Other methods of transmission are through the receipt of infected blood or blood products and donated organs and semen. Transmission also occurs through the sharing or reuse of contaminated needles by injecting drug users or for therapeutic procedures, and from mother to child. Transmission from mother to child occurs in utero, at birth (intrapartum) and through breast milk. There is no well documented evidence that the virus is spread by saliva. It is not spread by casual or social contact. Health-care workers can, however, be infected through

TABLE 11.2 Centers for Disease Control (CDC) revised classification system for HIV infection (1993)

New HIV infections in 1999

Clinical categories

A Asymptomatic, acute (primary) HIV or PGL

B Symptomatic, not (A) or (C) conditions

C AIDSindicator conditions

1. >500x10 /L

A1

B1

2. 200-499 x10 6 /L

A2

B2

3. <200x10 6 /L

A3

B3

C1 C2 C3

CD4+Tcell categories 6

TABLE 11.3 Global summary of the HIV/AIDS epidemic during December 1999 (WHO/UNAIDS)

PGL = persistent generalized lymphadenopathy.

Adults (women Children <15 years TOTAL

People living with HIV/AIDS

Adults (Women Children <15 years TOTAL

11

5 million 2.3 million) 570000 5.6 million 32.4 million 14.8 million) 1.2 million 33.6 million

Deaths due to HIV/AIDS, 1999

2.6 million

Cumulative number of deaths due to AIDS/HIV

16.3 million

needle-stick injuries and by skin and mucosal exposure to infected blood or body fluids. Finally, there is no evidence that the virus is spread by mosquitoes, lice, bedbugs, in swimming pools, or by sharing cups, eating and cooking utensils, toilets or air space with an infected individual. In other words, HIV and AIDS are not contagious.

Size of the problem Worldwide The WHO and UNAIDS estimate that by the end of 1999, 32.4 million adults and 1.2 million children worldwide will be living with HIV, and that cumulatively since the start of the epidemic, 16.3 million people will have died from HIV/AIDS (Table 11.3). The largest burden of infection has fallen on the developing world, particularly subSaharan Africa, which has close to 70% of the global total for HIV-infected people but is home to just 10% of the world (Fig. 11.3). In adults aged between 15 and 49 years it is estimated that the prevalence of those living with HIV/AIDS in 1999 was 8% in sub-Saharan Africa, compared to 0.56% in North America and 0.25% in western Europe. Recent studies have shown that although in the developing world HIV is transmitted predominantly by heterosexual intercourse, the female-to-male ratio of infected persons in sub-Saharan Africa is 1.2:1. There are several factors that might account for this, including the greater efficiency of male-to-female HIV transmission through sex, and the younger age at first sexual contact for women. Elsewhere in the world, estimated new infections continue to increase markedly in south and southeast Asia, including India, although preventative programmes in Thailand are beginning to be effective. Eastern Europe and central Asia, constituting Russia and the other former states of the USSR, also saw a marked increase in the estimated reported number of HIV-infected people, rising by one-third, and largely driven by transmission through injecting drug use. United Kingdom and United States By December 1999, 724636 cases of AIDS in adults had

FIG. 11.3 Adults and children estimated to be living with HIV/AIDS as of December 1999 (WHO/UNAIDS)

been reported in the United States, and in the United Kingdom the total was 16457. These figures are thought to be reasonably accurate and complete. Both countries have seen a significant decrease in the number of new cases of AIDS reported, a result of clinically effective antiretroviral therapy; however, new infections with HIV continue to occur and the prevalent population of people living with AIDS/HIV is rising. In North America and the UK the first wave of the epidemic occurred in homosexual men. Although new infections among men who have sex with men still arise, this group now accounts for less than 50% of those newly diagnosed with HIV infection per year in both countries. In the USA an increasing proportion are among injecting drug users who share needles or equipment, and in the UK among heterosexuals (Fig. 11.4). The latter is driven largely by an imported epidemic from subSaharan Africa, with only small numbers having acquired the infection through heterosexual contact within the UK. In the USA, 16% of AIDS cases have occurred in women, although the commonest route of transmission among such women is injecting drug use (42%). The next most common is heterosexual contact (40%).

441

FIG. 11.4 A changing epidemic: cases of HIV infection by exposure category diagnosed in the UK 1989-1999

THE HUMAN IMMUNODEFICIENCY VIRUS The human immunodeficiency virus type 1 (HIV 1) was first isolated in 1983 at the Pasteur Institute, Paris. It has a cylindrical core surrounded by a bilayered lipid envelope acquired as the new virion buds from its host cell. Implanted into this envelope are viral glycoproteins (Gp41 and 120 - Fig. 11.5) carrying receptor sites for the CD4 antigen expressed on the surface of host cells. Only cells bearing this antigen are susceptible to infection. The inner core of the virus is made up of the capsid protein P24 encasing two strands of viral RNA and viral enzyme molecules. The nucleic acid of HIV 1 has been cloned and sequenced. It has a basic gene structure common to retroviruses but is very different from the other human retroviruses, human T-lymphotrophic viruses 1 and 2. A second human immunodeficiency virus (HIV 2) was isolated in 1986 and has similar biological properties to HIV 1, but is limited to west Africa or countries with links to that area. On entry to the infected cell the viral reverse transcriptase enzyme (hence retrovirus) makes a DNA copy (proviral DNA) of the RNA genome (Fig. 11.6). The pro viral DNA is able to integrate into the host cell DNA. Latent or non-productive or productive viral replication may occur. During productive replication RNA transcripts are made from the proviral DNA, and complete virus particles

1

442

MCQ 11.1

FIG. 11.5 Schematic representation of the Human Immunodeficiency Virus

are assembled and released from infected cells by characteristic budding. The virus has three main genes: • 'Pol', which encodes for the three viral enzymes: reverse transcriptase, integrase and protease; • 'env' which encodes for the envelope glycoproteins; • 'gag', which encodes for core proteins. In addition, there are accessory genes which encode for regulatory proteins that control HIV replication within the target cell. 1

IMMUNOLOGY OF AIDS In the weeks following infection there is a sharp increase in plasma viral load, widespread dissemination of the virus, and occasionally clinically an acute seroconversion illness. An immune response to HIV develops, followed by a fall in detectable plasma viraemia. Although a period of prolonged clinical latency then ensues, viral replication continues within lymphoid tissue, together with a gradual fall in the number of circulating CD4 T cells in the peripheral blood. Destruction of lymphocytes bearing the T4 or CD4 differentiation antigen and loss of specific T-helper cell function seems to be the primary abnormality of immune dysfunction. The CD4 molecule, however, is also displayed

TABLE 11.4 Immunological abnormalities In HIV infection

11

Lymphopenia: CD4 T lymphocyte (naive and memory) cytotoxic responses: cell-mediated T cells (CD8) natural killer cells T-cell antigen repertoire 4 proliferative responses (mitogens, antigens) decrease skin allergy to common recall antigens e.g. Candida, PPD, tetanus immunoglobulins: polycolonal B-cell activation decreased de novo antibody response cytokine e.g. TNF-a, interleukins

FIG. 11.6 HIV replication cycle The main targets for antiretroviral treatments are inhibition of the viral enzymes: 1. Fusion of virus with CD4 receptor molecule; 2. Reverse transcriptase, which facilitates the transcription of viral RNA to a DNA copy; 3. Protease, which is necessary for cleavage of pre-cursor viral structure proteins after budding of HIV virion.

at lower density on other cells, such as monocytes, macrophages and some B lymphocytes, some of which may act as reservoirs of HIV infection. The CD4 lymphocyte has a pivotal role in the immune response (interacting with macrophages, other T cells, B cells and natural killer cells, either by direct contact or via the influence of lymphokines such as y-interferon and interleukin 2, see Ch. 4). The other immunological abnormalities seem to be largely secondary to the disorder of CD4 lymphocytes (Table 11.4). In vitro, HIV produces a cytopathic effect in susceptible cell lines, multinucleate giant cells (syncytia) form and cell death occurs. The mechanism(s) by which it produces immune dysfunction in vivo are not yet fully understood, but may include a direct cytopathic effect of HIV on the CD4 lymphocyte, HIV-specific cellular immune responses directed towards infected cells, and the triggering of programmed cell death (apoptosis) by viral proteins. Different factors, both viral and host determined, are likely to influence the rate of disease progression. A small number of HIV-infected persons, however, show little evidence of immunosuppression many years after acute infection (long-term non-progressors). It is thought that the breadth of the immune response to the different HIV anti-

gens at the time of primary infection is pivotal in determining disease progression. In most infected persons there is a rapid loss of specific T-helper cell and CTL responses to multiple HIV antigens during primary infection, which do not recover during chronic established infection. In persons where these specific immune responses are preserved, plasma HIV RNA levels are low, viral replication is contained and clinical disease progression slow. A better understanding of these specific protective immune responses may lead to a more complete knowledge of the immunopathogenesis of AIDS and the possible development of a protective vaccine.

NATURAL HISTORY Acute infection with HIV may be accompanied by a transient non-specific illness similar to glandular fever; it includes fever, malaise, myalgia, lymphadenopathy, pharyngitis and a rash. A transient aseptic meningoencephalitis may also occur. Most acute infections, however, are subclinical. The acute infection is accompanied by the development of antibodies to the core and surface proteins, usually in 2-6 weeks, although delayed seroconversions have been observed. Antibodies are usually detected by enzyme-linked immunoassays, and their presence can be confirmed by immunofluorescence or Western blot. Molecular techniques (PCR) have been developed to measure quantitative levels of plasma viral RNA and cellular viral DNA. High levels of plasma viraemia are detectable at times of seroconversion and before the appearance of antibodies. Diagnosis of acute primary infection can be confirmed by a positive PCR test for proviral DNA (or viral RNA using assays of high specificity) together with either a negative antibody test or the presence of an evolving antibody response. Following acute infection plasma levels rapidly decline and a chronic infection ensues (Fig. 11.7) which is asymptomatic in the early stages. Physical examination may show no abnormality, but about one-third of patients have persistent generalized lymphadenopathy (PGL). This is defined as lymph nodes of 1 cm or more in diameter in

443

FIG. 11.8 Oral candidiasis in the patient with symptomatic HIV infection The figure shows scattered white plaques with underlying erythema on the soft palate. FIG. 11.7 Progressive depletion of CD4 T cells in AIDS Plasma viraemia in peripheral blood, CD4 levels following primary infection with HIV. Once the CD4 count falls below a certain level the risk of major opportunistic infections increases substantially (adapted from Pantaleo G, Graziosi C, Fauci AS. New Eng J Med 1993 328: 327-334).

two or more non-contiguous extrainguinal sites, which cannot be explained by any other infection or condition. The commonest sites of lymphadenopathy are the cervical and axillary lymph nodes; it is unusual in the hilar lymph nodes. Biopsy usually shows a benign profuse follicular hyperplasia. At some time after the initial infection non-specific constitutional symptoms develop. These may be intermittent or persistent, and include fevers, night sweats, diarrhoea and weight loss. Patients may also be affected by several 'minor' opportunistic infections or conditions that tend to affect the mucous membranes and skin, such as oral candidiasis (Fig. 11.8), oral hairy leukoplakia (Fig. 11.9), herpes zoster, recurrent oral or anogenital herpes simplex, and other skin conditions such as seborrhoeic dermatitis, folliculitis, impetigo and tinea infections. This collection of symptoms and signs, which are often a prodrome to the development of major opportunistic infections or tumours, is referred to as symptomatic non-AIDS (Table 11.5). Without treatment with antiretroviral therapy, about 75% of HIV-infected people developed symptomatic disease (either AIDS or non-AIDS) and 45-50% develop AIDS in 9-10 years from primary infection. Retrospective testing of stored blood samples from early prospective

FIG. 11.9 Oral hairy leukoplakia in a patient with symptomatic HIV infection The figure shows a corrugated white patch on the lateral border of the tongue.

cohort studies have shown that levels of plasma viral load, combined with the absolute CD4 count, are highly predictive of the risk of clinical progression (Fig. 11.10). Infected persons with low plasma viral load levels following primary infection, presumably reflecting a better-preserved specific immune response to HIV, have a much lower risk of clinical progression than those with substantially higher plasma viral load. 12

OPPORTUNISTIC INFECTIONS

1

444

Case 11.1

2

MCQ 11.2

The incidence of opportunistic infections in populations of HIV-infected persons has fallen significantly in the developed world over the last few years with the widespread use of more effective antiretroviral therapies and better predictive monitoring of the risk of clinical progression in

11

TABLE 11.5 Common complications of symptomatic HIV disease Early

CD4 count 100-500 x106/L *Constitutional symptoms (fever, diarrhoea, weight loss) *Herpes zoster *0ral hairy leucoplakia *0ral Candida *Skin problems

Late

CD4 count 200-50 x10 6 /L Recurrent bacterial pneumonia Pulmonary tuberculosis Pneumocystis carinii pneumonia Oesophageal Candida Kaposi's sarcoma Cryptosporidial-microsporidial diarrhoea Cryptococcal meningitis Cerebral toxoplasmosis *HIV peripheral neuropathy

Advanced

C D4 count <50x10 6 /L CMV diseases: retinitis, colitis, polyradiculopathy, encephalitis Disseminated MAI infection Severe wasting syndrome AIDS-related dementia Lymphoma

FIG. 11.10 Risk of progression to AIDS at 3 years by CD4 count and plasma viral load (adapted from Mellers JW, Munoz A, Giorgi JB et al. Ann of Intern Med 1997 126: 946-954).

Progressive multifocal leucoencephalopathy * Symptomatic non-AIDS.

Pulmonary complications an individual. Opportunistic infections, however, remain common in the developing world, and in those who are newly diagnosed in late-stage disease or who decline antiretroviral therapy or who have experienced multiple treatment failures. The organisms responsible for the opportunistic infections that occur in patients with AIDS are unusual pathogens. Most of the infections are due to reactivation of latent organisms in the host or, in some cases, to ubiquitous organisms to which we are continually exposed. The infections are often difficult to diagnose because conventional serological tests are unhelpful, and treatment often suppresses rather than eradicates the organisms. If patients are able to respond to effective antiretroviral combinations, then immune reconstitution can occur and relapse of opportunistic infections is uncommon. In the absence of effective antiretroviral therapy severe immune deficiency persists, and relapse of opportunistic infections is common and continuous treatment with prophylactic drugs may be necessary. Three main organ systems are affected: the respiratory system, the gastrointestinal tract and the central nervous system. In addition, patients may present with a history of night sweets, chronic ill health, fevers or weight loss.

Pneumocystis carinii pneumonia (PCP) is one of the commonest life-threatening opportunistic infections in patients who progress from symptom-free infection to AIDS. The presentation is subacute, and malaise, fatigue, weight loss and shortness of breath often develop over several weeks. Typical retrosternal or subcostal chest discomfort, associated with increasing shortness of breath, a dry cough and fever, finally cause the patient to seek help. The chest X-ray at presentation may be normal or show bilateral fine infiltrates, which are typically perihilar (Fig. 11.11). The arterial oxygen tension is usually depressed, and the carbon monoxide transfer factor, when available, is low. A fall in oxygen saturation levels from normal at rest to below normal on exercise is a useful sign of early infection. The diagnosis is confirmed by cytological examination of induced sputum or of bronchoalveolar lavage (BAL) from fibreoptic bronchoscopy. First-line treatment is high-dose co-trimoxazole for 18-21 days, but intolerance may be a problem owing to nausea, rash and bone marrow suppression. Pyogenic bacterial causes of pneumonia should always be considered, particularly as its presentation may be atypical. The radiological appearances may include diffuse infiltrates as well as the more typical focal or lobar patterns. Bronchoscopy is frequently helpful to differentiate

445

Gastrointestinal and hepatic complications Oral and oesophageal complications Oral candidiasis occurs frequently (see Fig. 11.8). Oesophageal candidiasis is the commonest cause of dysphagia or retrosternal discomfort and may frequently be the first AIDS-defining illness (Fig. 11.12). Oral candidiasis alone does not fulfil the criteria for AIDS. Oesophageal infection is best shown by culture or biopsy at endoscopy, although plaques of Candida albicans can often be seen during a barium swallow. Ulceration may be focal or diffuse. Cytomegalovirus and herpes simplex virus may both cause a similar pattern of ulceration in the oesophagus (and also may affect the stomach and duodenum), and should always be considered if oesophageal symptoms do not resolve on treatment with systemic antifungal agents.

FIG. 11.11 Chest radiograph of Pneumocystis carinii pneumonia, showing bilateral interstitial shadowing

between PCP, pyogenic bacteria and other infections, particularly in patients with an atypical presentation or who fail to respond to treatment. Cryptococcal pneumonia may occur in patients with cryptococcal meningitis and disseminated infection. Cytomegalovirus may be seen in BAL specimens, but appears not to be a major cause of pneumonia. 1 Infection with Mycobacterium tuberculosis occurs and, since 1993, has constituted a diagnosis of AIDS (see Case study Fig 11.1.1, p. 454). The downward trend of reported cases of tuberculosis in the United States has slowed at the same rate as the HIV epidemic has grown. Pulmonary tuberculosis can occur at different stages of immunosuppression, but tends to appear earlier in HIV disease than the atypical mycobacterial infections that complicate severe immunosuppression of advanced AIDS. In subSaharan Africa tuberculosis, both extrapulmonary and pulmonary, is a major cause of morbidity and death in HIV-infected persons. 2

Diarrhoea, malabsorption and weight loss Diarrhoea is a common symptom in patients with chronic HIV infection, with or without other manifestations of AIDS. In the majority of cases a pathogen is found, although an enteropathy with malabsorption may occur secondary to HIV infection. Cryptosporidium, a coccidian protozoal parasite, is probably the commonest pathogen isolated from patients with AIDS who have diarrhoea. It is the commonest of the protozoal causes of diarrhoea, which also include Isospora belli and microsporidia. In immunocompetent human

SUMMARY 1 Pulmonary infiltrates in HIV disease 1. Differential diagnosis Infective:

• Pneumocystis carinii pneumonia • Bacterial infection:

•

• • •

Streptococcus pneumoniae Haemoph/lus influenzae Pseudomonas aeruginosa Mycobacterial: M. tuberculosis Atypical mycobacteria M. avium-intracellulare M. kansasii M. xenopi Cryptococcus neoformans Toxoplasma gondii Cytomegalovirus

Non-infective: • Kaposi's sarcoma • non-Hodgkin's lymphoma • non-specific interstitial pneumonitis 2. Investigation

1

446

Fig. 11.1

2

MCQ 11.3

3

Fig. 11.2

Arterial blood gases or oximetry Sputum examination: Smear for acid-fast bacilli, Culture Fibre-optic bronchoscopy and broncho-alveolar lavage CT scanning, Nuclear medicine scanning Open lung biopsy

More recently, with improved diagnostic techniques, microsporidium, a small obligate intracellular protozoa, has been increasingly identified as a possible cause of diarrhoea and weight loss in patients with AIDS where no other pathogen had previously been found. Cytomegalovirus and herpes simplex virus can cause focal or diffuse ulceration of the gut, from the mouth to the anus. Herpes simplex virus most commonly causes mucocutaneous lesions at the upper and the lower ends of the gastrointestinal tract, whereas cytomegalovirus most frequently causes a colitis or oesophageal ulceration. Atypical mycobacteria of the avium intracellular complex are ubiquitous organisms that have little virulence for the immunocompetent host. Disseminated infection with Mycobacterium avium-intracellulare (MAI) of several organs occurs in patients with AIDS. Gastrointestinal infection may be associated with fever, weight loss, abdominal pain, diarrhoea and malabsorption. Diagnosis can be made by acid-fast staining of the stool or biopsy material, or both, or culture of blood or tissue. Shedding of atypical mycobacteria in stools alone does not indicate disseminated disease, but may often precede the presenting symptoms of disseminated infection. These are typically high fevers, anaemia and weight loss. 0 The diagnosis is most frequently confirmed on blood cultures or bone marrow examination. Mycobacterium tuberculosis infection of the bowel does occur, but is less common. Campylobacter and Salmonella species infections may cause diarrhoea, but the latter more commonly presents as a fever of unknown origin with bacteraemia.

FIG. 11.12 Barium swallow of oesophageal candidiasis The figure shows extensive oesophageal ulceration.

hosts cryptosporidium produces a transient diarrhoeal illness. In people infected with HIV it can cause transient, intermittent or persistent diarrhoea, ranging from loose stools to watery diarrhoea, colic, and severe fluid and electrolyte loss. Oocysts can be found in stools. If direct smears of unconcentrated faecal samples stained with iodine or modified acid-fast stains fail to show the oocysts the samples should be concentrated. The diagnosis should not be discounted without examining multiple specimens.

Hepatitis and cholestasis At all stages of HIV infection the commonest causes of acute hepatitis, both symptomatic and asymptomatic, are either drug toxicity or coinfection with viral hepatitis B or C. Treatment with many of the current antiretroviral agents is associated with a risk of drug-induced hepatitis, e.g. the nucleoside analogue stavudine, the non-nucleoside reverse transcriptase inhibitor nevirapine (usually in association with a rash and fever) and the protease inhibitor ritonavir. Severe acute hepatitis with liver failure secondary to hepatitic steatosis (in which the liver cells are filled with small fat droplets) is a rare complication of treatment with the reverse transcriptase inhibitors. Drugs used to treat specific acute opportunistic infections may also be implicated, and include the antimycobacterial agents, systemic azoles and high-dose co-trimoxazole. When multiple drugs are being taken then drug-induced hepatitis must always be considered. Acute primary infection with viral hepatitis A, B or C should be considered. Chronic infection with viral hepatitis B and C is more common among homosexual men and injecting drug users. In those coinfected with HIV there is a higher likelihood of progression to more severe chronic liver disease and cirrhosis. In patients with AIDS specific opportunistic infections

11

447

can cause hepatitis, which may present as fever, abdominal pain and hepatomegaly. If an ultrasound does not show dilated intra- or extrahepatitic bile ducts, suggesting obstruction, needle biopsy often shows granuloma hepatitis, usually caused by atypical mycobacteria rather than M. tuberculosis. The herpes viruses may also occasionally cause hepatitis as part of a disseminated infection. Acalculous cholecystitis and cholangitis show an endoscopic retrograde cholangiographic picture similar to that of primary sclerosing cholangitis, with strictures and dilatation of the biliary tree. Cryptosporidium and cytomegalovirus have been isolated and are implicated as a cause of this syndrome (Fig. 11.13).

tomatic disease. At necropsy up to 90% of patients dying of AIDS have chronic subcortical encephalitis, characterized by infected macrophages and microglial cells that fuse to form multinucleate giant cells. There is also patchy demyelination and astrogliosis.

Neurological complications Cerebral toxoplasmosis is the commonest cause of intracranial mass lesions and usually presents with focal symptoms and signs (Fig. 11.14). The main differential diagnosis is primary intracerebral lymphoma. Cytomegalovirus commonly causes a retinitis (Fig. 11.15) and presents with blurring and/or partial loss of vision. It may eventually lead to blindness. CMV may also cause a polyradiculoneuropathy or encephalitis, which is more rapid in onset and progression than HIV encephalitis alone. Differentiation may be helped by the detection of cytomegalovirus DNA by polymerase chain reaction (PCR) in the cerebrospinal fluid. Meningitis due to Cryptococcus neoformans should always be considered in patients with advanced disease presenting with headache and fever, although symptoms may be more non-specific. Infection with the papovavirus JC is responsible for progressive multifocal leukoencephalopathy (PML), a form of demyelination which causes progressive neurological impairment such as hemiparesis, with or without aphasia. 1 The majority of persons who develop PML have a positive PCR test for JC virus in their cerebral spinal fluid.

FIG. 11.13 ERCP of HIV associated sclerosing cholangitis The figure shows dilation of the common bile duct and dilation and stricture of the hepatiticbilarytree.

HIV NEUROLOGICAL DISEASE Apart from opportunistic infections, chronic HIV infection is associated with several syndromes that affect the nervous system, in addition to the transient meningoencephalitis, myelopathy and peripheral neuropathy of acute infection. These neurological diseases are due to the direct or indirect effects of HIV and not to opportunistic infection. AIDS-related dementia (HIV encephalopathy) has been estimated to occur in 10-40% of patients with symp-

448

1

Case 11.2

4

MCQ 11.4

2 Fig. 11.3 5 Fig. 11.5

0 Fig. 11.4 FIG. 11.14 CT brain scan of cerebral toxoplasmosis The figure shows multiple ring enhancing lesions with oedema.

11

FIG. 11.15 Cytomegalovirus retinitis in a patient with AIDS The figure shows extensive areas of white granular retinal necrosis and haemorrhage. ©

The clinical features are characterized by cognitive and behavioural changes that include memory loss, apathy and impaired concentration and attention. Neurological examination may show hyperreflexia, hypertonia and frontal lobe signs. CT or MRI often show cerebral atrophy © and non-specific changes in the white matter. The cerebrospinal fluid findings are non-specific. Opportunistic infections, intracranial mass lesions, metabolic encephalopathy and neurosyphilis should be excluded. Treatment with antiretroviral therapies may improve symptoms in those with early cognitive impairment, but probably have little impact on those with severe AIDS dementia other than to reduce their risk of subsequent opportunistic infections. HIV infection is also implicated in vacuolar myelopathy, affecting primarily the posterior and lateral spinal cord, an aseptic meningitis and the following neuropathies: axonal sensory, chronic inflammatory demyelinating and mononeuropathies. 45

FIG. 11.16 Cutaneous lesions of kaposi sarcoma in AIDS, appearing as purple raised plaques or nodules on the skin surface

TUMOURS Kaposi's sarcoma Kaposi's sarcoma (KS) is a presenting feature in 16% of patients, though the incidence has fallen over recent years. It is commoner in homosexual men than in the other groups at risk. The Kaposi's sarcoma of AIDS differs from classic Kaposi's sarcoma in that widespread skin (Fig. 11.16), mucous membrane (particularly the oral cavity and palate), visceral and lymph node disease occurs. Visceral, particularly gastrointestinal, lesions are present in as many as half of all patients at presentation. Nodules of Kaposi's sarcoma also occur in the lungs. Chest X-ray appearances vary from confluent irregular masses to interstitial nodularity and are better illustrated on a CT scan of the thorax (Fig. 11.17). CT of the thorax

FIG. 11.17 CT scan of pulmonary kaposis sarcoma The figure shows irregular masses and nodules in the lung parenchyma.

may be useful in differential diagnosis. At bronchoscopy, endobronchial lesions may be seen. Kaposi's sarcoma consists of spindle-shaped cells arranged in nodules and broad bands, and contains vascular slits filled with extravasated

449

erythrocytes. The diagnosis of Kaposi's sarcoma in very early skin lesions may be extremely difficult, as little more may be seen than a few irregular dilated vascular channels in the mid dermis and a mild inflammatory cell infiltrate. A herpes virus, kaposi sarcoma-associated herpes virus 8 (KSHV8), has been identified in kaposi sarcoma lesions of all forms, including both classic and AIDS associated. The incidence of positive serology to KSHV8 in different HIVinfected patient populations reflects the incidence of KS in these patient groups. The presence of KSHV8 in blood is predictive of the subsequent risk of developing KS. These observations suggest a causal role for KSHV8 in the development of KS.

Non-Hodgkin's lymphoma As in other chronically immunosuppressed patients, there is a greatly increased risk of non-Hodgkin's lymphoma. Although the tumour typically affects lymph nodes, extranodal disease is common and affects the central nervous system, bone marrow and gastrointestinal tract. The diagnosis should also be considered in patients with weight loss, constitutional symptoms and anaemia. These are B-cell lymphomas of high or intermediate grade. Response to cytotoxic drugs occurs, but treatment is difficult and usually unsuccessful because of profound immune suppression. In addition to secondary involvement of the CNS a proportion of AIDS associated Non-Hodgkin's lymphoma presents as primary CNS lymphoma; tumours which are intracranial parenchyma lesions limited to the CNS. Patients usually present with single large mass lesions in the brain (Fig. 11.18). 1

Other tumours Squamous carcinomas of the mouth and anorectum occur in homosexual men with antibodies to HIV. Human papillomavirus may play a part. There is a higher proportion of cervical intraepithelial neoplasia (CIN) in HIV infected women and a higher rate of progression from lower to high grade CIN.

TREATMENT AntiretroviraIs Acute primary infection Treatment with antiretroviral agents should be considered in patients with acute primary HIV infection.

1 Fig. 11.6 450

FIG. 11.18 CT brain scan of primary cerebral lymphoma The figure shows a large single ring enhancing lesion with mid-line shift.

There is, however, no substantial evidence that antiretroviral therapy started at this stage results in greater longterm clinical benefit than if treatment is initiated during established chronic infection. However, the initiation of therapy within a few weeks of infection can result in immunological benefit, with preservation of specific Thelper cell responses to HIV, higher T-helper (CD4) cell counts and better HIV-specific cytotoxic T-lymphocyte responses. Although these preliminary immunological observations are impressive, their relative importance in the long-term treatment of HIV infection is not known. It is also uncertain whether the treatment initiated during primary infection should be continued long term or can be stopped after 1-2 years, with preservation of immunological benefit. Trials are under way to evaluate this treatment strategy. Until further evidence is available the decision to start antiretroviral therapy at this stage is a balance between preserving immune responses and the possible risk of drug toxicity, therapy failure, and the lack of evidence of long-term clinical benefit. Chronic adult infection The treatment of HIV infection has improved markedly over the last few years, and has resulted in decreases in the incidence of opportunistic infections and an improvement in survival. The results of large randomized clinical trials in the mid-1990s established a strong relationship between falls in plasma viral load occurring on therapy and clinical benefit in the short to medium term. It is now accepted that the objective of therapy is to sustain the suppression of viral load to levels that are preferably below the limits of

detection with the currently available ultrasensitive assays (less than 50 copies per mL). The suppression of viral replication results in sustained increases in CD4 cell counts (both memory and naive T cells), increases in the T-cell antigen repertoire, and re-establishment of T-cell responses to various antigens (e.g. CMV, mycobacteria). Failure to suppress plasma viral load optimally is associated with subsequent rebound of the viral load, the emergence of viral isolates with reduced drug sensitivity and subsequent falls in CD4 cell counts, and continued damage to the immune system. Currently there are three classes of antiretroviral drug available. These are nucleoside and non-nucleoside reverse transcriptase inhibitors (NRTIs and NNRTIs) and the protease inhibitors (PIs). The NRTIs are direct competitive inhibitors of the viral enzyme reverse transcriptase. NNRTIs are indirect inhibitors that bind to a site distant to the active site of the enzyme. The protease inhibitors inhibit the viral enzyme protease, which is responsible for the post-translational cleavage of precursor structural core proteins following budding of the HIV virion (see Fig. 11.6). Clinical trials have shown that a combination of three or more antiretroviral drugs results in greater and more sustained falls in plasma viral load, and increases the CD4 count more than with mono or dual therapy. It is currently usual standard clinical practice to combine two NRTIs with either one NNRTI or one or two PIs (Table 11.6). Comparative trials have shown that a triple combination containing either a PI or an NNRTI have similar antiviral efficacy at 12-18 months. Although there is more substantial trial evidence of the clinical benefit of PIs, it is generally accepted that NNRTI-containing regimens will result in similar clinical benefit, in view of their similar effects on viral load and CD4 count. It remains unclear, however, which regimen is the better for patients who have never received treatment. Recent trials have also demonstrated that combinations of 3 NRTIs as initial therapy to have similar efficacy to certain combinations to 2 NRTIs + either 1 NNRTI or 1 PI. It is also not yet clear which regimen should be used in patients who have experienced one or two failures of therapy. Trials are under way to assess different treatment strategies in each of these situations. There is no evidence from randomized clinical trials that helps to decide with the current therapies when, in the natural history of HIV infection, to initiate therapy. Treatment guidelines (Table 11.7) have used surrogate (laboratory-based) data from prospective studies of the natural history of the disease to guide the time to start therapy. These data have been added to the current knowledge of the effectiveness and tolerability of current combination treatments and the willingness of an infected person to start therapy. Most patients will start therapy once their CD4 count is 350 x 106/L or below, but national guidelines vary and so do the motivation and willingness of individual patients. A significant proportion of patients present and are

11

TABLE 11.6 Antiretroviral regimens (2001) 1. 2 NRTIs plus either: 1 NNRTI or 1PI or

2 PIs 2. 3 NRTIs

e.g. zidovudine or stavudine + lamivudine or didanosine e.g. nevirapine or efavirenz e.g. nelfinavir or lopinavir/r e.g. saquinavir + ritonavir zidovudine, lamivudine, abacavir

Recommended antiretroviral regimens for the initial treatment of established chronic infection in adults (2001) NRTis, nucleoside reverse transcriptase inhibitors NNRTIs, nonnucleoside reverse transcriptase inhibitors PIs, protease inhibitors.

TABLE 11.7 Treatment guidelines for starting antiretroviral therapy adults (2001) Disease stage

BHIVA(1)

US DHHS (2)

Symptomatic Asymptomatic (i) CD4<200x10 6 /L (ii) CD4 200-350 x106/L

Treat

Treat

Treat Consider treatment depending upon VL, rate of CD4 count decline symptoms and patient wishes Definitely

Treat Treatment should be generally offered

(ii) CD>350x10 6 /L

Defer or consider treatment if high VL

(1) BHIVA, British HIV Association Guidelines (Oct 2001). (2) USDHHS, United States Department of Health and Human Services (Feb 2001). VL, viral load. Treatment guidelines are constantly reviewed and updated.

diagnosed in latent disease when symptomatic and have CD4 counts of <200 x 106/L. Even in patients who are newly diagnosed at a late stage of disease, starting antiretroviral therapy can result in substantial increases in CD4 count and sustained clinical benefit. Although clinically effective, current antiretroviral drugs are not ideal. In patients not previously treated with antiretroviral drugs only 50-60% will sustain falls in plasma viral load to undetectable levels, as measured by the current ultrasensitive assays at 1 year of therapy. Poor tolerability and side-effects account for a high proportion of therapeutic failures, together with poor adherence and limited antiviral potency. Excellent compliance with combination therapies is a major determinant of therapeutic success. Every effort should be made, in both choice of

451

TABLE 11.8 Treatment guidelines for major opportunistic infections Infection

Drug

Dose

Duration 14-21 days 6 weeks

Pneumocystis carinii pneumonia

Trimethoprim/sulfamethoxazole

15-20/75-100 mg/kg/day (divided doses)

Cerebral toxoplasmosis

Pyrimethamine + sulfadoxine orclindamycin

Cytomegalovirus retinitis

Ganciclovir cidofovir

200 mg on first day then 50mg o.d. 4-6 g daily 600mg q.d.s. 10mg/kg/day/i.v. 5mg/kg/weekly/i.v.

Cryptococcal meningitis

(i) (ii)

Amphotericin B ± flucytosine or fluconazole

Disseminated MAI

Clarithromycin + ethambutol + rifabutin

Oesophageal candidiasis

Ketoconazole Fluconazole

0.5-1.0mg/kg/day/i.v. 75-100 mg/kg/day 600-800 mg/day 500mg-1gb.d. 15 mg/kg/day 400-600 mg/day 200-400 mg/day 50-100 mg/day

2 weeks 6 weeks

Indefinite

7-14 days

Following treatment course all major opportunistic infections require maintenance treatment regimens at adjusted doses. With effective antiretroviral therapy CD4 counts increase and secondary prophylaxis and treatment for MAI infection may be discontinued.

therapies and clinical and psychological support, to aid adherence to treatment. The emergence of viral isolates which are resistant to antiretroviral drugs is common in patients experiencing therapy failure. Phenotype and genotype assays have recently become available, and initial studies have demonstrated better virological outcome from their use in guiding the choice of new antiretroviral regimens in patients who are experiencing treatment failure. Recently long-term toxicities with antiretrovirals have been increasingly recognized, with patients experiencing metabolic toxicity (e.g. diabetes, hyperlipidaemia and lactic acidosis) and fat redistribution syndromes (lipodystrophy). The pathogenesis for some of these toxicities, particularly lipodystrophy, is not fully understood. Several predisposing factors seem to be involved, including specific actions of antiretroviral drugs, the previous treatment history and age of the patient. Antiretroviral agents with new sites of action are being developed. The binding of HIV to the CD4 molecule on the cell surface has long been a potential target for therapy. Recently, fusion inhibitors have been developed (T20) which bind to the HIV transmembrane protein GP41, preventing the conformational changes necessary for viral entry into the cell. Initial trials of T20 in patients who have experienced multiple therapy failures have shown good

452

1

MCQ 11.5

2

Case 11.3

3

Case 11.4

4

Case 11.5

antiviral response in the short term, suggesting that this new class of drugs maybe beneficial clinically. Another approach is through chemokine receptors, which act as coreceptors for HIV binding and determine the susceptibility of the target cell to infection by different HIV strains. New drugs are being developed which block chemokine receptor binding. 12

Opportunistic infections In the developed world, treatment of major opportunistic infections associated with HIV infection is usually effective. Mortality remains highest in those who present late with previously undiagnosed HIV infection, and treatment is frequently associated with a high incidence of drug sideeffects which limit tolerability. Treatment guidelines for the major opportunistic infections are outlined in Table 11.8. Alternative regimens for the treatment of Pneumocystis carinii pneumonia in those intolerant to high-dose co-trimoxazole are combinations of clindamycin and primaquine, or intravenous pentamidine and atovaquone. Steroid therapy is indicated in those who present with severe hypoxia (Pao2 < 9.3kPa), as this has

SUMMARY 2 AIDS prevention strategies • Control of generally sexually transmitted infections • Promoting safe sex behaviour in people with known HIV infection and those not infected or of unknown status • Promoting early diagnosis of HIV infection in those at risk • Ensuring good and equal access to HIV treatment services

been shown to improve clinical outcome and response. Although Pneumocystis resistance to co-trimoxazole has been demonstrated, poor clinical response to treatment is usually associated with late presentation or coinfection with bacterial or viral pathogens. Either intravenous ganciclovir or cidofovir is effective for the treatment of a first episode of CMV retinitis or other CMV end-organ disease, such as colitis or oesophageal ulceration. Although the dose scheduling is convenient, renal toxicity with cidofovir is common and frequent laboratory monitoring for toxicity is mandatory. With the initiation of effective antiretroviral therapy, relapse of CMV retinitis is uncommon. In those who do experience relapse, switching anti-CMV agents to combinations of foscarnet and ganciclovir or ganciclovir ocular implants maybe effective. For patients who present with a single or multiple intracerebral space-occupying lesion where toxoplasmosis is a probable diagnosis, the use of steroids should be avoided if possible because the interval MRI brain scan at 2 weeks, to determine an antitoxoplasmosis treatment response, may be difficult to interpret as a result of the effect of steroids on cerebral odema. In those not responding, a diagnosis of primary CNS lymphoma should be considered, particularly if it is a single lesion or there is supporting evidence from a gallium scan. The appearances on an MRI brain scan unfortunately do not aid differentiation. Gastrointestinal infection with cryptosporidium usually results in a self-limiting illness in patients with less advanced immunosuppression. However, in those with severe immunosuppression (CD4 count < 200) chronic infection can result in frequent diarrhoea, severe weight loss, nausea, vomiting, and salt and water depletion. Treatment with specific anticryptosporidial agents is disappointing. Initiation with antiretroviral combination regimens has, however, been shown to be effective as an effect of the improvement in the immune system. It is important for any patient presenting with a major opportunistic infection to be started on an effective antiretroviral combination regimen. Immune reconstitution on antiretroviral therapy will substantially reduce the risk of relapse and of additional opportunistic infections, and may allow the discontinuation of secondary prophylaxis. Effective immune reconstitution with antiretrovirals can also result in disease as T-lymphocyte responses to different antigens are regained. In patients with advanced immune deficiency (CD4 < 100 x 106/L) started on effective antiretroviral therapy, clinically apparent CMV and mycobacterial disease have been reported in the first few months of therapy. This is a result of the immune system mounting effective responses to previously silent infections. It is routine practice for primary prophylaxis against certain opportunistic infections to be started once the CD4 count falls to below 200 x 106/L, irrespective of whether the patient is on an antiretroviral regimen or not. Approxi-

mately one in five persons not on antiretroviral therapy will develop Pneumocystis carinii pneumonia within 1 year of their CD4 count falling to below 200 x 106/L. Primary prophylaxis with cotrimoxazole 960 mg once daily or an alternative regimen (e.g. dapsone 100 mg once daily) reduces this risk to a few per cent. In patients with a CD4 count less than 100 x 106/L and positive serology to toxoplasmosis, once-daily cotrimoxazole is also effective prophylaxis against cerebral toxoplasmosis. Dapsone 100 mg once daily and pyrimethamine 25 mg three times per week is an alternative in patients intolerant of co-trimoxazole. In patients with CD4 counts less than 50 x 106/L prophylaxis with azithromycin 1200 mg once weekly against disseminated Mycobacterium avium intracellulare infection should be considered. Alternative regimens are clarithromycin 500 mg twice daily, or weekly azithromycin plus rifabutin 300 mg daily. Chemoprophylaxis against mycobacterial tuberculosis may also be considered, particularly in those with a high background risk of exposure, but active disease needs to be excluded first, as this requires combination treatment. National guidelines differ, and prophylaxis may not be appropriate in populations with either very high or very low prevalence rates of clinical tuberculosis. Both secondary and primary prophylaxis to most opportunistic infections may be safely discontinued when the CD4 count on antiretroviral therapy has risen persistently to above 200 x 106/L, or even above 100 x 106/L for some opportunistic infections such as CMV disease. Prevention of environmental exposure to certain opportunistic infections may be possible, particularly to cyptosporidium and toxoplasmosis. Outbreaks of cryptosporidial diarrhoea from public water supplies can occur. Patients with severe immune deficiency should be advised to boil their water or use bottled water where the risk of contamination at source may be less than with public supplies. HIV-infected persons who lack IgG antibody response to toxoplasmosis should be counselled about the various sources of toxoplasma infection, and particularly should be advised not to eat raw or undercooked meat. 3 34

11

STRATEGY FOR CONTROL OF HIV AND AIDS Both HIV infection and other traditional sexually transmitted diseases (STDs) share similar risk factors in relation to their spread, but in addition they can potentate each other. Both of these facts need to be understood if effective control programmes are to be developed. The presence of an STD, particularly genital ulcers, can enhance both the acquisition and transmission of HIV by increased shedding of the virus within and from the genital tract. Several studies have shown that the risk of acquiring HIV in the presence of an ulcerative or non-ulcerative STD is increased 2-6-fold. The shared risk factors for the acquisi-

453

CASE STUDY 11.1 ACUTE DIARRHOEA IN HIV INFECTION A 45-year-old homosexual man presented with a 6-week history of liquid diarrhoea (5-6 times daily), weight loss of 6 kg, episodic abdominal pain and, more recently, nausea and anorexia. He had no recent travel history and no previous history of chronic diarrhoea. He had not noticed blood in his stool, or fever. He was not taking any medication and had no other history. He was a non-smoker and drank 15 units of alcohol per week. He had been found to be HIV antibody positive 6 years ago, but had not attended for regular monitoring and follow-up. On examination he was afebrile, he appeared wasted, had pseudomembranous oral Candida in his mouth and seborrhoeic dermatitis on his face. His abdomen was mildly distended with active bowel sounds, but was otherwise soft and nontender. His CD4 count on this occasion was 110 x 106/L, with a plasma HIV RNA level of 75000 copies per mL. Questions 1. What is the differential diagnosis? 2. How would you manage this patient? Discussion This man presents with chronic diarrhoea, abdominal pain and weight loss. He is known to be HIV antibody positive, and recent tests show that he is severely immunosuppressed, with a CD4 T-cell count of 110 x 106/L. On examination he has oral Candida and seborrhoeic dermatitis - both indicators of clinical progression to symptomatic HIV disease.

454

It is important to exclude an infective cause for his diarrhoea. Several stool samples should be sent for both bacterial culture and microscopy for ova cysts and parasites. (If the patient had taken antibiotics recently then diarrhoea due to Clostridium difficile toxin would also be considered). If repeat stool samples are found to be negative for pathogens, colonoscopy and duodenoscopy with biopsy of the bowel mucosa should be undertaken. If no pathogens are identified then a diagnosis of wasting syndrome secondary to HIV enteropathy can be made by exclusion. Patients with HIV enteropathy have partial villous atrophy of the small bowel mucosa. Gastrointestinal infection with cryptosporidium is frequently responsible for weight loss and persistent diarrhoea in HIV-positive patients with severe immunosuppression. Cryptosporidial infection (Case Fig. 11.1.1) causing diarrhoea for more than 1 month is an AIDS-defining illness. Both the large and the small bowel may

become infected. The small bowel involvement is associated with persistent nausea and recurrent vomiting. Prior to effective antiretroviral treatment gastrointestinal cryptosporidiosis in patients with CD4 counts less than 200 x 106/L was difficult to treat, and severe disease was associated with high morbidity and occasional mortality. There is uncontrolled evidence that a 4-week course of paramomycin may be helpful, but symptom control with antidiarrhoeal drugs was previously the mainstay of treatment. Today, with the availability of effective antiretroviral treatment, this man should be started on a triple combination regimen and given symptom control for his nausea and diarrhoea as required. Subsequent immune reconstitution on therapy should result in clearance of the cryptosporidium and resolution of the diarrhoea. HIV-infected persons with CD4 counts less than 200 should be advised to boil their water as a prophylactic measure against cryptosporidial infection.

CASE Fig. 11.1.1 Cryptosporidial infection of the small bowel

tion of HIV and other STDs, and the interaction between them, has led to the setting up of integrated control programmes in the hope of reducing STDs and slowing the spread of HIV. This strategy was reinforced by the results

of one study in Africa which showed a significant reduction of HIV transmission rates with the provision of STD treatment services. Control of HIV has two elements, primary and sec-

ondary. Primary prevention is aimed at stopping infection occurring in the first place. This is achieved through health education and programmes to market and encourage the use of condoms. In the UK, initial health education campaigns resulted in homosexual men, in particular, adopting safe sex practices. However, such changes are not always easy to sustain, and repeated reinforcement and monitoring is required. Secondary prevention is aimed at promoting health-care seeking behaviour, so that those who could be infected recognize this, regardless of symptoms, and as a result come forward for HIV testing and care if required. This is particularly important, as primary prophylaxis can be offered for PCP, and combination antiviral therapy can have an effect on morbidity and mortality. Everyone is potentially at risk of infection: even though the prevalence among those without recognized risk factors is currently low, it requires sexual contact with only one infected person for transmission to occur. Those who are HIV seronegative and in a mutually monogamous relationship have nothing to fear. The sensible message to everyone else must be: to reduce your risk, reduce your number of sexual partners, know about your partner's previous sexual and drug use history, and use a condom. Condoms may not provide total protection, but they will help considerably if used properly and every time. Preventing the transmission of HIV among injecting drug users must rely on stopping the sharing of needles and other paraphernalia that go with injecting drugs (syringes, mixing bowls, spoons etc.), as well as advising on safer sex practices. Users need to be advised of the risk of sharing; this applies to any form of injection, whether intravenous, intramuscular or subcutaneous. Equally important is advice on the risk of transmitting or acquiring the virus sexually, as well as the potential risk of both male and female prostitution, as this may be used to finance a drug habit. For any individual the best primary prevention is to stop using drugs, but if this is unrealistic at the time, the next option is to stop injecting and switch to sniffing, smoking or swallowing drugs. In the realization that people still continue to inject, a large number of needle exchange programmes have been set up throughout the UK, and are currently thought to be useful in cutting down the sharing of needles. Vertical transmission from mother to child is uncommon in the UK. The risk of an HIV-positive pregnant woman transmitting HIV to her unborn child is thought to be about 13% in developed countries, but up to 30% in the developing world. Intrapartum transmission accounts for the majority of cases of vertical transmission. Factors associated with a higher risk include prematurity, the use of instrumentation during birth, maternal diagnosis of AIDS, maternal viral load levels and breastfeeding. Seropositive women, or women who are considering parenthood, need to be counselled about the risk of pregnancy, to both the mother and her unborn child. A seropositive woman who becomes pregnant can obviously be offered termination, but treatment interventions substantially reduce the risk of perinatal transmission and therefore many wish to go to term.

Early clinical trials showed that zidovudine monotherapy given to mothers from the second trimester up to and including the intrapartum period, together with 6 weeks of zidovudine therapy given to the infant, results in an approximately 70% reduction in transmission rates. Subsequent studies have shown a relationship between the level of maternal plasma viral load on treatment at birth and the risk of transmission. In the developed world it is now common practice to initiate triple combination antiretroviral therapy to achieve and sustain maximum suppression of maternal plasma viral load levels during pregnancy and through the intrapartum period. The avoidance of breastfeeding is essential to sustain this reduction in risk. The implementation of these measures, however, is only possible if both the mother and her physician are aware that she is seropositive during pregnancy. HIV antibody testing should routinely be offered to all pregnant women, irrespective of the perception of risk. In the developing world vertical transmission is a major problem and complex triple combination therapies are too costly. Recent studies have investigated simpler and shorter-course treatments. A single dose of the nonnucleoside reverse transcriptase inhibitor nevirapine, given to the mother at the onset of labour and to the infant aged 48 hours, has been shown to reduce perinatal transmission by approximately 50%, compared to a short course of zidovudine given to the mother during labour and postnatally to the infant for 7 days. Peripartum therapeutic intervention will have no effect on intrauterine transmission. The reduction in transmission rates in infants at 6 weeks is likely to be lost in a breastfeeding population at 2 years, as a result of continued postnatal transmission. The widescale implementation of an effective therapeutic intervention to prevent perinatal transmission in the developing world is a challenge that remains to be met. The risk of occupational transmission of HIV as a result of needle-stick injuries has been estimated to be 0.3%. Although there is no substantial evidence of clinical efficacy, and based largely on biological rationale, it is recommended that health-care professionals who sustain a significant exposure to a bodily fluid from an HIV-infected person should take a triple combination regimen as postexposure prophylaxis for 4 weeks after the incident. Prophylaxis should be started as soon as possible after the incidence, preferably within 1-2 hours. The relative benefit of such a regimen remains uncertain. Postexposure prophylaxis has also been considered after sexual exposure, but in the absence of any data to support a relative benefit no specific guidelines have been recommended.

11

FURTHER READING Adler M W (ed) ABC of AIDS, 5th edn. London: BMJ Publishing, 2001. British HIV Association (BHIVA) Guidelines for the treatment of HIV infection. HIV infected adults with antiretroviral therapy. HIV Medicine 2001.

455

Granziosi C et al. Immunopathogenesis of HIV infection. AIDS Res Hum Retroviruses 1998; 14(Suppl 2):5135-5142.

SEXUALLY TRANSMITTED INFECTIONS GENITAL HERPES INFECTION The biology of herpes simplex infection and herpes labialis is dealt with in detail in Chapter 9 (p. 286). Genital herpes simplex virus infection is the commonest ulcerative STI in England and Wales and reported cases have increased severalfold over the last 20 years, particularly in females aged 16-24 years.

FIG. 11.19 A clinical episode of genital herpes affecting the penis The figure shows multiple small shallow ulcers around the foreskin.

Primary genital infection HSV 2 is largely associated with genital infection. Primary genital infection, or the first clinical episode, is most common in young, sexually active adults, and the risk of infection is proportional to the number of sexual partners. After an incubation period of 2-7 days, clusters of vesicles with surrounding erythema appear at the site of virus entry, on the glans or penile shaft in men, or on the vulva and vaginal mucosa in women. Homosexual men may develop rectal and perianal lesions. The lesions are usually painful and there may be systemic illness, with headache, myalgia, fever and local lymphadenopathy. Occasionally paraesthesia may develop in areas supplied by infected nerves, and there may be associated urinary retention or constipation. Skin and mucosal lesions resolve over 2-3 weeks. Virus may be shed from the lesions for up to 12 days. Infection of neonates is discussed on page 286.

Recurrent genital herpes Recurrent clinical episodes cause lesions similar to, but milder than, those of the primary episode (Fig. 11.19), with more localized lesions occasionally accompanied by shooting pains in the buttocks or legs. Recurrent episodes are associated with less pain and are shorter in duration. Systemic manifestations are rare. Often patients notice prodromal symptoms of local paraesthesia 24-48 hours before lesions appear. The rate of recurrence is variable, although the majority of patients experience a recurrence within 12 months of the first clinical episode. Asymptomatic secretion of virus can occur in the mouth or in the vagina, but this is rarely a source of transmission.

1 456

MCQ 11.6

2

Case 11.6

Management of genital herpes First clinical episodes of genital herpes, when severe, are usually treated with 5 days of oral aciclovir (200 mg five times daily), with analgesics and saline baths. For recurrences, local hygiene is usually all that is required. If episodes are frequent and severe suppressive aciclovir therapy may be necessary. Counselling on the management of recurrences and the prevention of transmission is required.

MOLLUSCUM CONTAGIOSUM Molluscum contagiosum is an infection of the skin caused by a poxvirus. The incubation period is variable but is usually around 14 days. A typical lesion is a smooth, firm papule, with the vertex being umbilicated and containing a waxy substance. Lesions commonly appear on the genitalia, suprapubic area and inner thighs, but can appear anywhere on the skin surface. They appear to be more common in people who are HIV positive, when the face, including the eyelids, may be involved. Spontaneous regression may occur. Therapies include curettage and liquid cryotherapy, or the application of phenol after disruption of the lesions.

GENITAL WARTS Warts are caused by a small DNA virus (papillomavirus) belonging to the papovavirus group. Anogenital warts are sexually transmitted (p. 393) and are regarded as important in view of increasing evidence associating them with cervical cancer.

have increased significantly over the last few years, the highest among 16-19-year-old females.

11

Clinical features Men usually have urethral discharge and dysuria about 10 days after exposure to infection. In women the infection is usually silent, but may present with an abnormal vaginal discharge or with symptoms of pelvic inflammatory disease. If uncomplicated rectal infection occurs there may be anal discharge, perianal dampness, irritation and tenesmus. Complications are more common in women than in men. In women they include bartholinitis, salpingitis and perihepatitis. In men the complications include epididymo-orchitis, prostatitis and sexually acquired reactive arthropathy (Reiter's syndrome, p. 1151). FIG. 11.20 Genital warts involving the vulva and perineum

In men warts can appear on any part of the genitalia, but especially the frenulum, the coronal sulcus and the inner surface of the foreskin. They are called condylomata acuminata. They may be found in the male urethra and appear as bright red lumps at the meatus. In women the vulva and cervix can be involved (Fig. 11.20), as can the perineal area in both sexes. Cervical condylomata may be flat and identifiable only by colposcopy. They may be associated with cervical dyskeratosis and koilocytosis.

Management Local application of cytotoxic agents such as podophyllin, or of destructive agents or methods such as trichloracetic acid, cautery, diathermy or cryosurgery, is still the main therapeutic approach. Ablation of cervical warts by laser therapy is effective. Sexual partners should be contact traced and treated. Women diagnosed with genital warts, and the female partners of men with genital warts, should also have cervical cytology.

CHLAMYDIA TRACHOMATIS Genital infection Chlamydia trachomatis is an obligatory intracellular parasite whose replication results in the death of the infected cell. During its lifecycle the infectious particle changes to an actively dividing form, which later reorganizes into the infectious form that is released on cell lysis. Non-specific genital infection (NSGI), of which chlamydia infection may account for up to 50% in the UK, may present as urethritis, cervicitis or proctitis. Other relevant organisms include mycoplasmas and uroplasmas. Genital chlamydia infection is thought to be the commonest STI in England, with prevalances of 2-12% detected in studies of GP attendees. The rates of diagnosis

Diagnosis In men a Gram-stained urethral smear shows an excess of polymorphonuclear leukocytes but no gonococci, and is supportive of a diagnosis of non-gonoccoal urethritis. In men and women chlamydial infection is confirmed by identifying antigen at the site of infection, or by culture. Management Chlamydial infection and most other NSGI responds to tetracycline therapy. The usual is doxycycline (100 mg daily) for at least 7 days, or azithromycin 1 g orally as a single dose. Erythromycin (500 mg twice daily) may be used as an alternative and is the first choice in pregnancy and in tetracycline allergy. Relapse or recurrent infection is common in both sexes. There may be symptoms of a postinfective urethritis, with no signs of infection. Sexual partners of both sexes should always be examined and treated to identify asymptomatic infection, to prevent serious complications of untreated infection in the contact, such as pelvic inflammatory disease, and to avoid reinfection of the index case.

Congenital Chlamydia trachomatis infection Chlamydia trachomatis infection is now frequently seen as a cause of ophthalmia neonatorum. The infected baby develops a conjunctivitis about 10 days after birth. Investigation should include culture of pus, or a monoclonal antibody test for Chlamydia trachomatis. The organism may also be a cause of pneumonia in the neonate. Treatment is with erythromycin syrup. The baby's parents should always be examined for Chlamydia and other genital infections. 12

LYMPHOGRANULOMA VENEREUM Lymphogranuloma venereum (LGV) is caused by L1, L2 or L3 serotypes of Chlamydia trachomatis. It is most prevalent in parts of Africa, Asia and South America.

457

Clinical features The most common primary lesion of LGV is a shallow ulcer appearing on the genitals of men or women 3-10 days after exposure to infection. In men there may be an associated lymphangitis of the dorsal penis. More commonly in men there may be an inflammation and swelling of the inguinal lymph nodes 10 days to 6 months after infection. The inguinal bubo is unilateral in about 60% of cases and may enlarge, causing pain, and rupture. About 20% also have involvement of the femoral lymph nodes. This creates a groove over the inguinal ligament - the characteristic 'groove sign'. LGV may also cause a rectal infection. After initial inflammation of the rectal mucosa a chronic inflammatory process may follow; this invades the bowel wall, with noncaseous granulomas and crypt abscesses forming. A partial or complete stricture of the rectum may develop.

the extremities. Gonococcus is an important cause of acute septic arthritis (p. 1158). Disseminated gonococcal infection is commoner in women than in men. Signs include fever, pustular rash and arthritis. Infection in pregnancy may cause abortion or premature labour and is associated with high perinatal mortality. The baby can be infected in utero, at birth or postnatally, and at worst develops disseminated infection. Ophthalmia neonatorum is the more common sign of neonatal infection.

Diagnosis and management Diagnosis is based on a positive complement fixation or microimmunofluorescent (micro-IF) test. Doxycycline (100 mg daily for 2-4 weeks) is the therapy of choice.

Management A typical treatment regimen is ampicillin (3g) plus probenecid (1 g orally). This combination allows singledose treatment, probenecid delaying excretion of the penicillin. Recommended alternative regimens are ciprafloxacillin 500 mg orally, or ofloxacin 400 mg orally, both as a single dose. These are preferred for the treatment of pharyngeal infection as ampicillin (and spectinomycin) has poor efficacy in eradicating infection from this site. The prevalence of penicillinase-producing strains in the UK is less than 5%, and quinolones remain highly active against these resistant strains. The prevalence of quinolone-resistance strains in the UK is less than 2%, but has risen since 1993. The prevalence of strains resistant to both quinolones and penicillin is high in some parts of the world, particularly southeast Asia, and this needs to be considered in treating possible imported infection. In these cases alternative regimens include cefotaxime 500mg (i.m.) or spectinomycin 2mg (i.m.), both as a single dose. Coinfection with Chlamydia trachomatis occurs in up to 20% of men and 40% of women presenting with gonohorroea, and should be suspected if symptoms persist a week after treatment of gonoccocal infection. Combining effective anti-microbial therapy against Chlamydia with singledose therapy for gonorrhoea at first presentation may be appropriate, and should be considered if there is doubt about follow-up. Sexual contacts should be traced, examined and treated. 1

NEISSERIA GONORRHOEAE (GONOCOCCUS) INFECTION Gonorrhoea is an important STD that causes urethritis, endocervicitis or proctitis. The annual incidence of cases reported by clinics in the UK declined markedly in the 1980s, probably as a result of behaviour changes in response to the HIV epidemic. Since 1994 there has been a rise in the incidence of gonorrhoea cases, both heterosexually and homosexually acquired, suggesting that such behaviour modifications may not have been sustained. In 1998,20% of all cases of gonorrhoea were among homosexual men. Clinical features The incubation period is between 2 and 7 days. In men there is urethritis, with a creamy discharge, dysuria and frequency. Homosexual men also develop proctitis, with pruritus, tenesmus and rectal discharge, often containing blood. In women the infection is usually asymptomatic (discovered by contact tracing or during a routine examination at a GUM clinic) or causes vaginal discharge, dysuria, frequency and lower abdominal pain. Pharyngeal infection follows oral sex but is usually asymptomatic. Complications Ascending infection causes pelvic inflammatory disease, with salpingitis, peritonitis or abscess formation in 10-20% of infected women. Septicaemia may occur, with a rash on

1 458

MCQ 11.7

Diagnosis Microscopy of Gram-stained smears shows pus cells and both intra- and extracellular Gram-negative diplococci. Specimens should be taken from the endocervix in women: a high vaginal swab is inadequate. Culture is essential, as penicillin-resistant gonococci are becoming more common. Serology is not helpful in diagnosis.

Syphilis Syphilis is caused by Treponema pallidum (TP), one of a small group of treponemes. TP is 6-15 um in length and is narrow, with regular tight spirals, and mobile in all axes. Transmission is usually by sexual contact, although transplacental infection may occur. Exposure to moist mucosal or cutaneous surfaces is required; transmission is commonly during the first 1-4 years of infection. The incidence of syphilis in England and Wales fell after the intro-

11

CASE STUDY 11.2 URETHRAL DISCHARGE A 22-year-old homosexual man presented to a genitourinary medicine clinic with a 2-day history of urethral discharge and dysuria. He had no other symptoms but reported a history of oral sex and protected receptive anal intercourse with a new male partner 10 days ago. He also had a history of a condom accident with a casual partner 6 months ago, but no other history of unsafe episodes of sexual intercourse. He had been treated 4 years previously by his GP for scabies, but otherwise had no other relevant history. On examination he had a mucopurulent urethral discharge. Testis, scrotum and anorectal examination was normal. Small bilateral inguinal lymph nodes were palpable. Questions 1. What is the differential diagnosis? 2. How would you manage this patient?

Discussion This 22-year-old man presents with a urethral discharge and a recent history of a new sexual partner. Clinically he has urethritis and the differential diagnosis includes infection with either Neisseria gonorrhoeae, or Chlamydia triachomatis, or a non-specific urethritis. Microscopy of a urethral smear will show polymorphonuclear leukocytes, and a Gram stain should be undertaken to look for Neisseria gonorrhoeae. In men presenting with symptomatic gonococcal infection urethral microscopy is diagnostic in 90% of cases. In asymptomatic men the diagnostic sensitivity of a urethral smear is lower. A urethral swab for culture of Neisseria gonorrhoeae should be taken to confirm infection and allow for antimicrobial sensitivity testing. Urethral screening tests for Chlamydia infection should be performed. In view of this man's sexual history, rectal and pharyngeal cultures for gonorrhoea and routine serology tests

duction of antibiotic therapy (see Fig. 11.1), but there was a resurgence during the 1970s in males. These cases were mainly in homosexual men with numerous sexual partners. In the 1980s and 1990s there was a decline in the incidence of syphilis, with homosexual transmission now accounting for only one-quarter of all cases diagnosed in GUM clinics in 1998. Fifty per cent of all infectious syphilis in England was acquired outside the UK. However, very recently, in 2000, there has been reported a 2-3 fold increase in the number of cases of infectious syphilis amongst homosexual men in different cities in the UK. Worldwide there was a marked increase in reported cases in eastern Europe and Russia in the 1990s. Clinical features Syphilis is classified as acquired or congenital. Acquired syphilis may be divided into early or late. The stages of early syphilis are primary, secondary and early latent (less than 2 years of infection), and of late syphilis are late latent (more than 2 years of infection), tertiary and quaternary. Primary The primary lesion is a chancre (Fig. 11.21a), which starts as a painless pink papule and may develop into an ulcer.

for syphilis should be undertaken. A full sexual history should be taken from any patient presenting with symptoms of a possible sexually transmitted infection, and the appropriate diagnostic screen for a coincident sexually transmitted infection be offered. Referral to a genitourinary medicine clinic for management is strongly encouraged. In this man an HIV antibody test should be discussed and offered. In addition, if serological tests for either viral hepatitis B or A show he is nonimmune, he should be offered appropriate vaccination. His recent sexual partner should be traced and screened for sexually transmitted infection. He should avoid sexual intercourse until he and his partners have completed treatment and follow-up. The opportunity should be taken to give further advice on safer sex practices.

The ulcer has a button-like feel on palpation. The base is initially pink and shiny, becoming grey later. Painless inguinal lymphadenopathy is common and may be unilateral. On occasions the chancres may be multiple and painful. A chancre usually appears about 3 weeks after exposure to infection (range 10-90 days), and heals. Men, especially homosexual men, are infected more often than women. In the male the common genital sites of infection are the coronal sulcus, glans penis and prepuce. In the female chancres may be found in the vagina, and on the labia, fourchette and cervix. In both sexes infection can occur in the pharynx and anus. In the anal area a chancre resembles a fissure and may be painful. Chancres, because they can be small and painless, are often missed. Secondary The secondary stage may take 3 months or more after infection to appear (average 6 weeks). At this stage there is bacteraemia. Systemic upset is common, including low-grade fever, malaise, headache, generalized lymphadenopathy, including the epitrochlear glands, and rarely hepatosplenomegaly and transient proteinuria. Arthralgia, acute meningitis and cranial nerve palsies are rare features.

459

The skin is most frequently involved; a non-pruritic rash is common. The lesions are usually papules or plaques, and early on tend to be symmetrical and widespread. The palms, soles and face are often involved. Scaling is common and the lesions can resemble psoriasis. Later lesions tend to be fewer and asymmetrical. Patchy and partial hair loss can occur, giving a moth-eaten appearance (Fig. 11.21b). Other signs include patchy alopecia, erosive lesions of the mucous membranes (mucous patches), common in the mouth, and soft, flat-topped and highly infectious warts found in the perianal area, known as condylomata lata. At this stage all tests for syphilis are positive. The venereal disease research laboratory slide test (VDRL, see below) shows a very high titre. 1 12

skin ulcers, and may destroy the nasal septum. Lesions in bones and joints lead to fractures and joint destruction. Liver involvement causes fever and local tenderness. Gummas respond remarkably rapidly to treatment. Quaternary Quaternary syphilis is either neurosyphilis (p. 1428) or cardiovascular syphilis, which is due to endarteritis obliterans

Latent period If untreated, syphilis will go into a latent phase, when clinical symptoms and signs are absent. This is termed early latent syphilis if it is within 2 years of infection, or late latent after 2 years. Before the latent period there may be periods of infectivity: over 90% occur in the first year. These relapses may be characterized by rashes and lymphadenopathy. During the latent phase sexual transmission is rare, although fetomaternal transmission may occur. Transplacental passage becomes unlikely in late latent syphilis. The VDRL will be negative or have very low titre, e.g. 1:8. The T. pallidum haemagglutination assay (TPHA) and fluorescent treponemal antibody absorption test (FTA-Abs) - (see below) - are always positive. Late (tertiary) syphilis Tertiary syphilis is now rare, as effective treatment prevents progression. The characteristic pathological lesion at this stage is a gumma, which is a destructive granuloma sometimes associated with vasculitis. Treponemes are very scanty and not usually seen under the microscope. Gummas can involve any organ. They cause punched-out

SUMMARY 3 Sites of involvement in secondary syphilis The proportion of cases is given in parentheses Skin (90%) Lymph glands (50%) Mouth and throat (35%) CNS symptomatic (20%) asymptomatic (2%) Kidneys (rare) Gastrointestinal tract and liver (rare) Bones and joints (rare)

1

460

Case 11.7

2

Fig. 11.7

3

Figs 11.8,11.9

FIG. 11.21 Clinical features of syphilis A Chancre of primary syphilis B Alopecia in secondary syphilis.

of the vasa vasorum of the aorta. This causes aneurysm, most commonly of the ascending aorta (p. 600). Congenital syphilis TP crosses the placenta after the third month of pregnancy, usually when the mother has early disease. If she is not treated, infection of the fetus may cause abortion, stillbirth, or the birth of a child with congenital syphilis ©• The child may have signs of infection at birth, similar to those of secondary syphilis in adults, or may be asymptomatic but develop signs of late congenital syphilis, equivalent to adult tertiary, after the age of 2 years. The predominant signs in early presentation are snuffles and a rash, which may vesiculate. Possible late signs are facial abnormalities, including frontal bosses, saddle nose, protruding mandible, Hutchinson's incisors (peg-shaped, splayed), interstitial keratitis, eighth-nerve deafness, periostitis and Clutton's joints. As in adult syphilis, diagnosis depends on serology in the later stages. Diagnosis Primary syphilis is diagnosed by identifying TP with darkground microscopy. TP can be identified in serous exudate from the surface of the chancre. A negative result does not exclude the diagnosis, as topical (e.g. antibiotic ointments) or systemic therapy may alter the findings. The FTA-Abs is positive in approximately 70% of cases at this stage. Later stages require serological testing for confirmation of the diagnosis. Serological examination of the CSF is necessary if neurosyphilis is suspected. Antibodies to treponemal antigens are present, i.e. positive TPHA and FTA-Abs, in 80% of cases of primary syphilis. Two types of antibody test are employed (Table 11.9): • Non-treponemal tests (e.g. VDRL; rapid plasma reagin card test or RPR). These detect cardiolipin antibodies in serum. A high titre indicates an active infection. They

become negative after therapy or during the latent phase of untreated syphilis. False positive results may be found in acute febrile illness, autoimmune diseases, and after immunizations. False negative tests may occur in secondary syphilis due to the prozone phenomena from using undiluted serum. • Treponemal tests (e.g. FTA-Abs; TPHA). These detect antibodies against the treponemal antigen. They are positive earlier in the infection, but remain positive after therapy and in latent syphilis. Yaws, another treponemal disease caused by T. pertenue (see p. 329), gives a serological picture similar to that of untreated syphilis. A careful history is necessary to differentiate the cause, and if doubt about the diagnosis persists, treatment for syphilis is given. Management Table 11.10 shows acceptable treatment regimens. Benzathine penicillin does not effectively cross the blood-brain barrier. Patients may (but do not commonly) experience Jarisch-Herxheimer reaction (a transient fever) 1-2 hours after the first dose of penicillin. This is more commonly seen in secondary syphilis. In cardiovascular syphilis the inflammatory response to antibiotic therapy may result in small vessel occlusion in the coronary ostia. The use of prednisolone (24 hours before and up to 48 hours after commencing therapy) may prevent this. Serological examinations are repeated for at least 1 year after treatment, to

TABLE 11.10 Treatment for syphillis Primary, secondary or early latent syphilis Late latent syphilis or tertiary syphilis (not CNS) Neurosyphilis

TABLE 11.9 Serological tests for syphilis Test Disease/stage

VDRL or RPR*

TPHAf

FTA-Absf

Primary syphilis Secondary syphilis Latent (over 2 years) syphilis Latent (over 2 years) late syphilis Yaws Treated syphilis

Negative >1:64 <1:16

Positive/negative Positive Positive

Positive Positive Positive

Negative/<1 : 2

Positive

Positive

Negative/<1 : 2 Negative/<1 : 2

Positive Positive

Positive Positive

* Tests for cardiolipin antibodies: VDRL = Venereal Disease Research Laboratory slide test; RPR = rapid plasma reagin card test. Tests specific for treponemal antibodies: FTA-Abs = fluorescent treponemal antibody absorption test; TPHA = T. pallidum haemagglutination assay. f

11

HIV positive all stages Penicilin allergy Acquired syphilis <2yrs Penicilin allergy Acquired syphilis >2yrs Pregnant and penicillin allergy Acquired syphilis <2yrs Pregnant and penicillin allergy Acquired syphilis >2yrs

600 000 units procaine penicillin daily 600 000 units procaine benzylpenicillin (procaine penicillin) daily 2.4 million with procaine benzylpenicillin (procaine penicillin) daily 500 mg oral probenecid q.d.s. Neurosyphilis regimen Oral doxycycline 100 mg b.d. (or oxtetracycline 500 mg q.d.s.) Oral doxycyline l00 mg b.d. (or oxtetracycline 500 mg q.d.s.) Oral erymromycin 500 mg q.d.s. followed by oral doxycycline 100 mg b.d. after delivery* Oral erythromycin 500 mg q.d.s. followed by oral doxycycline 100 mg b.d. after delivery

10-14 days 14-21 days

17-21 days

21-28 days 14-21 days 30 days 15 days 15 days 30 days 30 days

* Treat newborn with procaine benzylpenicillin (procaine penicillin) at birth and advise mother not to breast feed during doxycycline treatment. An alternative to procaine benzylpenicillin (procaine penicillin) as first line therapy is: Eary syphilis: IM benzylpenicillin 2. 4 MU x 1-2 (days 1 & 8) Late latent syphilis: IM benzylpencillin 2.4 MU x 3 (days 1, 7, 14).

461

ensure treatment response and identify relapse. This is of especial value in primary and secondary syphilis. Tracing the partners of patients is an integral part of treatment. If the erythromycin regimen is used for a pregnant patient it is always advisable to treat the baby at birth with the appropriate dose of bicillin. Because erythromycin is not a fully reliable treatment, the mother should be treated with doxycycline after delivery. Like all tetracyclines, doxycycline is of course contraindicated during breastfeeding. 1

TRICHOMONIASIS Trichomonas vaginalis (TV) is a pear-shaped flagellate protozoan that causes vaginal discharge and urethritis in women and urethritis in men. It is a sexually transmitted infection. Clinical features Most women with TV have an offensive vaginal discharge associated with vulvovaginal pruritus, and occasionally

1 MCQ 11.8 462

dysuria and dyspareunia. Clinically the vulva is erythematous and excoriated; the vaginal discharge may be seen at the introitus. The discharge is grey, yellow or green and may be frothy. Vaginal erythema and punctate haemorrhages of the cervix may be seen. It is a rare cause of nonspecific urethritis in men. Diagnosis and management Examination of the discharge on a wet mount is essential. The trichomonads are easily seen. Metronidazole (2 g as a single dose, or 400 mg twice daily for 5 days) is effective, but recurrence is common. Sexual partners should be treated.

FURTHER READING ON GENITOURINARY MEDICINE Adler M W (ed) ABC of sexually transmitted diseases, 4th edn. London: British Medical Association, 1999. UK National Guidelines for Sexually Transmitted Infections and closely related conditions. Sexually Transmitted Infections 1999;75:Suppl 1.

12

Cardiovascular Didease J Malcolm Walker and Lip-Bun Tan

The scope of modern cardiology 463

Normal and disordered physiology

Complications of myocardial infarction 560

463

Prognosis after myocardial infarction 566

Symptoms and signs of heart disease 467

Management of myocardial infarction survivors 567

Cardiac investigation 476

Diseases of heart muscle 570

Heart failure 488

Infective endocarditis 574

Arrhythmia 500

Non-infective endocarditis 580

Resuscitation 517

Cardiac tumours 581

Valvular disease of the heart 520

Cardiac disease and pregnancy

Congenital heart disease 533 Cyanotic congenital heart disease

582

Secondary heart disease 583

536

Pericardial disease 586

Acyanotic congenital heart disease 538

Systemic blood pressure and hypertension 589

Coronary heart disease (CHD)

Vascular disease 600

542

Peripheral vascular disease 602

Myocardial infarction 554

THE SCOPE OF MODERN CARDIOLOGY More than 40% of patients admitted to medical wards in the UK have some form of heart disease. Coronary artery disease, heart failure and cardiac arrhythmias are among the commonest causes of emergency hospital admission. Ischaemic heart disease is the most important cause of death in the western world. It causes more than a million deaths per year in the USA and about 160000 deaths annually in the UK (Fig. 12.1), where it accounts for 40-50% of all deaths and is the main reason for men not reaching their seventh decade. A recent concerted effort to change lifestyles with respect to smoking, diet and exercise, as well as better control of hypertension, appears to have led to a decline in coronary artery disease deaths in the USA and the more

affluent northern European countries, but has yet to make a major impact in the UK. Rheumatic heart disease has declined rapidly in importance in the west over the last 30-40 years (p. 580). It now accounts for less than half the cases of heart valve disease coming to surgery and, with only a small number of new cases presenting, its importance will probably decline further. However, it is still a major cause of morbidity and mortality in the developing world. Congenital heart disease is one of the principal causes of neonatal death and is likely to remain a significant cause of morbidity and mortality in childhood. However, more cases are now treatable by cardiac surgery in childhood, and major cardiac defects can now be detected during pregnancy, raising the possibility of neonatal, or even antepartum, cardiac surgery. The development of sophisticated methods of investigation in cardiology now allows great precision in diagnosis, enabling full advantage to be taken of the advances in cardiac surgery and catheter-based treatments. Diagnostic methods have expanded from simple ECG and X-ray to include an ever-widening range of techniques, using ultrasound (both imaging and Doppler), isotope imaging, angiography, intracardiac electrophysiology and 24-hour ECG recording, pressure measurement and magnetic resonance imaging. Therapeutic options have also increased, to include complex cardiac surgery possible at all ages, pacemaker therapy and non-surgical interventions such as angioplasty. The efforts of cardiac surgeons and bioengineers have produced a generation of patients with prosthetic heart valves, implanted pacemakers, modified congenital heart disease and heart transplants. These present a new spectrum of pathology and altered natural history.

NORMAL AND DISORDERED PHYSIOLOGY CONTRACTION Cardiac muscle is striated and shares with skeletal muscle the same basic contractile unit, the sarcomere. The components of the sarcomere are the myosin thick filaments, and the thin filaments composed of actin, troponin and tropomyosin. A local increase in calcium concentration allows an interaction to occur between the heads of myosin molecules and the thin filaments, which produces crossbridges between the filaments. A confirmational change between the globular head of the myosin molecule and its rod-like tail causes a tension in the cross-bridge, making the filaments slide past each other (Fig. 12.2). Relaxation is brought about by a decrease in intracellular calcium and the hydrolysis of adenosine triphosphate (ATP) by myosin ATP-ase to produce adenosine diphosphate (ADP) and inorganic phosphate (Pi).

463

FIG. 12.2 Diagrammatic representation of sarcomere contraction A In the relaxed state (diastole) tropomyosin inhibitory subunits prevent the interaction of the myosin heads with actin. B In the presence of calcium, this interaction can take place enabling contraction (systole) to occur,

FIG. 12.1 Mortality from cardiovascular disease A Causes of death from cardiac disease (England and Wales, 1999). B Deaths due to coronary artery disease in Scotland and the USA in the years 1968-1992, England and Wales to 1999 (UK Government Statistics). Note the steep decline in mortality in males and females in the USA.

Energy for contraction ATP is stored in only small amounts in heart muscle, so its production must occur at a rate equal to its consumption. Free fatty acids (FFA) are the major source of myocardial energy, yielding 2.5 times more ATP per gram weight than glucose. Glucose breakdown can occur in the absence of oxygen, but only two molecules of ATP are produced per molecule of glucose metabolized to pyruvate (instead of the 36 obtained by aerobic metabolism of glucose to carbon dioxide and water). The normal heart consumes lactate, which, by the action of lactate dehydrogenase, yields pyruvate.

The action potential 464

Rhythmic contraction and relaxation of myocardial cells

is achieved by the sequential release of calcium near the myofibrils and its subsequent removal. The sequence of events is outlined in Figure 12.3. The action potentials from different regions of the heart have different configurations, the most important of which is the spontaneous depolarization during diastole in the pacemaker cells of the sinoatrial node (Fig. 12.4, action potential A). This rise in membrane potential in diastole triggers the next depolarization and sets the heart rate. Areas distal to the sinoatrial node are inhibited from showing their intrinsically slower pacemaker potential, but will do so if the impulse from the sinoatrial node is absent or delayed. The rate of sinoatrial node depolarization is greatly influenced by p-adrenoreceptors. Sympathetic stimulation increases this rate and therefore speeds the heart rate, whereas parasympathetic stimulation produces opposite effects mediated by acetylcholine. Circulating catecholamines and changes in serum potassium and magnesium are among the other factors that affect the basic electrophysiological properties of cardiac cells. In non-pacemaker cells the wave of depolarization from adjacent cells is the stimulus for the inward flow of sodium ions that starts the cardiac cycle. Periods of absolute and relative refractoriness follow the prolonged plateau phase. Average action potential durations are about 300ms, but shorten in response to increasing heart rate and adrenaline. The rate of conduction of impulses through the heart is determined by the action potential shape. Depolarization is accompanied by a short-lived inward current, producing the fast upstroke of the classic cardiac action potential (Fig. 12.4, B, D, E). The fast inward current is maximal when the

12

FIG. 12.4 Action potentials from different regions of the heart A Sinoatrial node; B atria; C atrioventricular node; D Purkinje network; E myocardium; F is the electrocardiogram. FIG. 12.3 Relationship between action potential, transmembrane ionic movements, sarcomere contraction and muscle tension A With the onset of the action potential, sarcolemmal sodium channels open. B This is followed by release of calcium from the sarcoplasmic reticulum into the cytosol and sarcomere shortening. c Relaxation follows resequestration of calcium into the sarcoplasmic reticulum. Repolarization is accompanied by loss of potassium from the cell.

foregoing resting membrane potential (Vm) is most negative, the situation found in the Purkinje cells of the ventricle, where conduction velocities are at their highest. In the specialized tissues of the sinoatrial and atrioventricular nodes the upstroke of the action potential is slow and dependent on calcium influx (Fig. 12.4, A and C). These features account for the slow conduction through the atrioventricular node. Ischaemic injury can cause depolarization of cells in affected regions of the heart, leading to regional variation in conduction velocities and areas of conduction block.

HEART PUMP PERFORMANCE Normal function The basic physiological factors that influence the power of myocardial contraction and pump performance are: • Inotropy. dependent on the neurohormonal and ionic environment • Preload: the myocardial fibre length at which the contraction begins • Afterload: the force against which the muscle contracts • Rate and rhythm: dependent on the rate and pattern of stimulation. Inotropy The myofibrils produce tension in proportion to the

free calcium released into the cytosol, until saturation is reached and tension rises no further. Calcium entry into the cell occurs during the action potential plateau, and variations in contraction strength occur by the release of variable amounts of calcium from intracellular stores, or by a change in sensitivity of the myofilaments to calcium. In a steady state, calcium entry into the cell with each action potential is balanced by a loss to the extracellular space. Accumulation of calcium in intracellular stores is facilitated by increased levels of extracellular calcium or by high intracellular sodium, which may occur when the sodium pump is blocked by digoxin. Under normal conditions extracellular calcium concentration plays no part in the beat-to-beat control of the heart's performance. This is mainly affected by sympathetic discharge and circulating catecholamines, which produce an increased heart rate (chronotropic drive) and also enhance myocardial contraction. The final common pathway of this latter effect is achieved by increased myofibrillar calcium sensitivity. Changes in extracellular concentrations of potassium and pH affect myocardial contraction, but, as these concentrations are generally kept constant by homeostatic mechanisms, they do not control heart activity in normal circumstances, although they may considerably affect it in disease. Preload Venous return to the heart is dependent on right atrial pressure. During inspiration intrathoracic pressure falls, causing an accentuation of venous return to the right heart. Standing up tends to diminish this, but in the upright posture exercising muscles provide extravascular compression and maintain venous return. Cardiac output in the normal heart is very responsive to venous return (the Frank-Starling curve), but the failing heart is less sensitive owing to a flatter ventricular function curve (Fig. 12.5).

465

FIG. 12.6 The relationship of LV volume to pressure in a normal cardiac cycle (A-B-C-D) At A systole begins. The pressure rises isovolumetrically until at B the aortic valve opens. The ventricle ejects to C and then isovolumic relaxation occurs until D at which point filling occurs. The dotted blue line shows the relationship in a failing ventricle with reduced inotropy and increased diastolic stiffness. Following the cardiac cycle A-B-C-D it can be seen that a smaller stroke volume is generated (ABCD). Increasing inotropy (dotted red line) would restore or increase stroke volume.

FIG. 12.5 Relationship between ventricular output and filling pressure A The black line shows a normal right ventricular (RV) function curve, where an RV output of 7 L/min (dotted line) is achieved with a filling pressure of 1 mmHg. The left ventricle (LV) must produce the same output (A1), so that the normal LV (blue line) requires a filling pressure of 8 mmHg. If the LV is failing (LVF, light blue line), a filling pressure of 18 mmHg (A2) is required to maintain this output. B A similar relationship is shown for the normal (black line) and failing (grey line) right ventricle.

An increase in intrapericardial pressure with the accumulation of an effusion impedes venous return and cardiac filling and, with marked effusion, a low stroke volume and cardiovascular collapse follow, a condition called tamponade (p. 587). Some drugs affect venous tone: for example, some of the therapeutic benefits of nitrates are related to venodilatation, lowering wall stress and thus reducing myocardial oxygen demand. Normal ventricular filling occurs principally in early diastole, with atrial systole giving a final 'top-up'. The latter may be particularly important in hypertrophied hearts, which fill poorly in early diastole. Here atrial systole provides a much greater fraction of ventricular filling. In these

1

466

MCQ 12.1

patients, loss of atrial contraction (due to atrial fibrillation) may lead to marked reduction in stroke volume. Afterload During systole the ventricles eject blood once the diastolic pressure in the arterial circuits is exceeded. Afterload is the load that affects the ventricles during ejection. This can be equated with the peripheral vascular resistance, which, together with vessel wall elasticity, determines the diastolic arterial pressure. Increasing afterloads (Fig. 12.6) decrease cardiac output, and this relationship is worse in failing hearts. The treatment of patients with impaired ventricular performance now includes the use of drugs that reduce the peripheral vascular resistance and hence reduce afterload (p. 496). Rate and rhythm Changes in the rate of stimulation of the heart muscle can alter contractile performance, independently of changes in muscle length (or preload). These mechanisms affect output to a lesser degree than preload and afterload. However, increases in heart rate may also change cardiac output by altering the time available for cardiac filling and coronary flow. 1

Ventricular dilatation The cardiac chambers dilate either with abnormal volume loads or as a consequence of impaired myocyte function,

as in cardiomyopathy. The larger-volume ventricle is an adaptive response, allowing the maintenance of stroke volume despite a reduction in the shortening capability of the myofibrils. However, although the response is initially adaptive, there is a cost. From the Laplace equation, a dilated ventricle with an increased diameter can only generate the same pressure with an increased wall tension. This increased wall tension can only be sustained by an increase in oxygen consumption, which may outrun the supply, leading to further ischaemia and dilatation.

Ventricular hypertrophy Cardiac myocytes hypertrophy in response to: • Loss of adjacent myocytes. • An increase in ventricular wall stress - the resultant increase in wall thickness effectively reduces the stress each sarcomere has to generate. With mild hypertrophy, the capillaries increase in density proportionally to the hypertrophy, but beyond a certain extent the capillary growth lags behind the hypertrophic process, thus rendering the less well supplied subendocardium liable to ischaemia. The hypertrophic process, accompanied by deposition of more collagen fibres, leads to reduced compliance and impaired diastolic function. • Cellular mechanisms of cardiac adaptation. The adult cardiac myocyte is not capable of cell division but remains synthetically active, regenerating contractile proteins every 30-90 days. Under conditions of increased stress the myocyte is stimulated by a large number of factors via sarcolemmal receptors. These include growth hormone, angiotensin II, noradrenaline (norepinephrine), tumour necrosis factor-a, interleukins, growth factors and nitric oxide, among others. Acting via intracellular second messengers, the cells revert to a more fetal phenotype, resulting in cell growth and hypertrophy, often accompanied by impaired contractile function.

SYMPTOMS AND SIGNS OF

HEART DISEASE Symptoms arising from the heart are not specific and it often requires a careful history to differentiate them from similar complaints arising from chest disease, dyspepsia or anxiety. Chest pain, breathlessness, oedema, palpitations, dizziness, syncope and fatigue can all be caused by heart disease. Exercise capacity is limited in most forms of heart disease, and the limiting symptom may be obvious, such as pain and breathlessness, or much more subtle, such as excessive exhaustion. The degree to which symptoms have begun to limit normal activities requires careful definition, as patients may not always volunteer the information, wrongly thinking that age alone is enough to account for their increasing inability to exert themselves without dyspnoea or dizziness. Others may prevent angina or

breathlessness by greatly restricting their lives. In assessing the effectiveness of treatment it is useful to have an idea of the activities that can and cannot be performed, using relevant examples from the patients' own lives, the ease with which they can walk and talk, or how they manage household chores. Improvements in patients' performance may then be judged by their own yardstick. Patients sometimes need considerable encouragement to re-establish more normal patterns of activity after being treated.

12

Chest pain Chest pain arising from the heart and great vessels may be caused by cardiac ischaemia, pericardial inflammation, aortic dissection or massive pulmonary embolism. The history may allow differentiation between the various possibilities, but the severity of the symptom has virtually no relationship to the potential severity of the underlying cause. Many patients suffer pains in the chest which are probably cardiac in origin but for which no adequate explanation is possible; for example, some patients may describe abnormal sensations produced by ectopic beats as pain. Cardiac ischaemic pain In a typical case the discomfort associated with myocardial ischaemia is described as a pressure or tightness in the chest, which may also be felt in the throat, producing the choking feeling that led to the symptom being called angina pectoris. The characteristics of angina include common patterns of radiation to the arms, more commonly the left, the jaw and teeth and, less commonly, through to the back. The pain, and its sites of radiation, are often reproducible, but incomplete or mild attacks may not have the same full distribution of the more severe episodes. In angina, which represents episodes of reversible ischaemia, the precipitating causes are typically those that will increase myocardial oxygen demand beyond the coronary supply. The patient may report discomfort only when having to run, or when they try to walk in the face of a cold wind or after a heavy meal. Emotional upsets are powerful precipitants of angina in some patients, even without physical exertion. The discomfort may be very severe and be associated with frightening feelings of impending death. It can also be a mild discomfort or ache in the chest, which may be thought trivial and ignored. Many patients do not experience their angina as a pain, and feel it as a faint constriction in the chest. Some attacks of myocardial ischaemia are unaccompanied by any discomfort (silent ischaemia}. Angina usually disappears fairly rapidly if the patient rests, or takes glyceryl trinitrate (GTN). Angina builds up over several seconds and is usually not greatly influenced by posture, unlike musculoskeletal chest pain (which may also occur with exertion). Although similar in nature and radiation, the pain of myocardial infarction

467

lasts much longer than angina. It is more intense and does not pass off with rest or GTN. In an attack, a patient with cardiac ischaemia usually looks pale and sweaty, unlike someone with an attack of indigestion, who often appears flushed. Other causes of chest pain Pericardial pain (p. 587) can be easily confused with the pain of myocardial infarction or angina, but is often influenced by posture, sometimes being relieved or worsened by leaning forwards, and aggravated by swallowing. It is more often described as a burning or dull pain, and is less likely to be associated with breathlessness than is myocardial ischaemia. The pain of dissection of the ascending aorta (p. 600) is a severe, tearing pain, often starting suddenly, usually felt retrosternally at first, and sometimes radiating through to the back or to the left shoulder. Massive pulmonary embolism can also produce a retrosternal constricting discomfort indistinguishable from myocardial ischaemia; smaller peripheral emboli may be responsible for more pleuritic-type pains (p. 681). Oesophageal spasm due to reflux can produce a severe retrosternal pain which may be confused with that of myocardial ischaemia, and may be relieved by GTN. The pain of oesophageal rupture can be confused with myocardial infarction. Sharp, stabbing precordial pains are common in highly strung patients. They may have a muscular origin, but occasionally coincide with the accentuated postectopic contraction of the heart, which is perceived by the patient as a knife-like pain. 1

Dyspnoea or shortness of breath This symptom is a feeling of laboured, or unnaturally difficult, breathing. In heart failure it is due to the lungs becoming stiff and difficult to ventilate, owing to the rise in pulmonary venous pressure. The reduced cardiac output limits the oxygen-carrying capacity of the circulation, precipitating anaerobic metabolism. The acidosis produced causes increased ventilation, which persists for longer than normal after exertion. Thus, even minor exertion produces disproportionate and prolonged dyspnoea. Finally, in severe left heart failure with pulmonary oedema, arterial oxygen desaturation adds to the distress. Dyspnoea (or angina) can be categorized using the New York Heart Association scale (Table 12.1). Severe forms of left heart failure are associated with dyspnoea at rest, worse when supine. This degree of pulmonary venous hypertension is commonly accompanied by a dry, nonproductive cough. Dyspnoea at night can be graded on

1

468

MCQ 12.2

TABLE 12.1 New York Heart Association grading of dyspnoea or angina Grade

Criteria

1

Symptoms only occur on severe exertion. Almost normal lifestyle possible Symptoms occur on moderate exertion. Patient has to avoid certain situations, such as carrying shopping up several flights of stairs Symptoms occur on mild exertion. Activity is markedly restricted Symptoms occur frequently, even at rest

2 3 4

the basis of the number of pillows required to enable the patient to sleep (orthopnoea). Patients with left heart failure may be woken from their sleep by severe dyspnoea resulting from pulmonary congestion. This passes off on sitting up, or walking about slowly, and is known as paroxysmal nocturnal dyspnoea (PND). This may be the first symptom of left heart failure and may be difficult to distinguish from nocturnal asthma, as some patients with left heart failure develop a considerable degree of wheeze with bronchospasm.

Palpitations A careful interpretation of what the patient means by the word 'palpitations' is essential. It may be simply an awareness of the normal heart beat, which may be more forceful or faster than usual because of anxiety. It may, however, be indicative of a serious cardiac arrhythmia. It may be useful to ask the patient to tap on the table to indicate the rate and rhythm during the attack. It is helpful to determine the following: • Onset and cessation of the attacks. Is it sudden, with a thump (ectopic), or gradual? • Are there any symptoms to suggest cardiac decompensation (failure) during the attacks? • Was there irregular thumping (ectopics or atrial fibrillation)?

• Was there chest pain (the occurrence of angina implies a very fast and potentially dangerous heart rate)? • Has there been any fainting or near fainting with the attacks? True syncope indicates an urgent need for investigation, as the arrhythmia is potentially fatal. Polyuria sometimes occurs on cessation of supraventricular tachycardias, but is not usually described by the patient unless a leading question is asked.

Syncope and presyncope (dizziness) Sudden collapse with loss of consciousness during exertion almost always has a cardiac cause. It is due either to an obstruction to outflow from the left or right ventricle, which prevents an adequate increase in cardiac output on exertion, or to a cardiac arrhythmia induced by the

exertion. Both may result in a sudden fall in blood pressure and cerebral perfusion. Some mental confusion due to cerebral anoxia is common. Syncope after exertion is not uncommon and may simply be due to blood pooling in the legs and a poor venous return. However, syncope under other circumstances may be cardiac, and an eyewitness account is invaluable. Cardiogenic syncope usually occurs without warning; the collapsed patient is vasoconstricted and grey, and the pulse is either absent or very slow. In a classic 'Stokes-Adams' attack (p. 503) the unconscious patient blushes dramatically with the return of circulation and subsequent return of consciousness. Syncope may also occur in severe tachycardia. A prolonged period of unconsciousness is unlikely to be cardiogenic, as cardiac asystole or a malignant cardiac arrhythmia sufficiently rapid to cause unconsciousness is unlikely to last long without proving fatal. Other forms of syncope 'Vasovagal' syncope. The common faint, often triggered by pain or an unpleasant sight, gastrointestinal upset, haemorrhage or pyrexial illness, is a vagal phenomenon, and the syncopal episode is followed by profuse cold sweat, a feeling of sickness or actual vomiting, and bradycardia. Micturition syncope and cough syncope are rarely cardiogenic, but are probably triggered by a Valsalva manoeuvre, the first when initiating micturition with a full bladder and the latter after repeated bouts of coughing, which inhibit venous return. Presyncope or dizziness is a common and much less well defined symptom, and has a variety of causes as well as heart disease. It may result from a cardiac arrhythmia, but even then is much more common in elderly patients with associated cerebrovascular disease.

Oedema Right heart failure causes pitting oedema of the feet and legs, worse at the end of the day and relieved by rest and elevating the legs. Unlike other causes of oedema it may be associated with other symptoms of heart failure, such as dyspnoea and fatigue. The oedema may spread to the thighs, abdominal wall, sacrum and back. There may be ascites and hepatic congestion, with abdominal distension. The hepatic congestion may be worse on exertion, with pain over the liver (usually epigastric) during exercise and for several minutes afterwards - hepatic angina. The differential diagnosis of cardiac oedema includes fluid retention from other causes, such as nephrotic syndrome and cirrhosis of the liver, oedema of one or both legs from venous insufficiency, or lymphatic insufficiency.

Fatigue Fatigue is a non-specific and often neglected symptom. It is, however, a very real feature of heart disease, a result of the reduced effort tolerance and lactic acidosis produced by anaerobic muscle metabolism and changes in the skele-

tal muscle. It is a prominent symptom in low cardiac output states, e.g. severe left ventricular failure.

12

EXAMINATION OF THE CARDIOVASCULAR SYSTEM The general examination Traditionally, and ideally, the cardiac examination is undertaken in a warm, quiet and well-lit room with the patient se'ated on the bed or couch with the upper body supported and at an angle of 45° to the horizontal. The general appearance of the patient and his or her comfort and general wellbeing or illness is noted. Certain cardiac diseases are associated with anormalities of habitus, specifically Marfan's syndrome (p. 526), where aortic dissection and mitral valve disease are to be expected, or other chromosomal abnormalities that produce associated congenital cardiac lesions, such as Down's and Turner's syndromes (p. 535). The 'mitral fades' A thin face with prominent 'rosy' cheeks, which on closer examination are tinged with blue (cyanosis), used commonly to be seen as a consequence of chronic mitral stenosis, but now is more frequently associated with any condition where the cardiac output is low and the venous pressure high. Skin - pallor As in any medical examination a note is made of pallor, as symptoms indistinguishable from those of cardiac failure and even angina may be caused by profound anaemia. A low cardiac output may itself produce generalized pallor, often associated with cool fingers, toes and nose. Skin - cyanosis Peripheral cyanosis can also be produced by low cardiac output, when the cool extremities take on the characteristic blue tinge. Central cyanosis is caused by intracardiac shunting (right to left heart, as in the Eisenmenger reaction, see p. 536); the lips and tongue are blue and the colouring made more intense by the frequently associated polycythaemia. The periphery is often warm in these circumstances, and in congenital cardiac defects clubbing of the finger- and toenails is found. Skin - hands and nails Clubbing may occur in cyanotic congenital heart disease, but may be acquired in chronic infective endocarditis (p. 576). Infective endocarditis also accounts for several other findings, including splinter haemorrhages, which appear as small dark subungual petechiae. They are also seen in other diseases, such as polyarteritis nodosa. The appearance of new splinter haemorrhages during an acute episode of pyrexia in a patient with a murmur strongly suggests endocarditis or persistent infection in an established treated case. Osier's nodes are tender palpable spots on the pulps of the fingers occasionally detected in endocarditis.

469

Skin - xanthomata and arcus senilis Characteristic firm nodules form on the tendons of the knuckles and over the elbow, and thicken the Achilles tendon in hypercholesteraemia. 1 Indeed, these xanthomata might be the first sign of familial hypercholesterolaemia (p. 1023) and bring the patient to the doctor before there are manifestations of coronary heart disease. More commonly there is arcus senilis, which is a pale ring, usually incomplete, encircling the iris in young persons with hyperlipidaemia; the ring is commonly present in older people even without any lipid disorder. In hyperlipidaemia the periorbital area may also develop flat, mildly raised, pale yellow xanthomata.

Arterial pulses The information sought in examining the arterial pulses includes the following:

Mouth Poor dentition and recent major dental work are documented sources of infection in endocarditis, so an assessment of the oral hygiene of patients is essential.

The pulse volume rate and rhythm are assessed from the right radial pulse. The contralateral radial should also be briefly examined to confirm its presence and volume. Some indication of the state of the arterial wall can be obtained from the thickness of the radial and brachial arteries. Next the carotid pulses are examined, where the pulse character can be appreciated. At the carotid pulses thrills may be felt and bruits heard. The same examination is made of the femoral arteries, where in addition the pulse is simultaneously felt in the radial and femoral arteries. Coarctation of the aorta (p. 540) produces either absence of the femoral pulse or a reduced volume beat, which is delayed in comparison to the radial pulse upstroke so-called radiofemoral delay. In the legs the popliteal arteries and both dorsalis pedis and posterior tibial arteries are examined.

Dyspnoea Movement, dressing and undressing may reveal breathlessness, and orthopnoeic patients may not be able to tolerate lying flat. Oedema Pitting oedema is classically seen in the dependent regions: the ankles and feet in ambulant patients, 2 and the sacral area in those in bed. The oedema may be extreme and involve the thighs, pudenda and lower abdominal wall, when a 'peau d'orange' effect is seen. In these circumstances ascites is nearly always present, but may surprisingly occur before much peripheral oedema has developed. Cardiac cachexia Although weight gain due to fluid accumulation is usual in heart failure, late stages of the severe disease may produce profound muscle wasting and weight loss closely analogous to disseminated malignancy; this condition is called cardiac cachexia.

Arterial pulses and blood pressure Blood pressure A mercury sphygmomanometer is usually used to measure the systemic arterial blood pressure. This is often performed early in the examination; however, blood pressure is often highest at this time, owing to an 'alerting' response in the patient, and a lower reading may be obtained at the end of the consultation. The room should be warm and the measurement repeated in both arms and again with the patient standing. Details of blood pressure measurement are to be found on page 590.

1 470

Figs 12.1-12.3

2

Fig. 12.4

• Presence and equality of volume of the pulses on both sides of the body • Rate, rhythm and character of the pulse waveform • The state of the arterial wall • The presence of arterial bruits • The carotid pulse, which can help to time events when auscultating the heart (S1 just precedes the upstroke of the pulse) • The presence or absence of radiofemoral delay.

Rate and rhythm It is important to count the pulse accurately when there is a tachycardia (pulse >100) or a bradycardia (pulse <60). Heart rates of 150 bpm (±5bpm) are almost always due to atrial flutter. A small variation of heart rate with respiration is normal, and more marked in young people (sinus arrhythmia). Intermittent irregularities in the pulse are often due to ectopic activity, producing a slight pause in the normal regular beat followed by a beat of larger than usual volume. A pulse which is totally irregular with constantly varying volumes is likely to be due to atrial fibrillation (AF) and is called the irregularly irregular pulse. Frequent ectopics can cause confusion with AF, and the irregular rhythm may be missed if the fibrillation is well controlled owing to therapy. The heart rate is best assessed by auscultation when the pulse is irregular, although the use of the difference in pulse rate measured at the wrist and by auscultation at the cardiac apex (the apex-radial deficit) is of limited practical value as there is always a difference in the presence of a fast apical heart rate in AF. Pulse amplitude The pulse amplitude at the wrist depends on the stroke volume, which varies with cardiac output and pulse interval. It also depends on the state of vasodilatation of the skin of the hands, with a smaller radial pulse volume when cold than when warm. A thready, low-volume and rapid

pulse may be felt in shock due to haemorrhage, and a largevolume pulse in thyrotoxicosis or CO2 retention in pulmonary disease. Pulse character With the exception of collapsing pulses, this is most reliably assessed by examining the carotid pulses in the neck, although the femoral and brachial arteries are useful when the neck is difficult to examine. The collapsing pulse. This is a sign of aortic regurgitation but may be present in any condition that produces the combination of a large pulse volume with a rapid run-off into the periphery owing to a reduced impedance to flow (e.g. pyrexia, pregnancy, thyrotoxicosis, arteriovenous fistula). The sign is best appreciated by gripping the forearm with the left hand, the examiner's fingers being pressed fairly firmly against the medial aspect of the forearm. The arm is then raised up by the examiner's right hand, which is holding the patient's hand in a handshake position. In the presence of a collapsing pulse the fingers will feel a 'knock' in the forearm muscles. Associated features are a low diastolic blood pressure reading and, in severe cases, visible arterial pulsations in the neck, Corrigan's sign. The slow rising, anacrotic or plateau pulse. This sign of aortic stenosis is usually most reliably found by examining the carotid pulses. The carotids are best felt with the patient sitting at 45°, resting the head on a pillow and looking straight ahead. The pulse is felt halfway up and medial to the sternocleidomastoid muscle at the level of the major skin crease. The slow rising pulse contrasts with the transient flick that the normal carotid systolic pulse produces. It can be a difficult physical sign to appreciate and, as its name implies, the pulse may be felt to caress the fingertips as it rises slowly to a more sustained peak. It is often of low volume and may have an associated thrill or judder. It may be masked in stiff arteries of the elderly. The jerky pulse. This is felt at the carotids as described, and is one of the few physical signs to be found in hypertrophic cardiomyopathy (p. 571). It is therefore potentially important, as it should alert the physician to the possibility of the presence of this serious condition affecting the young. Severe mitral regurgitation can produce the same physical sign. Rarer forms of abnormal arterial pulse. Pulsus alternans consists of a beat-to-beat alternating large and small pulse volume with a regular rhythm, and is seen in severe left heart failure. Pulsus paradoxus represents an accentuation of a normal physiological variation in the pulse volume with respiration. It consists of a greater than l0mmHg fall in systolic pulse pressure with inspiration and occurs in pericardial tamponade or constriction (pp. 587 and 588), but may be seen in cardiomyopathy, massive pulmonary embolism, and sometimes with normal cardiac function in severe asthma. It may be felt by the fingers on the radial pulse but, if mild, is better elicited by using a sphygmomanometer and detecting the different systolic pressures during inspiration and expiration. Pulsus bisfe-

riens is a pulse that can be felt to have a double systolic peak. It can be a feature of mixed aortic valve disease, where a degree of stenosis complicates dominant aortic regurgitation.

12

The jugular venous pulse (JVP) The JVP can be a difficult physical sign to appreciate, and the subtleties of its waveforms have fascinated physicians and terrorized medical students for generations. Despite its archaic nature the JVP remains a useful bedside indicator of circulating blood volume, it being invisible in fluid depletion or haemorrhage and raised in fluid overload or heart failure. Its waveforms can indicate the nature of some cardiac arrhythmias and make the diagnosis of pericardial tamponade. The internal jugular is used; despite its being deep in the tissues under the sternocleidomastoid muscle, pulsations arising from this vein can be seen at the skin surface. The height of the JVP is measured vertically from the sternal angle; normally, in a patient seated at 45° it does not exceed 3cm. The normal waveform can be appreciated in most people at the root of the neck, but they may have to be lying rather flatter than at 45°. What is seen is a regular pulsation in the neck, with two peaks and two troughs for each radial pulse (if sinus rhythm is present). Unlike arterial pulses, the venous pulse is most obvious by the rhythmic inward movement of the skin overlying the vein; arterial pulses produce a dominant outward, brisk skin movement. Inspiration causes a fall in the JVP (Table 12.2). The JVP can be so high that the neck looks generally full and the upper limit of the venous column is not appreciated; sitting the patient upright or making him stand may make the top of the JVP column come into view. A low JVP can be made to appear by gentle abdominal pressure - hepatojugular reflux: in a normal person the JVP may rise 2-3 cm, but rapidly falls, whereas in cardiac failure the bolus of extra blood supplied to the right heart by squashing the splanchnic bed cannot be accommodated, so the JVP rises, and this is sustained for several seconds. When

TABLE 12.2 Differentiation of the JVP from an arterial (carotid) pulse

Waveform Pressing at root of neck? Variation with respiration? Palpable pulse? Change with posture? Pressure on abdomen (hepatojugular reflux)

Venous

Arterial

Two peaks per beat (in SR) Obliterates wave form Normal JVP falls with inspiration Not usually, except in tricuspid reflux May fall if patient sits up or stands JVP rises

Single peak No effect No effect Yes No change No effect

471

• • • •

FIG. 12.7 Relationship of venous waveform to heart sounds The a wave of atrial contraction is followed by the x descent. The cwave is not usually visible. The I'wave coincides with ventricular systole.

the JVP is massively raised but does not show waveforms during the cardiac cycle, it is likely to be owing to superior vena caval obstruction, such as in malignancy. The JVP waveform The two peaks, a and v, are due to atrial and ventricular systole (Fig. 12.7). The two descents are x and y, with a transient pressure 'hump' on the first (x) descent being called the c wave. The a wave is generated by atrial systole, and therefore disappears in atrial fibrillation. Extra-large a waves are called cannon waves and are seen when atrial systole occurs with a closed tricuspid valve; this occurs intermittently in complete heart block, occasionally with ventricular ectopics, and regularly in nodal rhythm. Large, or giant, a waves are associated with a poorly compliant right ventricle increasing the impedance against which the atrium has to eject blood, such as in pulmonary hypertension. When the tricuspid valve leaks, the right ventricle is in communication with the great veins and the rise in pressure during systole produces a very large v wave, sometimes called the s wave. Abnormalities of the x and y descents can be difficult to time correctly. The y descent is slow in tricuspid stenosis, which is a rare valve lesion. In pericardial tamponade and constriction (pp. 587 and 588) the JVP is very high and demonstrates an abrupt fall in systole (x descent); complicated by myocardial involvement the dominant descent may occur in the early part of diastole (y descent). 1

The precordium Scars and deformities may give useful clues to the past history; the former might indicate a previous closed mitral valvotomy (generally a subtle left-sided inframammary thoracotomy scar) and the latter congenital heart defects. Pulsations of the apex beat or dyskinetic movements of a left ventricular aneurysm may be seen. Palpation is undertaken to detect the following: • Presence of thrills - these are palpable murmurs felt as high-frequency vibrations 1

472

MCQ 12.3

The position of the apex beat The character of the apex beat The presence of a parasternal heave Palpable heart sounds.

Thrills The commonest thrills are those arising from the aortic valve in aortic stenosis, and are felt either side of the sternum in the upper part of the anterior chest wall. Diastolic thrills down the sternal edge and at the apex can occasionally be felt in very severe aortic regurgitation; mid to low central sternal thrills may occur in small ventricular septal defects. The apex beat This is defined as the lowest, most lateral detectable cardiac pulsation; its position is normally in the fourth intercostal space at or medial to the midclavicular line, and is felt when the patient is sitting in the standard 45° position. Displacement of the apex downwards and laterally is due to left ventricular dilatation. Many adults, especially those who are obese or have overexpanded lungs, do not have palpable apex beats; in a very small minority this will be due to dextrocardia. The character of the apex beat can be determined in most cases by palpating the apical region with the patient in the left lateral position. It is generally a very localized pulsation covering no more than the pulp of two examining fingers. A thrusting apex beat pushes the fingers outwards with each systole and is a feature of left ventricular hypertrophy. In aortic stenosis the apex may be thrusting and not displaced. In a volume-overloaded ventricle, such as in mitral regurgitation, the apex beat is usually displaced and may be thrusting in nature. A tapping apex which is not displaced is a feature of mitral stenosis and is caused by a loud (palpable) first heart sound. Dyskinesia of the apex may affect larger portions of the precordium and are associated with ventricular aneurysms. The whole apex may then rock under the palpating hand, and the apex beat may have a double impulse. This latter sign may also be felt in hypertrophic cardiomyopathy. The apex beat is abnormal in dilated poorly functioning ventricles when it may be very diffuse, with a weak but widespread pulsation. Parasternal heave This is felt by applying the heel of the hand to the lower third of the sternum with modest pressure. Right ventricular hypertrophy produces a feeling of lifting of the hand during systole. Similar lifting sensations, slightly more to the left of the sternum, can be produced by the rapid expansion of the left atrium in systole due to severe mitral regurgitation.

Auscultation Successful cardiac auscultation requires a comfortable

and familiar stethoscope of good quality and a relatively quiet environment. The bell is designed to accentuate low-frequency sounds and is therefore particularly suited to detect third and fourth heart sounds, as well as the diastolic murmur of mitral stenosis, but it can perform more like the diaphragm if pressed hard against the skin, a feature used to advantage on occasion (see below). The diaphragm filters out low-frequency sounds and is therefore most suitable for listening to the first and second heart sounds, as well as high-frequency murmurs such as that produced by aortic regurgitation. The aim of cardiac auscultation is to detect the following:

12

• Normality of the first (S1) and second heart sounds (S2), loudness and splitting - diaphragm; • Presence of added third (S3) and fourth (S4) heart sounds - bell, lightly applied; • Presence of clicks - diaphragm; • Presence of murmurs, their timing, location, intensity and radiation - both diaphragm and bell; • Other noises, such as rubs and venous hums - both diaphragm and bell. A suitable examination begins with the diaphragm applied to the upper left sternal edge to determine the first and second heart sounds - a finger or thumb on a carotid pulse will aid the timing. The diaphragm may then be moved over the chest, down the sternal edge to the apex; after listening to both S1 and S2 here, the change is made to the bell and the process repeated in reverse. At this point the patient can be moved into the positions necessary to bring out murmurs (see below). The relationship between the heart sounds and various types of murmurs is illustrated in Figure 12.8A. Heart sounds First and second heart sounds (SI and S2) The first heart sound slightly precedes the carotid pulse upstroke. It signals the beginning of systole (Fig. 12.9) and, although complex in origin, is produced mainly by the closure of the mitral and tricuspid valves. Normally, only the mitral component is audible, but delayed systolic events in the right heart (e.g. in right bundle branch block, or volume overload in atrial septal defect) may allow splitting of S1 to be heard (mitral followed by tricuspid closure). A loud S1 is heard in mitral stenosis (p. 522) and in situations where atrial systole is very close to ventricular systole, such as with a short PR interval; conversely, mitral regurgitation and long PR intervals are associated with a quiet S1. The second heart sound is caused by closure of the aortic and pulmonary valves. Audible splitting of S2 (aortic valve A2 followed by pulmonary valve P2) occurs in most normal people during inspiration and can be heard with the diaphragm applied to the upper left sternum (pulmonary area), but not at the apex where P2 does not

FIG. 12.8 Heart murmurs A Diagrammatic representation of murmurs. S1 and S2 = first and second heart sounds; A2 = aortic valve closure sound; P2 = pulmonary valve closure sound; OS = opening snap. B The precordial areas where murmurs are usually heard maximally. A = aortic; P = pulmonary; M = mitral.

radiate unless it is abnormally loud, e.g. in pulmonary hypertension. Third and fourth heart sounds (S3 and S4) Both these sounds are produced by ventricular filling, are very low pitched and require the bell to be lightly applied to be heard. It can be very difficult to tell them apart, particularly at fast heart rates. The third sound occurs early in diastole and can be heard in normal young

473

An opening snap is characteristic of mitral stenosis, where the valve remains compliant. It occurs early in diastole owing to the sudden tensioning of the mitral valve apparatus and can be differentiated from an S3 by its higher pitch and sometimes wide radiation. An opening snap is audible with diaphragm and bell, S3 only with the bell. Pericardial rubs are variously heard as creaky rustling, shuffling noises which can be differentiated from pleural rubs by their continuation even during breath-holding. Tumour plops can uncommonly be associated with a large atrial myxoma, when a curious 'plop' may be produced intermittently, and sometimes with a change in posture - usually as the patient takes up the left lateral position.

FIG. 12.9 Relationship between pressures in the aorta, left ventricle and left atrium The timing of the heart sounds (S1 and S2) are shown (a and v represent the a and v waves).

people and some slim athletic adults, who have a relative bradycardia. It is also heard in hyperdynamic states and volume-overloaded ventricles, such as with thyrotoxicosis and significant mitral regurgitation, respectively. In older people an S3 is taken as indicative of heart failure. A higher-pitched and loud S3, sometimes called a 'pericardial knock' is produced by the very abrupt end to ventricular filling that occurs early in diastole owing to pericardial constriction. The fourth heart sound is generated by atrial systole in the presence of a raised left ventricular enddiastolic pressure. It indicates an impairment of ventricular compliance, often due to hypertrophy. It is typically a feature of hypertension, but can also be heard during myocardial ischaemia. Gallop or triple rhythm In heart failure, if S3 and S4 are loud, a characteristic cadence of the heart sounds is heard which sounds like a galloping horse. A loud combination of the sounds leads to the summation gallop, which may be heard when a tachycardia complicates the cardiac failure. Other added sounds Ejection clicks are very high pitched and sound as their name implies. They are a feature of young patients with pulmonary and aortic valve stenosis when there is a pliable valve which domes upwards in systole. Midsystolic or non-ejection clicks may be single or multiple and are a feature of floppy or prolapsing mitral valves (p. 525), and much less commonly of tricuspid valves.

1

474

MCQ 12.4

Heart murmurs Murmurs are found where there is turbulent flow, and their radiation follows the direction of blood flow. Although the precordium is classically divided into various valve areas (Fig. 12.8B) they may not always be the sites where murmurs arising from those particular valves are best heard. Murmurs may be palpable and produce precordial thrills. Their full characterization and clinical importance may depend more on the associated physical signs elicited on examination than purely on their ausculatory features. Thus a loud ejection systolic murmur in a patient with a normal apex beat and carotid pulse is of less concern than a quieter but similar murmur associated with a displaced, thrusting apex and a slow carotid pulse upstroke. The former may be a benign flow murmur, the latter represent critical aortic stenosis. Causes of murmurs Murmurs may be generated by a high-velocity jet, such as a stenotic or regurgitant jet, passing from a high-pressure to a lower-pressure chamber or vessel. They may occur with increased flow velocity in a normal vessel, or with flow into a dilated or distorted vessel, such as the aorta in hypertension or the pulmonary artery in sternal depression. Murmurs do not therefore always indicate valve pathology. They may indicate a high flow, such as in pregnancy, or a minor anatomical distortion not amounting to true pathology, so-called innocent murmurs. Description of murmurs Murmurs are described in terms of their timing, precordial location, radiation, relative intensity and, if warranted, the influence of respiration and posture (Fig. 12.8). It is possible to grade the loudness of murmurs from 1/6 (almost inaudible) to 6/6 (a murmur which can be heard in a quiet room without using a stethoscope). These grades are, however, of limited value, as they are highly subjective and the loudness of a murmur varies considerably depending on circumstances. The length of a diastolic murmur can be expressed on a scale of quarters, 1/4 being very short, and 3/4 an almost full-length diastolic murmur.

Systolic murmurs Ejection murmurs are crescendo-decrescendo (diamondshaped) (Fig. 12.8) and are characteristic of flow across a stenotic aortic or pulmonary valve, or flow into a dilated aorta or pulmonary artery. They may also occur with increased flow across these valves (innocent murmur). Sometimes, in severe stenoses it is impossible to distinguish an ejection from a pansystolic murmur. The associated signs will often help. Pansystolic murmurs are generated by jets passing from a high-pressure to a low-pressure chamber throughout systole, including the normally isovolumetric phases of contraction and relaxation. The best example is mitral regurgitation, but tricuspid regurgitation and ventricular septal defects produce similar murmurs. The murmur obscures both S1 and S2 and is of almost even intensity throughout systole. Late systolic murmurs. Mitral valve prolapse and papillary muscle dysfunction give rise to a high-pitched murmur of even intensity starting halfway through systole, often preceded by a midsystolic click. Posterior papillary muscle dysfunction can give rise to a mitral pansystolic murmur, which may be loud in the aortic area. Diastolic murmurs Diastolic murmurs are always pathological, but are often much harder to detect than systolic murmurs. Mitral and tricuspid mid-diastolic murmurs. The mitral (and tricuspid) stenotic jets induce a low-frequency, largely subsonic vibration. These are heard as low-pitched, rumbling murmurs starting after the opening snap, best heard with the bell. They are localized and are often missed by the inexperienced, who are expecting to hear a much higher-pitched sound. The mitral diastolic murmur is accentuated by the patient lying on the left. If short, it can be confused with a third heart sound. Early diastolic murmurs start immediately after the second heart sound. They are high-pitched noises, which can be very short or persist throughout diastole. Aortic diastolic murmurs are often faint and are best heard with the patient leaning forwards in expiration. A soft aortic diastolic murmur is often missed. Other types of murmur Musical or mewing murmurs may occur in systole or diastole if a valve cusp adjacent to a jet vibrates at a characteristic frequency. These murmurs can be very loud, but the jet causing them may be small. A mewing murmur is characteristic of a small hole in an aortic valve cusp as a complication of endocarditis. 'Innocent' murmurs Many midsystolic and ejection systolic murmurs are unaccompanied by any signs or features of heart disease. These murmurs are common in children and in pregnancy. They are maximal in the pulmonary area, but may also be heard along the left sternal edge and at the apex, and often have a low-pitched vibratory quality. They are usually of low

intensity. If there is doubt as to their 'innocent' nature, an echocardiogram and, particularly, a Doppler study can rule out a small ventricular septal defect, pulmonary stenosis or unsuspected trivial mitral regurgitation.

12

Accentuating murmurs - dynamic auscultation Unless murmurs are perfectly well heard with the patient sitting at 45°, all patients should be examined in the left lateral position (mitral murmurs) and sitting forward in exhalation (which accentuates the early diastolic murmurs of aortic regurgitation). Exercise, such as sit-ups, can be used to increase cardiac output and so increase the loudness of mitral stenotic murmurs. Increased venous return, with inspiration or passive leg raising, will transiently increase the loudness of many right heart murmurs. In hypertrophic cardiomyopathy the LV outflow tract murmur can be diminished by asking the patient to squat, whereas in mitral valve prolapse this manoeuvre will increase the murmur loudness. 1

Concluding a cardiovascular examination The examination is incomplete without examination of the lung bases for signs of cardiac failure, the sacrum and ankles for oedema, and assessment of the peripheral pulses, dentition, skin and nails. • The lungs are examined not only for signs of failure, but also because chronic lung disease can complicate and add to the symptoms and signs generated by heart disease. • A patient with angina may well have atheromatous peripheral vascular disease, shown by loss of pulses in the legs. Examination of the carotid arteries for signs of atheromatous carotid stenosis may reveal loss of a carotid pulsation, a carotid bruit or loud systolic murmur. Atrial fibrillation may have caused embolic loss of some pulses. If the patient is hypertensive with a systolic murmur, radial and femoral pulses should be felt to detect any delay and weakness of the femoral pulse indicating coarctation. Possible aortic aneurysm should lead to careful assessment of the pulses and the blood pressure in both arms. • Tricuspid regurgitation is confirmed by finding a pulsatile liver. If the patient has hypertension, the abdomen is examined for renal enlargement or bruits. A patient in right heart failure may well have ascites and considerable liver enlargement. • Atrial fibrillation or tachycardia may be caused by thyrotoxicosis, so the thyroid is examined for goitre and signs of thyroid overactivity. • If the patient is hypertensive, the optic fundi are examined for hypertensive retinopathy. They should be examined regularly in patients suspected of having endocarditis.

475

FIG. 12.10 Normal ECG

CARDIAC INVESTIGATION THE ELECTROCARDIOGRAM The electrocardiogram (ECG) is a fundamental part of cardiovascular assessment. It provides information on the heart rhythm and underlying cardiac morphology. A 'normal ECG' (Fig. 12.10) at rest can be reassuring but also misleading, as it can be found even with life-threatening coronary disease; in the absence of symptoms, a routine ECG is thus of limited value.

Recording an ECG The patient should be relaxed and warm (not shivering) before the 10 skin electrodes are applied to skin previously mildly abraded with spirit. The limb leads provide an electrical 'view' around the heart in the frontal plane, and the

1

476

Fig. 12.5

2

Fig. 12.6

six ventricular leads view the heart from the complementary horizontal plane (Fig. 12.11). By convention, the lead connections have been made to show a positive deflection when the wave of depolarizing current travels towards the electrode. The final shape of the ECG is determined by the lead being sampled and by the angle (vector) at which it 'views' the activity in the heart; the magnitude of the recorded deflection is affected by the mass of tissue undergoing electrical activation. Understanding the ECG therefore requires that the anatomy of the heart, the route taken by electrical activation and the viewpoint of the electrode are known. Differences in anatomy and minor variations in lead positions can account for some of the variability in clinical ECGs; pathological processes account for the rest. Standard gain settings (1 cm/mV deflection) and paper speed (25 mm/s) complete the ECG recording.

Atria I activity and the P wave In the normal heart the dominant pacemaker is to be found in the high right atrium at the sinoatrial node. The depolarizing wave front moves through the right atrium first and arrives at the left atrium fractionally later. The wave front moves towards +60° (see Fig. 12.14). The surface ECG records a P wave which is positive and best seen in

12

FIG. 12.12 Normal and abnormal P waves A On the left is shown the direction of atrial activation that produces the normal P waves shown in lead II. In P pulmonale (PP) the P wave is abnormally tall and peaked. When activation is retrograde (nodal rhythm shown by dashed line), the P wave is inverted, B In left atrial hypertrophy (e.g. in mitral stenosis) a broad bifid P wave is seen in lead II and a prominent negative deflection in lead V1; this is P mitrale (PM).

the depolarizing wave front, and may be seen when atrial activation is initiated at the atrioventricular (AV) node low in the right atrium. Other ectopic atrial rhythms may produce P waves of varying shapes.

The PR interval

FIG. 12.11 Position of leads and electrodes in the ECG [A] Frontal plane. B Horizontal plane. Four limb leads and six precordial leads are routinely used, the anatomical landmarks for the latter being: V1, right sternal edge, 4th intercostal space; V2, complementary position to left of sternum; V3, between V2 and V4; V4, midclavicular line, 5th space; V5, anterior axillary line; V6, midaxillary line, horizontally in line with V4.

leads II, V1 and I. Right atrial hypertrophy produces a tall (>2.5mm) peaked P wave, sometimes called the P pulmonale 1 from its association with the pulmonary hypertension of pulmonary disease. Left atrial hypertrophy (P mitrale) 0 produces a broad notched P wave (>0.12s duration) in leads II or III, with a negative secondary deflection in V1 which exceeds the right atrial positive deflection in that lead (Fig. 12.12). P waves that are inverted in lead II are caused by an abnormal direction of

Between atrial and ventricular depolarization, the ECG signal records a flat isoelectric line for 0.12-0.20 s (Fig. 12.13). This is the delay encountered by the depolarizing wave front in the AV node before rapid propagation to the ventricular myocardium. The PR interval will diminish at increased heart rates and under the influence of adrenaline (epinephrine). It increases in diseased AV nodes, with certain drugs and, occasionally, in normal people.

The QRS complex and ventricular activation Normal QRS The normal QRS parameters are shown in Figure 12.13. Ventricular activation is complete within 0.12s, with virtually simultaneous activation of the left and right ventricles owing to the system of specialized rapidly conducting muscle cells that form the bundle of His and the Purkinje network. Activation of the interventricular septum begins first from left to right and accounts for the small q wave (duration <0.04s and depth <25% of R wave height) seen in leads V5, V6 and, usually, V4. This wave is matched by the r wave in V1 and V2. Because the left ventricular

477

pattern, with a small initial r wave and large secondary R wave, is seen in the right-facing chest leads V1, V2 ± V3, as right ventricular forces occur late and unopposed by left ventricular contraction; this is also manifested by late deep slurred S waves appearing in V4 to V6. Left bundle branch block causes a delay in left ventricular activation, which occurs from right to left, causing a marked axis shift (greater than -60°) and loss of q waves in V5 and V6. Lead V1 shows only a qS wave and the lateral leads (V5 and V6) the characteristic slurred RsR or 'M' pattern. Left bundle branch block is virtually always pathological, the commonest cause being ischaemic heart disease. Right bundle branch block may be a normal finding, but the change from normal conduction to one or other of these patterns implies heart disease. FIG. 12.13 Normal PQRST parameters A Duration. QT interval varies with rate. B Height.

muscle mass is greater than that of the right ventricle, it is the left ventricular forces that dominate the chest leads, with rS waves in leads V1 and V2 and qR waves in leads V5 and V6, with an intermediate transition zone to be found between leads V3 and V4. The mean frontal plane vector (QRS axis) This is obtained from the limb leads and describes the direction of the dominant electrical forces. It can be determined for all the major ECG deflections (P wave, QRS complex and T wave), but is mainly used to summarize ventricular depolarization as the mean QRS axis. It is roughly estimated by choosing the lead with the smallest equiphasic deflection and assuming that the vector is at 90° to this lead. On this vectorial line the limb lead that shows a positive deflection indicates the direction of the axis. The units are degrees, with values above the equator (0°) being prefixed negative (Fig. 12.14). The vector is normally about +60°. Left axis deviation is variously defined, but a shift further than -30° is definitely abnormal; right axis deviation is definite at greater than 120°. The latter can be due to right ventricular hypertrophy or, less commonly, to abnormal patterns of activation caused by interruption of the posterior fascicle of the left bundle branch, in left posterior hemiblock. Left axis deviation is an 'electrical' phenomenon, caused by interruption of the anterior fascicle of the left bundle branch. It is associated with the loss of the normal septal depolarization and hence the loss of q waves in V5 and V6. This is called left anterior hemiblock; in these conditions the QRS duration remains less than 0.12s.

478

Bundle branch block Interruption of the right or left bundle branches produces a bundle branch block pattern characterized by broad QRS complexes of more than 0.12s duration (Fig. 12.15). With right bundle branch block an rSR or 'M'-shaped

Rate-dependent bundle branch block Occasionally, normal QRS morphology is present at low heart rates, but a bundle branch block pattern appears at faster rates. This is usually right bundle branch block, as its refractory period is generally longer than that of the left bundle branch. Rate-dependent bundle branch block may occur with a supraventricular tachycardia, where the broad QRS complexes may wrongly suggest the diagnosis of ventricular tachycardia. Ventricular hypertrophy The tallest R wave should not exceed 27mm, nor the deepest S wave 30mm, and the sum of the tallest R wave and the deepest S wave should not exceed 40mm (Fig. 12.13). Left ventricular hypertrophy (LVH) is diagnosed when these criteria are exceeded, with a normal total QRS duration of less than 0.12s. The intrinsic deflection is delayed by more than 0.04s in the left-facing leads, this measurement being taken from the onset of the QRS deflection to the point at which the R wave swings abruptly downwards. Associated ST segment depression and T inversion occur in the left-facing limb and precordial leads, particularly with pressure-overloaded ventricles, and is said to represent ventricular 'strain'. Right ventricular hypertrophy (RVH) is suspected when there is a dominant R wave in V1, associated with a rightward QRS axis shift (>+90°) in a QRS complex of less than 0.12s duration. There is also ST segment depression and T inversion in the right precordial leads, and often also clockwise rotation of the ventricular leads with the normal transition zone (V3-V4) moving towards V5-V6.

The ST segment The ST segment begins where the QRS ends, and is normally on the isoelectric line curving upwards into the T wave (the J or junction point). There may be 1-2 mm of elevation in the high anterior chest leads (V1 and V2), but greater elevation is seen with myocardial infarction, the so-called current of injury (p. 556, Fig. 12.63), and pericar-

12

FIG. 12.14 Determining the mean frontal plane axis A Lead aVL has the smallest biphasic signal. The axis is perpendicular to this and its direction is given by the upward deflection produced in lead II. The axis, shown by the arrow, is therefore +60° (normal axis). B The axis is -90° (perpendicular to I and away from aVF). This is left axis deviation. C The smallest and most biphasic complex lies between II and aVF and the axis is therefore greater than 120° (severe right axis deviation).

ditis (p. 587, Fig. 12.66). A normal variant, in which there is marked anterior ST elevation, is frequent in young black men. Depression of the ST segment (p. 545, Fig. 12.57) is the hallmark of cardiac ischaemia. This can vary from subtle loss of the normal slight upward curvature and acute angulation with the T-wave origin, to more marked depression, sloping downwards before abruptly changing direction with the T-wave onset. The degree of depression of the ST segment is a rough indication of the severity of the myocar-

dial injury. The ST segment is encompassed within the QT interval. This shortens with increasing heart rate, but may be prolonged (>0.40s or >50% of previous R-R interval) by hypocalcaemia or hypomagnesaemia.

The T wave: myocardial repolarization There is more variability in T-wave amplitude and orientation than in other parts of the ECG, so that criteria for normality have to be more flexible. The T wave may be flat

479

amplitude and shares the T-wave orientation. It is best seen in lead V3 and is enlarged in hypokalaemia. Its genesis is uncertain. 1

Exercise electrocardiography The normal 12-lead resting ECG often shows no abnormalities in patients who develop symptoms on exercise. The exercise test, on a treadmill or cycle ergometer, may be performed for the following reasons: • To reproduce the symptoms, so that the patient can describe them as they happen under standardized exercise conditions; • To assess 'functional capacity', i.e. the amount of work that can be performed before the patient becomes limited by symptoms, and also how rapidly the patient recovers from exercise; • To ascertain whether the symptoms are accompanied by any objective evidence of myocardial ischaemia, cardiac arrhythmia, hypertension or other events that would make unsupervised exercise hazardous; • To monitor treatment with serial studies; • To assess prognosis in patients with known cardiac disease.

FIG. 12.15 Bundle branch block A Left bundle branch block. B Right bundle branch block.

or abnormal in non-cardiac disease and in some normal patients. It is asymmetrical in shape, with a slower upstroke than downstroke and a smooth rounded peak. Symmetrical peaked T waves may indicate ischaemia, whereas notching may be a feature of pericarditis. T-wave height should not exceed 10mm in the precordial leads; larger heights without LVH suggest acute myocardial infarction or hyperkalaemia. T-wave orientation is normally upright in leads with R waves over 5 mm in height, so that, in the normal adult, T waves are upright in I, II and V3 to V6, are always inverted in aVR, and may be inverted in aVL, aVF, III, V1 and V2. The evolution of acute infarction (Fig. 12.63, p. 556) involves T inversion in leads affected by the injury, and these changes may persist.

The U wave The U wave follows the T wave (Fig. 12.13), is small in

1

480

MCQ 12.5

The exercise load is started at an easily managed level, kept stable for 1.5-3 minutes at each load, and then serially increased with the aim of 'exhausting' the subject within 15 minutes (although in submaximal tests exhaustion is not usually the end point). The patient is connected to an ECG monitor and the blood pressure is also recorded. A defibrillator is available. The absolute indications for stopping the test are: • • • • • • • • •

Severe ischaemic symptoms Fall in systolic blood pressure >10mmHg ST depression >3 mm New arrhythmia Patient requests it. The relative indications for stopping are: Achieves target heart rate Systolic blood pressure >220 mmHg Tired patient Angina.

Interpretation The exercise test is conclusively positive if the patient develops classic anginal pain, and downsloping ST segment depression (>2mm) is shown on the ECG (Fig. 12.57, p. 545). Minor degrees of upsloping ST segment depression are common, particularly in anxious individuals and especially in young women. T-wave inversion is not necessarily ischaemic in origin. A commonly used criterion is 1mm ST depression more than 0.08s after the previous QRS complex. Severe ischaemia is shown by widespread ST depression in several leads which takes several minutes to return to normal. There may be a sudden drop in blood

pressure during exercise; this may be unaccompanied by ECG changes but is diagnostic of severe ischaemia and is an absolute indication to stop the test.

CHEST X-RAY Because of the almost uniform density of its components, the cardiac shadow on a PA chest film is a silhouette of a three-dimensional structure displayed on a twodimensional film (Fig. 12.16). Examination of a cardiac X-ray requires a logical approach, and includes the following:

which may result in an abnormal appearance of the heart. Lateral films may show a depressed sternum or abnormal sternal segmentation. There may be rib notching, suggesting coarctation.

12

Heart size and cardiothoracic ratio Minor variations in heart size can be technical in origin. It is usual to measure the cardiothoracic ratio (the maximum transverse diameter of the heart compared with that of the ribcage), which is 0.5 or less in normals. This commonly accepted criterion is, however, rather rigid and needs to be modified with athletes and people of African descent, many of whose hearts exceed this limit but are by other criteria physiologically normal. When comparing successive chest X-rays, an increase in transverse diameter of the heart of more than 1.5cm is regarded as significant.

• A check of the technical quality of the film (PA film, full inspiration, no rotation); • An assessment of heart size (and changes from previous films if available); • A search for evidence of specific chamber enlargement and abnormalities of cardiac anatomy; • An examination of the lung fields and vessels; • Identification of valve or other calcification.

Anatomy and specific chamber enlargement

An AP film is only of value in assessing the lung fields or gross changes in heart size. There may be evidence of thoracotomy, rib resection, scoliosis or severe kyphosis,

The border-forming structures of the cardiac silhouette in the PA and lateral views are shown in Figure 12.16. The approximate positions of the heart valves are also shown.

FIG. 12.16 Composition of the normal cardiac silhouette on chest X-rays [2 PA view. B Lateral view. RV and LV = right and left ventricles; RA = right atrium; LA = left atrium; SVC = superior vena cava; Ao = aorta; RPA and LPA = right and left pulmonary arteries.

481

Left atrium (LA) LA enlargement may produce widening of the angle of bifurcation (carina) into the main bronchi, with lifting of the left main bronchus. The increased density of an enlarged LA may make it visible within the right heart border, or it may protrude beyond it. Enlargement of the LA can cause straightening, and even bulging, of the left heart border. On lateral chest X-ray, enlargement produces posterior displacement of the upper posterior cardiac border. Left ventricle (LV) The LV makes up the cardiac apex and part of the left heart border on PA X-rays, and much of the posterior heart border on lateral views. LV enlargement usually displaces the apex of the heart downwards and laterally on the PA view, whereas on the lateral view the LV produces a prominent convexity of the lower half of the posterior border of the heart.

Aorto In young adults the upper right cardiac border may not include the ascending aorta, but with ageing the aorta elongates and becomes more tortuous and prominent on chest X-ray. Prominence of the ascending aorta may thus be due to age as well as to hypertension, poststenotic dilatation or aortic dilatation from aortic valve disease, aneurysm or dissection. CT, ultrasound or angiography may be needed to define the cause of a prominent aorta. The aortic arch may also be prominent in any of these conditions. Calcification of the ascending aorta occurs in syphilis, but is very common in the arch and descending aorta in normal individuals.

Lung fields and vessels Assessment of pulmonary vascularity tends to be subjective and demands a good-quality film. Normal pulmonary vascularity will show larger vessels in the lower zones, with symmetry between the two lungs. There is a steady diminution in vessel size peripherally. Pulmonary venous hypertension If the pressure in the pulmonary veins rises above about 15mmHg the upper lobe vessels dilate and become larger than the lower lobe vessels, which constrict with perivascular cuffing. There is redistribution of flow to the upper lobes. If the venous pressure rises higher than about 20-25 mmHg the pulmonary lymphatics become engorged, and appearances of interstitial oedema are seen on the chest X-ray. These are the appearance of Kerley B lines (horizontal septal lines best seen at the costophrenic angles) and increased haziness of the perihilar regions, sometimes resulting in the classic but uncommon 'bat's wing' appearance of the chest X-ray, often with small effusions (Fig. 12.17). Further rises in pressure above 30 mmHg result in alveolar oedema, with dense opacification of the lung from fluid collection. Increased lung blood flow With increased blood flow to the lungs (as in a left-toright shunt) the lungs are vascular, with quite large vessels running to the periphery. This can be distinguished from left heart failure by the lower lobe vessels remaining larger than the upper lobe vessels. Pulmonary oligaemia Pulmonary oligaemia (as in severe pulmonary stenosis) is

Right atrium (RA) Enlargement of the RA is seen as prominence of the lower half of the right heart border and is usually accompanied by superior vena cava (SVC) prominence. The SVC forms the highest part of the right border of the cardiac silhouette and is prominent in right heart failure, SVC obstruction, or displacement of the SVC by the aorta. Right ventricle (RV) The RV is an 'internal' structure on the PA X-ray unless there is massive RV dilatation with the LV rotated posteriorly. RV dilatation results in a more globular appearance to the heart on PA film, with the cardiac apex being more elevated - the so-called coeur en sabot.

482

Pulmonary conus and trunk The pulmonary trunk is more prominent in young subjects than later in life. A dilated pulmonary trunk may indicate a poststenotic dilatation, pulmonary hypertension, increased pulmonary flow, or can be idiopathic. It therefore needs to be interpreted along with the appearance of the lung fields.

FIG. 12.17 Radiological features of left heart failure PA chest X-ray showing bilateral shadowing spreading from the hila, typical of acute pulmonary oedema. The heart size is increased.

associated with an avascular appearance of the lung fields but normal-sized central pulmonary arteries (although the pulmonary trunk may be dilated as a poststenotic dilatation in pulmonary stenosis). A region of oligaemia (e.g. left upper zone) may be detected after a moderate or large pulmonary embolism. Pulmonary arterial hypertension The presence of pulmonary arterial hypertension gives rise to large central pulmonary arteries which are rapidly pruned peripherally, with relatively avascular-looking lung fields.

Calcification Calcification is usually best seen on penetrated PA and lateral films, as much of the calcification seen within the heart overlies the spine on PA films. The mitral and aortic valves may both calcify. On lateral films the latter is usually lying above a line joining the tracheal bifurcation and the costophrenic angle, whereas the mitral is predominantly below the line. Calcification is also sometimes seen in the great vessels, the wall of a rheumatic LA, the pericardium, and within the LV in thrombus in an LV aneurysm.

ULTRASOUND TECHNIQUES If a beam of ultrasound is directed into the thorax through a suitable 'acoustic window' the various tissue interfaces in the line of the ultrasound beam reflect some of the sound back towards the transmitting transducer, and the red blood cells in its path scatter the ultrasound. The reflected ultrasound takes varying times to get back to the transducer, depending on the depth of the interface from which the reflection occurs. Ultrasound imaging allows for a detailed, painless and safe investigation of the heart and great vessels which can be repeated many times. Anatomical and functional information can be obtained rapidly in virtually any circumstance, from the bedside in and out of hospital to intensive care units and operating theatres, and even in the unborn fetus. The limitation of obtaining a suitable imaging window on the chest wall (transthoracic echocardiography, TTE) has been surmounted, at least in part, by using small probes inserted endoscopically and viewing the heart from the oesophagus and stomach (transoesophageal echocardiography, TOE).

Echocardiography The method of obtaining the M-mode echocardiogram is illustrated in Figure 12.18. This imaging method remains important in producing a very accurate resolution of distance and time (the ECG provides the timing index) so that the movement of ventricular walls and valve leaflets can be measured to within 2mm and 10ms. The dimensions

of the left ventricular walls in systole and diastole, plus the size of the left atrium, proximal aorta and right ventricle, are the parameters most commonly taken from these recordings. Others, such as valve leaflet excursion and the extent of thickening and thinning of the LV walls, are also used. The two-dimensional echocardiogram, produced by swinging the ultrasound beam through an arc or sector, produces pictures that are immediately recognizable as cardiac anatomical structures (see Figs 12.19 and 12.20A). Their pattern of movement can also be seen in 'real time', adding to the value of the assessment. This method is particularly well suited to the demonstration of LV function, being able to demonstrate regional as well as global disturbances of contraction. Valves can be seen to open and close during the cardiac cycle, and vegetations of >3.0mm in size may be detected (see bacterial endocarditis, p. 574). It is the method of choice for the demonstration of pericardial fluid and atrial tumours, and can reveal intracardiac thrombus following myocardial infarction. Using peripheral venous injections of fluids containing microscopic bubbles (usually agitated normal saline in a bolus of 1025 ml) intracardiac shunts can be demonstrated - contrast echocardiography. Newer commercial contrast agents have been developed with echogenic particles of >6um which can pass through the pulmonary capillary bed (10 um). These agents may become useful in defining left ventricular function and viability. Echocardiography can be used to map regional left ventricular function to compare images obtained at rest with those immediately after exercise or during infusion of an inotropic agent such as dobutamine. Stress echocardiography can indicate the extent of regional dysfunction produced by ischaemia in coronary artery disease, particularly in patients with atypical symptoms and normal LV function at rest. An improvement in contraction of a region of LV with inotropic stimulation differentiates stunned from dead myocardium, an important difference if surgery or angioplasty (PTCA) is being considered.

12

Doppler ultrasound The technique of Doppler ultrasound makes use of the change in frequency of the backscattered ultrasound energy from the red blood cells. This change of frequency is proportional to the blood velocity; thus, if the beam of ultrasound is aligned with the jet of blood within the heart, accurate measurements of blood velocity can be made and gradients across the valves calculated (Fig. 12.20B) from the equation P = 4V2

where P is the pressure gradient in mmHg and V the velocity in ms. The predicted pressure gradient from Doppler measurements is the instantaneous pressure gradient, whereas cardiac catheterization usually measures the peak-to-peak pull-back gradient (Fig. 12.22).

483

FIG. 12.18 M-mode echocardiography The diagram shows the position of the transducer and directions of the ultrasonic beam. In A the beam traverses the left ventricle (LV) below the mitral valve. Systole (S) and diastole (D) are indicated. In B the beam is aimed to demonstrate the mitral valve (MV). In C the aortic valve (AV) and aorta (Ao) are shown.

FIG. 12.19 Two-dimensional echocardiography Parasternal long axis views of the heart in systole and diastole are shown. (The direction of the beam is equivalent to that shown by line B in Figure 12.18A.) The opening and closure of the aortic and mitral valves are clearly shown.

Doppler ultrasound is used predominantly for measuring pressure gradients and can also be used to demonstrate the presence and severity of valvular regurgitation. It is extremely sensitive in detecting minor degrees of regurgitation, which are often unexpected. Colour flow Doppler imaging Colour flow Doppler uses a form of pulsed wave Doppler to superimpose a colour image of the blood velocity and

1

484

Figs 12.7, 12.8

direction on a two-dimensional image of the heart. By convention, blood flowing away from the probe is coded blue and that towards it red. Very high velocities are associated with multicoloured pixels. This type of imaging does not lend itself to quantification, but is very sensitive at detecting regurgitant jets and unusually directed high-velocity jets such as appear in leaks around prosthetic heart valves. Colour flow Doppler has now become a routine part of the echocardiographic examination; it gives a rapid overall assessment of intracardiac flows and makes the rest of Doppler technology easier to use and appreciate. Transoesophageal echocardiography (TOE) All ultrasound techniques are limited by the size of the acoustic windows, which must avoid bone and lung. The windows are large in babies and children (almost complete in the fetus) but become smaller in adults, especially those with lung disease. Miniaturized probes mounted on endoscopes allow some of these deficiencies to be overcome (Table 12.3). Virtually all the imaging planes available transthoracically can be achieved, and in addition the atria, their appendages and the descending aorta can be seen. The technique is more invasive, being equivalent to an upper gastrointestinal endoscopic procedure. Doppler tissue imaging (DTI) This is a modification of conventional colour flow imaging where low-frequency motion is encoded and the high velocities of the red cells are filtered out. The velocity, direction and acceleration of motion of the myocardial

12 FIG. 12.20 Two-dimensional echocardiogram and Doppler ultrasound A Apical four-chamber view of the heart. B The continuous wave Doppler signal shows a diastolic jet (upwards towards the transducer). An abnormal systolic jet away from the transducer, due to mitral regurgitation, is also shown. The velocity trace indicates peak velocities of 2ms and 5ms respectively (corresponding to peak gradients of 16 and l00mmHg).

TABLE 12.3 Indications for transoesophageal echocardiography • • • • • • • • •

Poor views on transthoracic echocardiography Investigations of prosthetic heart valves, especially mitral and aortic Endocarditis (better resolution, to 1 mm with high-frequency transducers) Aortic dissection or aneurysm (to view descending aorta) Mitral stenosis - intra-atrial clot can be excluded prior to balloon valvuloplasty Mitral valve disease prior to surgical repair Systemic embolism (looking for a cardiac source) Suspected atrial mass (tumour or clot) Intraoperative, e.g. exclusion of mitral regurgitation during valve repair

cells and valve structures are colour coded, as in conventional colour flow imaging. This technique is used to detect regions of wall motion abnormality and is particularly useful in stress echocardiography as well as in quantifying regional diastolic dysfunction.

Limitations on use The combined techniques of echocardiography and Doppler ultrasound save many patients, particularly children, from invasive studies, as in many cases the complete pathology can be demonstrated with ultrasound. These techniques cannot, however, measure chamber pressures, such as the left ventricular end diastolic pressure, or show the coronary arterial anatomy.

NUCLEAR IMAGING Two major nuclear techniques are used in cardiology, myocardial perfusion imaging (MPI) and radionuclide ventriculography. Isotopes, principally y-emitting agents such as technetium-99m and thallium 201, are used to image heart function. The commonest use for nuclear techniques in cardiology is in coronary heart disease, where perfusion imaging can differentiate normal from reversibly ischaemic from infarcted myocardium.

Ventriculography, using gated pool scanning, is used to quantify ventricular function by calculating the ejection fraction (EF) of the ventricle. Very short-acting positronemitting isotopes have been used to measure myocardial metabolism and its alteration in ischaemic states, but they remain a research tool as the isotopes must be produced on site by a cyclotron.

Nuclear ventriculography Nuclear ventriculography uses intravenously injected 99m Tc-pertechnetate bound to erythrocytes. In the first-pass technique, a discrete bolus of labelled blood is followed through the heart and data are collected by an externally placed gamma camera for 30-50s. A fresh injection is required for each view. Short-lived isotopes such as 195 Au enable rest and exercise studies to be performed with relatively low cumulative exposure and background. In an equilibrium study the whole of the circulating blood is labelled and data collection can take much longer, with serial views obtained over time. Resolution of the gamma camera image is increased by taking a large number of images over several cardiac cycles, and the images from different phases of the cardiac cycle are averaged. This process is accomplished by 'gating' with reference to the ECG as a time marker and dividing the cardiac cycle into a series of time intervals. This provides a non-invasive assessment of the ventricular ejection fraction (EF) and wall motion. 1 Derived images and measurements Ejection fraction (EF) The ejection fraction is normally above 50%, and should rise with exercise by at least 5%. Regional wall motion abnormalities are' shown by quantitative colour coding of emission over the heart. Limitations of nuclear ventriculography In severe right heart failure, a peripheral venous injection of isotope is dissipated through a large mixing volume before it reaches the heart. Tricuspid regurgitation aggravates the problem, leading to greatly degraded imaging. Time-gating over many cycles is not usually possible in patients with atrial fibrillation or other significant arrhyth-

485

mias. More than a minor degree of movement of the thorax relative to the gamma camera also tends to limit studies during exertion.

Myocardial perfusion imaging (MPI) The technique of myocardial perfusion imaging depends on the radionuclide being taken up by viable myocardium that is normally perfused. Generally two sets of images are required, one obtained at peak stress and the other at rest. 1 Areas of viable myocardium which are not perfused during stress appear as areas that 'fill in' during the resting scan. Stress defects on the scan that do not perfuse on the resting images represent infarcted myocardium. Although bicycle ergometry is the commonest stress used in MPI, those patients who are unable to exercise are given vasodilator agents such as adenosine or dipyridamole by infusion. Alternatively, catecholamine derivatives, dobutamine or arbutamine are used. Radionuclides used for MPIU studies are thallium-201 (201T1) and 99mTc-MIBI (Cardiolite) on 99mTc-tetrofosmin (Myoview). The latter two agents require reinjection, but no overall superiority of one agent over the others has been established. • Perfusion imaging has a high prognostic capability. The risk of a future cardiac event in a patient with normal myocardial perfusion is less than 1% per year, even in the presence of angiographically proven coronary artery disease. The converse is true, severe defects being associated with a poor prognosis. • In addition to providing vital prognostic information in patients with known or suspected stable coronary disease, MPI performed within 3 days of myocardial infarction can identify low-risk patients. It is used to stratify the risk of non-cardiac surgery in patients likely to have widespread vascular disease.

MAGNETIC RESONANCE IMAGING (MRI) Cardiovascular MRI (CMR) has the capacity to provide high-quality tomographic images of the heart and great vessels, combined with reproducible functional assessment. New technology means that scans may take only 20 minutes to acquire. The advantages of CMR include its non-invasive nature and freedom from ionizing radiation. Cost and patient tolerability (because of claustrophobia) have been limitations which are diminishing with speedier image acquisition. CMR has a particular role in the diagnosis and assessment of anatomical abnormalities, such as aortic aneurysm, dissection and coarctation. Its use in complex congenital

A Fig. 12.9 0 Fig. 12.10

486

TABLE 12.4 Indications for MPI in chronic coronary artery disease • Detection and quantification of ischaemia in symptomatic or selected asymptomatic patients • Risk stratification prior to non-cardiac surgery • Selection of 'culprit' coronary lesion prior to PTCA • Symptom evaluation after PTCA or CABG • Evaluation of viable myocardium

heart disease is becoming routine in some centres. CMR can provide imaging where echocardiography is impossible, and it can usefully quantify valve regurgitation to a degree unobtainable with other techniques. Regional and global ventricular function can be obtained, as well as unique insights into myocardial tissue in diseases such as sarcoid and iron overload due to transfusion (e.g. thalassaemia) or haemochromatosis. Angiography of the neck vessels and the renal arterial system is now well established. Imaging of coronary arteries is improving but only superior to current radiological methods in certain rare circumstances, such as the assessment of aberrant coronary arteries.

COMPUTED TOMOGRAPHY X-ray technology using ultrafast CT scanners has developed to the point of allowing rapid scans of the heart, flow and morphological detail to be obtained. The spiral CT scan is now the diagnostic tool of choice in suspected moderate or large pulmonary embolism.

CARDIAC CATHETERIZATION Cardiac catheterization is a diagnostic procedure in which a flexible tube is introduced under local anaesthetic, into the brachial or femoral veins and passed into the right side of the heart, or from the brachial or femoral arteries to the left side of the heart. In children, the procedure may require a general anaesthetic. In right heart catheterization the catheter can be passed to the right atrium, right ventricle and pulmonary artery in turn. Using an end-hole catheter, an indirect LA pressure can be obtained by wedging the catheter in a small pulmonary artery. Left heart catheterization from a peripheral artery enables cannulation of the aorta, the coronaries and, by passing the catheter retrogradely through the aortic valve, the LV cavity.

Use of cardiac catheterization Catheterization allows pressure measurements to be made and cine angiograms performed by the injection of contrast agents. Selective catheterization of the coronary arteries provides high-quality angiographic views of the coronary

12

FIG. 12.21 Radionuclide ventriculography: the multiple uptake gated acquisition (MUGA) scan The blood pool is labelled and counts over the left ventricle are related to defined proportions of the cardiac cycle (1-15 in the diagram). These are summated to give an average radioactivity for end systole (B) and end diastole (A), from which the ejection fraction can be calculated.

tree to demonstrate its anatomy. 2 This has become the major use of cardiac catheterization, complementing noninvasive techniques such as echocardiography and nuclear imaging. In addition to its diagnostic role catheterization is now the tool by which many therapies are achieved, in particular the use of angioplasty balloons and stents to dilate narrowed blood vessels in the heart and circulation. Devices can be introduced to close shunts, e.g. atrial septal defects or patent ducts, as well as instruments to retrieve foreign bodies lost in the circulation. In valvular heart disease many of the uses of catheterization have been supplanted by the non-invasive techniques discussed in this chapter. Nevertheless, catheterization may still be required to obtain the following information: • To measure chamber pressures and valve pressure gradients (Fig. 12.22); • To estimate valve areas in stenoses from the approximation: Valve area in cm2 =

Cardiac output in L/min Pressure gradient in nimHg mmHg

• To perform angiography by using radio-opaque contrast agents to show cardiac anatomy, assess valve or conduit regurgitation and assess ventricular function. • Measurement of oxygen content for demonstration of shunts (Fig. 12.23). Sampling in the different chambers of the right side of the heart can be used to demonstrate the presence and size of a left-to-right shunt by the rise in oxygen content (saturation) of the blood. Persistent arterial desaturation, despite the patient breathing 100% oxygen, demonstrates the presence of a right-to-left shunt. Measurement of pulmonary artery and systemic arterial blood oxygen content, together with mea-

FIG. 12.22 Cardiac catheter pressure tracing in aortic stenosis As the catheter is withdrawn from the left ventricle into the aorta, there is an abrupt drop in pressure demonstrating a gradient across the aortic valve of 140mmHg. Lower tracing is the ECG.

TABLE 12.5 Complications of cardiac catheterization • Peripheral arterial damage (dissection, haematoma) • Cardiac arrhythmia (SVT, VT, VF) • Thromboembolism, due to clot on catheter, air embolism, dislodged atheroma • Dissection of a coronary artery • Cardiac tamponade • Renal dysfunction - especially in diabetes, or - when performed in the presence of other nephrotoxic drugs, including biguanides.

surement of oxygen uptake by timed expired air gas collection (or, in many centres, estimated oxygen uptake from normograms of patient's surface area), are used to calculate cardiac output. The use of cardiac catheterization in coronary angiography is described on page 546.

Complications of cardiac catheterization The complications of cardiac catheterization are shown in Table 12.5. The risks are small (<0.1%) and are minimized by current techniques. They are increased if the patient has small vessels, is in heart failure, or has very severe coronary artery or valvular disease, and at the extremes of age (Table 12.6).

FURTHER READING ON CARDIAC INVESTIGATION ACC/AHA Task Force Report 1995 Guidelines for clinical use of cardiac radionuclide imaging. J Am Coll Cardiol 25: 521-547. ACC/AHA Guidelines for cardiac catheterization and cardiac catheterization laboratories 1991 ACC/AHA Ad Hoc Task Force on Cardiac Catheterization. J Am Coll Cardiol 18:1149-1182.

487

treatment of heart failure is solely to correct these secondary defects and to relieve the symptoms, missing out therapy to improve the heart's function. Secondly, pathological states of the vasculature may produce syndromes similar to heart failure, but the heart may not actually be failing. Examples are severe hypertension (when there is excessive vasoconstriction) and the so-called 'high-output failure' (such as that produced by significant arteriovenous shunting). In these conditions the heart is required to perform supernormally and is unable to do so. To refer to these conditions as 'heart failure' is a misnomer, and to treat them in the same way as other forms of heart failure, e.g. by administering positive inotropic agents to augment cardiac pumping, is inappropriate. FIG. 12.23 Oxygen saturation in the cardiac chambers and great vessels in ventricular septal defect A step-up in saturation occurs in the right ventricle. Ao = aorta; PA = pulmonary artery; RV = right ventricle; LV = left ventricle; RA = right atrium; LA = left atrium.

TABLE 12.6 Relative contraindications to cardiac catheterization • Recent CVA • Significant renal impairment or worsening renal function • PUO • Anaemia, Hb<10g/dL • Uncontrolled hypertension • Significant electrolyte disturbance • Uncontrolled diabetes or thyrotoxicosis • Anticoagulated state (warfarin, INR < 2.0; heparin, ACT > 200s) • History of previous contrast reaction • Severe anxiety state or psychological disturbance • Use of biguanide drugs within 48 hours

Bellenger N, Pennell D J 1999 Magnetic resonance imaging in cardiology. J Roy Coll Phys (Lond) 33:12-18. Brown K A 1996 Prognostic value of myocardial perfusion imaging: state of the art and new developments. J Nucl Cardiol 3: 516-538. Chambers J B, de Belder M A, Moore D 1997 Echocardiography in stroke and transient ischaemic attack. Heart J 78(suppl 1): 2-6. Krahninkel W, Marwick T, Gulker H (eds) 1997 Stress echocardiography: a well established diagnostic tool. Eur Heart J 18(suppl D).

HEART FAILURE

488

Heart failure can be defined simply as failure of the heart to function as it should to maintain an adequate circulation. Such failure has several physiological consequences, which together produce the clinical syndrome of heart failure. It is not enough to define heart failure as a clinical syndrome consisting of congestion and hypoperfusion. First, this might lead to the wrong assumption that the definitive

The spectrum of cardiac performance The main difference between the hearts of an elite athlete, a sedentary healthy subject and a heart failure patient lies in their peak cardiac performance, not in their cardiac performance at basal resting states (they all have comparable cardiac outputs of 5-7 L/min and cardiac power outputs of 0.8-1.3 W). Elite athletes can attain cardiac outputs of up to 35 L/min and power outputs of up to 11 W, whereas the values for normal subjects are 15-20 L/min and 4-6 W. Patients with moderate heart failure achieve 10 L/min and 2.5 W. It is the deterioration in peak cardiac performance that heralds the onset of heart failure; this deterioration is progressive. A point will be reached when, even at maximal stimulation, the hearts of some patients in severe heart failure or in cardiogenic shock are not able to attain the basal resting values observed in normals. These patients have a very grave prognosis.

PATHOPHYSIOLOGY OF HEART FAILURE The major effects of heart failure are organ hypoperfusion, arrhythmia, congestion, vasoconstriction and redistribution of regional blood flow. Arrhythmia in heart failure may be secondary to ischaemia, abnormal mechanical loads and the production of areas conducive to the formation of re-entry circuits (see p. 506). It is important to remember that the primary defect lies in the heart. The secondary defects, such as fluid retention and excessive vasoconstriction, are compensatory adaptive mechanisms invoked to maintain as normal a circulation as possible. They are mediated by neuroendocrine activation and produced by the primary cardiac dysfunction. Together with the primary defects, they produce the symptoms and signs of heart failure (see below). The tertiary defects are the detrimental effects of treatment. Categorizing the defects in this way is helpful in planning treatment (Table 12.7).

Primary defects The definitive treatment of heart failure should be directed

12

TABLE 12.7 Defects in heart failure and their management Defect

Clinical response

Primary Pump dysfunction

Establish a cause and remedy if possible

Secondary Fluid retention Vasoconstriction Hypoperfusion Flow redistribution Arrhythmia Tertiary (usually due to treatment) Electrolyte imbalance (diuretic Px) Arrhythmia (diuretics and digoxin) Skeletal muscle deconditioning (bed rest)

Diuretics ACE inhibitors; other vasodilators (e.g. nitrates, hydralazine) Volume replacement (if acute); decrease diuretics Inotropes (if acute); ACE inhibitors Correct tertiary defect; antiarrhythmic drugs

Monitor biochemistry, potassiumsparing diuretics Review drug (proarrhythmic effects, e.g. digoxin) Exercise training

at diagnosing and correcting the primary defects. Examples might be the removal of cardiodepressant drugs, aspiration of pericardial fluid in tamponade, control of arrhythmias, a pacemaker for complete heart block, repair or replacement of faulty valves, stripping of pericardium in constrictive pericarditis, or surgical closure of intracardiac shunts. By far the most common defect causing heart failure is cardiac myocyte loss, as a consequence of myocardial infarction or cardiomyopathy. Because cardiac myocytes are unable to undergo regeneration, the necrotic myocytes are replaced by connective tissue. The remaining viable myocardium contracts harder and undergoes hypertrophy to compensate for the loss. There is as yet no definitive therapy for this type of defect. Ultimately cardiac transplantation or cardiomyoplasty may be necessary. The extent to which a primary defect compromises overall cardiac performance can be measured by how much cardiac reserve is reduced. A schematic representation of progressive damage to cardiac functional reserve by sequential myocardial infarctions is shown in Figure 12.24.

Secondary defects In the event of heart failure or a normal heart facing excessive haemodynamic burden, the heart relies on cardiovascular compensatory mechanisms to maintain an adequate circulation. The main mechanisms are: • Fluid and salt retention, invoking the Starling effect through an increase in ventricular filling. • Neuroendocrine responses (sympathetic stimulation followed by activation of the renin-angiotensin-aldos-

FIG. 12.24 Progressive damage to the heart by sequential myocardial infarctions With the first infarction there is initial loss of cardiac functional reserve with no compromise of cardiac performance at rest. Further damage results in more loss of reserve, which may be accompanied by depressed cardiac function at rest. Finally, when the cardiac reserve is insufficient even to maintain adequate output at rest (cardiogenic shock), death ensues.

FIG. 12.25 The renin-angiotensin-aldosterone system in cardiac failure Activation of the renin-angiotensin system by failing cardiac output leads to salt and water retention and worsening cardiac output (ACE = angiotensin-converting enzyme).

terone system; Fig. 12.25) to increase cardiac pumping rate and force, and increase vasoconstriction to maintain normal blood pressure and divert the limited blood flow to the vital organs. • Myocardial hypertrophy, which is variable and incompletely understood but may reduce ventricular wall stress and thus augment function. • Readjustment of skeletal muscle metabolism such that the same physical work can be performed with a reduced cardiac effort.

489

Decompensation occurs in heart failure when some of these compensatory mechanisms become excessive. Too much fluid and salt retention results in tissue oedema, which impedes tissue perfusion and reduces vasodilatory capacity. Gut and hepatic congestion impairs absorption and metabolism, contributing to cardiac cachexia (and rendering enteral medication less effective). The ensuing hypoalbuminaemia further worsens the oedema. Pulmonary oedema causes orthopnoea and dyspnoea, impairs gaseous exchange and increases the work of ventilatory muscles, which in turn imposes an extra demand on cardiac performance. Excessive cardiac dilatation causes functional AV valve regurgitation, further reducing ventricular pumping efficiency. Excessive vasoconstriction (due to activation of the sympathetic, renin-angiotensin and vasopressin systems) imposes a higher afterload on the heart. Vasoconstriction is exacerbated by diuretic therapy, which stimulates the renin-angiotensin system and the release of vasopressin. Excessive vasoconstriction of the renal arterioles leads to reduction in renal blood flow and glomerular filtration rates. The resulting azotaemia and oliguria compounds the management of heart failure. Chronic overstimulation of the sympathetic system depletes myocardial noradrenaline (norepinephrine) and causes B-receptor downregulation, reducing myocardial chronotropic and inotropic responses. Excessive myocardial hypertrophy causes cardiac diastolic dysfunction and makes the subendocardial region more prone to ischaemia.

Tertiary defects

490

Pharmacological treatments that alter physiological homeostatic controls lead to complications, some of which can be detrimental. The commonest is probably electrolyte imbalance. The powerful loop diuretics lead to loss of potassium and magnesium, both of which can increase the likelihood of tachyarrhythmias. Dietary replenishment of potassium supplements does not fully compensate for the losses. A better way of avoiding this complication is to use the so-called potassium-sparing diuretics, which also conserve magnesium. Diuretic therapy in severe heart failure can also lead to significant hyponatraemia, which is associated with a poorer prognosis. Diuretics must be used carefully to avoid this pitfall. Arrhythmia is an important consequence of electrolyte imbalance. Hypokalaemia, hypomagnesaemia and hyponatraemia all cause arrhythmia, the latter being especially hard to treat. Overdiuresis can also produce hypovolaemia. This leads to further hypoperfusion, postural hypotension and worsening renal function. Vasodilators have preferential effects on different vascular beds. Such selective vasodilation may counter the 'natural' redistribution of blood flow that preserves flow to vital organs in low-output states. For example, a-adrenergic blockers such as prazosin counteract the redistribution of blood flow effected by noradrenaline (norepinephrine), but prazosin does not improve exercise

tolerance or the prognosis of heart failure patients. Vasodilators that redistribute blood into the splanchnic or cutaneous vasculature may help in acute severe heart failure, but do not improve exercise capacity or mortality in chronic heart failure. Angiotensin-converting enzyme (ACE) inhibitors are immensely useful, but can be counterproductive by worsening renal function in patients dependent on the angiotensin-induced renal efferent arteriole vasoconstriction to maintain glomerular filtration rates. Certain positive inotropic agents (e.g. catecholamines) are potentially toxic, causing arrhythmias and cardiac myocyte necrosis.

CLASSIFICATION There are three main ways of classifying heart failure: • Acute vs. chronic. Although this temporal classification may seem clinically obvious, the pathophysiology and therapy of the two conditions have important differences. In acute heart failure the haemodynamic derangement is severe enough to result in obvious symptoms at rest. In chronic heart failure, however, the circulation at rest is adequate but there is inadequate cardiac reserve to pursue daily activities. Acute heart failure may occur as a consequence of a sudden recent myocardial injury, e.g. myocardial infarction, or as an exacerbation of previously compensated chronic heart failure. • Systolic vs. diastolic (or forward vs. backward). Systolic or forward failure results in low output and low arterial pressure states, whereas diastolic or backward failure results in congestion. Systolic failure is due to inadequate overall myocardial contractile capability, and diastolic failure relates to abnormalities in relaxation. Although heart failure is due most often to systolic dysfunction, some patients have normal systolic function and the failure is due to abnormal diastolic function. The cause is either within the heart muscle itself (hypertrophy, restriction) or in the pericardium (effusion or constriction). In ischaemic heart disease systolic and diastolic dysfunction often coexist, because ischaemia also impairs myocardial relaxation. • Left- vs. right-sided. This classification arose from a simplistic clinical division into whether there is pulmonary congestion (left heart failure) or systemic congestion (right heart failure). However, the right and left sides of the heart are connected in series, and in juxtaposition, enclosed in the pericardium. This complicates the simple division into right and left heart failure. Marked hypertrophy and dilatation of the left ventricle can impinge on right ventricular filling (the Bernheim effect) and cause systemic congestion without intrinsic right ventricular dysfunction.

Epidemiology In western countries approximately 0.5-2% of the popula-

tion is affected by heart failure, and the prevalence is greater with increasing age to about 10% in those over 75 years old. The annual incidence of heart failure is about 3 per 1000 for men aged 45-65 years and 8 per 1000 for men aged 65-75 years. Both the prevalence and incidence have more or less doubled since the 1960s.

Aetiology The main causes of heart failure in the west are ischaemic heart disease, idiopathic cardiomyopathy, hypertension and myocarditis. In the Framingham (Massachusetts, USA) population study (1955-1971), over 75% of those who developed heart failure had antecedent hypertension. More recently the incidence of hypertensive heart failure has decreased markedly, which is probably owing to improved antihypertensive treatment.

Prognosis and natural history Like neural cells, cardiac myocytes do not undergo division: growth is by hypertrophy. In adulthood there is a natural attrition rate of cardiac cells with increasing age. It has been estimated that there is an aggregate loss of about 35% of myocyte cells in the ventricles from young adulthood to old age. This rate is increased by disease. The prognosis of patients with heart failure is poor, and is worse in those with advanced heart failure. Overall, the 5-year survival rate of patients with all degrees of heart failure is approximately 50%. In patients with severe heart failure, the 1-year mortality rate is greater than 50%. About half of these patients die suddenly and the other half die of progressive heart failure.

PROGRESSIVE VENTRICULAR DYSFUNCTION Clinical features The main symptoms and signs of heart failure are those related to congestion and hypoperfusion, either at rest (as in acute heart failure) or during exertion (as in chronic heart failure). In taking a history, the symptoms and signs must be noted and information about the likely cause of heart failure and the severity of its effect on the patient's life should be sought.

Symptoms and signs The symptoms of congestion in the lungs are exertional dyspnoea, orthopnoea (dyspnoea and cough on lying down), paroxysmal nocturnal dyspnoea (breathlessness that wakes the patient from sleep), 'cardiac asthma' (wheeziness due to congestion of the bronchial mucosa) and, rarely, haemoptysis (see p. 522). In some patients the dyspnoea at rest may be episodic, raising the possibility of paroxysmal arrhythmia. The physical signs of pulmonary congestion are fine

SUMMARY 1 Symptoms and signs of heart failure Symptoms • Fatigue • Shortness of breath - on exertion - on lying flat • Paroxysmal nocturnal dyspnoea • Haemoptysis • Weight gain • Swollen ankles • Abdominal distension • Abdominal pain • Lethargy

Signs • Raised JVP • Gallop rhythm • Pitting oedema • Hepatomegaly • Basal crackles • Positive hepatojugular reflux • Cardiomegaly • Ascites • Pallor • Peripheral cyanosis • Cold extremities • Pulsus alternans

inspiratory crackles at the lung bases, but throughout the lung fields in severe cases. These are usually associated with tachypnoea. Expiratory wheeze may be present and should not be assumed to be due to airways disease. Occasionally dyspnoea is atypical, more like Kussmaul respiration (deeper and slower), and is due to lactic acidosis secondary to marked tissue hypoperfusion. Some of these patients may be labelled 'hyperventilators' until the severity of the underlying left ventricular dysfunction is appreciated. In severe pulmonary oedema due to acute left heart failure the patient is in respiratory distress and producing white frothy sputum through the mouth or nostrils. They are pale and cyanosed, and, if pulmonary oedema is unrelieved they progress to a moribund or unconscious state. The clinical features of systemic congestion are raised JVP, peripheral pitting oedema, abdominal distension with discomfort or pain, and dyspnoea due to pleural effusion. In mild failure the oedema only involves the ankles and feet, and it may be dependent (occurring after prolonged upright posture). Patients with chronic leg oedema may develop lower leg ulcers with or without cellulitis. With more severe failure the oedema may involve the thighs, sacrum, abdomen and upper body. Abdominal distension may be due to ascites or hepatic congestion. The latter may induce pain, especially during exercise (hepatic angina). Long-term hepatic congestion and gut oedema lead to weight loss and cardiac cachexia. Occasionally there may be sparing of lower limb oedema, and the patient presents with abdominal and/or upper body congestive features. In bedridden patients oedema may only be found on the sacrum. The symptoms of hypoperfusion include fatigue, lethargy, cold peripheries and dizziness. The latter is often due to postural hypotension. The physical signs include tachycardia, hypotension with a small pulse pressure, pallor and feeble peripheral pulses. In more severe failure or cardiogenic shock, the signs are of cool, clammy and cyanotic peripheries, oliguria or anuria, and blunted mental function. Cardiac examination and the history alone may reveal

12

491

SUMMARY 2 Treatment of acute pulmonary oedema Emergency measures • Sit patient upright, with the legs over the edge of the bed if necessary • Oxygen by 28% mask at 5-10L/min; non-invasive ventilation, it necessary • Reassure the patient whilst drugs are prepared • Apply one or two 10mg GTN patches • i.v. access by large-gauge cannula in peripheral vein (12-17 gauge in forearm or brachial vein) • i.v. bolus furosemide (frusemide) 40-80 mg or bumetanide 1-2 mg (higher dose only if previously known to require large doses, such as in renal failure patients) • i.v. bolus diamorphine 1-5 mg plus 10mg metoclopramide, if patient in pain or severe distress • Extreme hypertension can produce pulmonary oedema, especially if the pressure rise has been rapid. Lowering blood pressure has to be done progressively, and although i.v. GTN may be adequate to achieve this, more reliable and controllable pressure reduction can be performed by i.v. sodium nitroprusside infusion (40-75 g/min). The precise ideal rate of blood pressure reduction is not known. • Monitor EGG for HR & rhythm, BP, respiratory rate, oxymetry, urine output • Investigations: ECG, chest X-ray, blood gas (arterial), blood count and urea and electrolytes (venous) as emergency • Venesection of 200 ml aliquots of blood remains an effective and rapid method of alleviating pulmonary oedema in the volumeoverloaded or overtransfused patient. Pulmonary wedge pressure measurement helps to determine how much blood to remove. Stabilization period (10-20 minutes after emergency measures) • If systolic BP>100mmHg set up an i.v. infusion of nitrate: - i.v. GTN 1 mg/h (or isosorbide dinitrate). Titrate upwards to 10mg/h. Dose can be increased each 5min according to response, heart rate (aim <120) and systolic blood pressure (>90)

the cause of heart failure (valvular lesions, acute infarction, rhythm or conduction defects, pericardial effusion or constriction). Frequently, a chest X-ray, ECG, and, increasingly, echocardiography, are needed to diagnose the primary defect.

ACUTE HEART FAILURE Diagnosis and management

Pulmonary oedema Although the commonest cause of pulmonary oedema is acute left ventricular failure following acute myocardial infarction, there are many important non-cardiac causes. These include overtransfusion, shock lung in septicaemia,

O Figs 12.11,12.12

492

0 Case 12.1

If systolic BP<100mmHg patient requires CCU or ITU care for Swan-Ganz catheterization via a central vein to monitor filling pressures of the right heart and the pulmonary capillary wedge pressure (the indirect measure of LA pressure), and possibly direct intra-arterial pressure monitoring SBP<100mmHg and oliguric/anuric- consider inotropic support: - i.v. infusion dobutamine: 2.5-10 g/kg/min and dopamine infusion at 2.5-5 g/kg/min - When severe hypotension is present (SBP <70mmHg) consider use of intra-aortic balloon pump in centres well practised in its use Arterial gases po2 < 9.3kPa or pco2 > 7kPa - consider positivepressure ventilation. It is worth trying the effect of mask positivepressure (nasal cPAP) or non-invasive ventilation first. Treat arrhythmia. Fast AF, VT or atrial flutter are all best treated by emergency DC cardioversion and maintained in sinus rhythm with drugs. Dual-chamber pacing may be required to treat heart block.

Maintenance and diagnostic stage • Establish the precise cause of left ventricular failure. If LVF has complicated acute myocardial infarction, there is a need to consider the use of ACE inhibitors, -blockers when haemodynamically stable. • Echocardiography is an important investigation to establish the nature and extent of left ventricular dysfunction and valve dysfunction, or the presence of other acquired defects, such as VSD, aneurysm, thrombus. • In the majority of patients, treatment with ACE inhibitors and blockers should be started and doses titrated up cautiously. Exceptions are those with renal artery stenosis for ACEI, and true asthmatics and claudication for -blockers.

aspiration of gastric acid, paraquat poisoning, and inhalation of certain toxic fumes. The common feature is accumulation of oedema fluid in alveolar spaces, leading to dyspnoea. The lungs become stiff, increasing the work of breathing. Gas exchange is impaired and, if unrelieved, progressive hypoxaemia and hypercapnia follow as the patients tires; this can lead to death. Not all inspiratory crackles are due to pulmonary oedema. Experienced clinicians may be able to distinguish crackles of oedema from other pulmonary pathology. Crackles in the absence of upper lobe blood diversion (pulmonary venous hypertension) on chest X-ray are most unlikely to be due to pulmonary oedema. Confirmation with direct measurement of pulmonary venous pressure may rarely be necessary. Pneumonia often coexists with pulmonary oedema, especially in elderly patients. The production of discoloured sputum (other than white and frothy) should alert the clinician. The chest X-ray shows upper lobe venous engorgement and blood diversion; intralobular oedema in the small horizontal lines best seen at the lung bases (Kerley's 'B' lines); severe intra-alveolar oedema

causing acute bilateral lung shadowing often most prominent in the perihilar regions. O

Treatment Treatment is summarized in Summary 2, page 492. In acute left ventricular failure initial management consists of manoeuvres to decrease left atrial pressure, increase oxygenation, increase cardiac output and BP if significantly depressed, and diminish fear and anxiety, which increase the sympathetic drive and lead to increased myocardial work. Diamorphine and morphine have threefold beneficial effects - anxiolytic, analgesic and venodilation - but are also powerful respiratory suppressants, which may precipitate Cheyne-Stokes respiration or respiratory arrest. They should therefore be used with great caution. Venodilation may be more safely achieved with i.v. nitrates and anxiolysis by agents such as benzodiazepines. Treatment of brady- or tachyarrhythmias is crucial. Maintenance of a regular sinus rhythm is the objective in the presence of haemodynamic compromise. Withdrawal and counteraction of culprit agents (e.g. digoxin) may be necessary. Physical therapy (DC cardioversion, artificial pacing) is usually more efficacious, but appropriate pharmacotherapy is often required to maintain sinus rhythm and to avoid recurrences. In general, the reduction in preload with diuretics and nitrates is associated with some impairment of tissue perfusion (demonstrated, for example, by minor elevations in urea and creatinine). This is usually transient. If the reduction in preload by diuretics and nitrates is excessive or prolonged, tissue perfusion may become greatly compromised and the patient develop multiple organ failure (e.g. renal failure) that is more difficult to treat than the original oedema. These situations can be avoided by reducing diuretics once euhydrate, maintaining adequate systolic pressures, or using inotropic support if necessary. The use of positive inotropic agents and, to a lesser extent vasodilators and diuretics, is contraindicated when treating hypertrophic cardiomyopathy, especially of the obstructive type. © Acute right heart failure Pulmonary embolism (see p. 675), cor pulmonale (see p. 662) or acute inferior or posterior myocardial infarction are the usual causes. The features of acute right heart failure in myocardial infarction include a raised JVP associated with a low blood pressure and usually clear lung fields. The right ventricular function curve has flattened (Fig. 12.5, p. 466), so that a higher right atrial filling pressure is needed to support the same cardiac output. These patients need expansion of the circulating volume with synthetic plasma expanders to improve output and blood pressure. Injudicious use of diuretics or nitrates to 'treat' the high JVP can lead to serious falls in cardiac output, and precipitate shock. Acute tamponade Cardiac tamponade presents as acute heart failure, and is described on p. 587.

Cardiogenic shock In cardiogenic shock severe cardiac pump dysfunction causes tissue perfusion to be so inadequate as to fail to meet basal metabolic demands. Clinically it is recognized by hypotension (systolic BP <80mmHg or mean BP <60 in normotensives, and <110 in chronically hypertensive subjects), associated with oliguria or anuria, cold sweaty skin, mental confusion and lactic acidosis. It is not necessary for all these features to be present before making the diagnosis of cardiogenic shock. Causes of cardiogenic shock are massive myocardial infarction (usually loss of >40% of total myocardial muscle), mechanical complication of infarction (e.g. VSD and papillary muscle dysfunction or rupture), severe brady- or tachyarrhythmia, rejection in the transplanted heart, myocarditis, and infiltration of the previously normal myocardium. It is important to exclude the effects of any myocardial depressant or vasodilatory agents (e.g. excessive diamorphine; -blockers or verapamil. In the presence of right ventricular infarction, a shift of blood volume centrally by elevation or the application of positive pressure to the lower body, or a fluid challenge to raise the pulmonary wedge pressure to >15mmHg should be made before making the diagnosis of cardiogenic shock. The incidence of cardiogenic shock in patients admitted with acute myocardial infarction is 5-10%. The advent of thrombolytic therapy has not altered this significantly. The hospital mortality rate is still 50-70%. The continuation of function in the heart is also dependent on its own output. When the aortic pressure falls below the critical pressure for coronary perfusion (mean 60mmHg), left ventricular myocardium becomes more ischaemic, further impairing performance. Medical therapy of cardiogenic shock is described on page 564. The following general principles are the basis of management:

12

• Optimizing heart rate, ideally between 80 and 130bpm, with pacing if necessary (preferably AV sequential pacing; see p. 504). • Treating tachyarrhythmia using agents with minimal negative inotropic effects. Digoxin is still useful in fast AF. Give 500 g in 100 mL 5% dextrose over 30 minutes. If required, this may be repeated over 1 hour. A third dose over 2 hours is rarely given. Amiodarone i.v. is frequently used but requires central venous access to avoid serious thrombophlebitis. Give 300 mg (<5mg/kg) in 5%

SUMMARY 3 Clinical features of cardiogenic shock Hypotension (<60mmHg mean aortic pressure) Oliguria/anuria (<60mL/h) Cold sweaty skin Mental confusion Lactic acidosis Biventricular heart failure is the rule

493

•

•

• •

dextrose in 1 hour, 900 mg i.v. total in 2-4 hours, with a maintenance infusion of 1200 mg in 24 hours. Directcurrent cardioversion is preferred over 'cocktails' of antiarrhythmic drugs, provided the patient can be maintained in sinus rhythm afterwards. Optimizing LV filling pressure by monitoring PA wedge pressure using a balloon-tipped flotation pulmonary artery catheter (Swan-Ganz) and aiming to maintain PAW pressure between 15 and 20mmHg by balancing diuresis and fluid intake. Correcting metabolic acidosis with bicarbonate is rarely necessary. Intermittent positive-pressure ventilation, either non-invasively or via endotracheal intubation, may be required (although this may occasionally worsen the haemodynamic status). Avoiding pure vasodilators (e.g. nitroprusside, ACE inhibitors), negative inotropic agents (e.g. -blockers, verapamil, disopyramide). Using positive inotropic agents (see p. 497).

In managing cardiogenic shock it is important to investigate whether the heart, which is obviously depressed in the basal resting state, possesses any reserve function, and whether this reserve exceeds that necessary to sustain normal body basal metabolism (e.g. peak cardiac power output >1 W, see Fig. 12.25, p. 489). Cardiac reserve can be measured by stepwise infusion of dobutamine and noting the peak cardiac power output attained. If the peak is >1 W, medical therapy, including the use of inotropic agents to maintain adequate pressure and flow, can be sufficient. Nevertheless, remediable defects should be dealt with. If the peak is below 1W (or simple bedside observation of peak attainable systolic BP is <90mmHg despite maximal dobutamine stimulation), then death is expected unless urgent and appropriate definitive treatment of primary defects is obtained (e.g. urgent mitral valve replacement, coronary angioplasty, bypass surgery, other curative operation) or insertion of an artificial mechanical heart pump or cardiac transplantation. If the definitive treatment is delayed, artificial circulatory support in the form of intra-aortic balloon counterpulsation or ventricular assist devices should be instituted (p. 520). If the surgical option is ruled out, frantic therapeutic efforts should be avoided and the patient allowed to die peacefully.

SEVERE CONGESTIVE CARDIAC FAILURE

494

Many patients, particularly the elderly, are admitted to hospital with acute exacerbations of chronic heart failure. As in all cases, whether acute or chronic, the cause must be established to allow a rational approach to therapy. A frequently encountered syndrome consists of a patient with a dilated, poorly contracting ventricle, who suffers increasing degrees of fluid retention, sometimes to enormous proportions. Gross leg oedema may prevent walking, tense oedematous skin with poor perfusion is subject to accidental laceration, infection and the development of cel-

lulitis and venous thrombosis, both superficial and deep. Ascites may be present and contribute to poor breathing and orthopnoea due to diaphragmatic splinting. Associated oedema of the intestinal mucosa may produce diarrhoea, but more frequently accounts for variably poor absorption of drugs (especially diuretics). Many of these patients have been previously treated with diuretics and digoxin, so that electrolyte disturbance and digitalis toxicity may complicate the picture. The mainstay of initial management in hospital is to generate a gradual and progressive loss of oedema fluid, best monitored by daily weighing. A rate of weight loss of 0.5-1.5 kg per day is ideal. This is best achieved by using intravenous diuretics combined with fluid restriction to 1.5L/day and a reduction in salt intake. Daily or twicedaily bolus injections of frusemide 40-160mg or bumetanide 1-4 mg are common (patients with moderate to severe renal impairment may need very much higher doses, e.g. 120-500mg furosemide (frusemide)). Electrolytes are maintained by simultaneously prescribing a potassium-sparing diuretic (amiloride 5-10 mg daily, or spironolactone 25-100mg twice daily). Occasional patients are resistant to these measures. In these cases adding a thiazide diuretic orally for several days may initiate the diuresis (e.g. metolazone 5-10mg/day, or bendroflumethiazide (bendrofluazide) 2.5-5 mg/day). An alternative method is to infuse dopamine (2.5-5.0 g/kg/min) via a central line to avoid thrombophlebitis. Very rarely haemofiltration, ultrafiltration or dialysis may be needed to relieve the patient of intractable fluid congestion. Hypokalaemia needs to be corrected in patients who have been on long-term diuretics. Because of a simultaneous loss of magnesium, replacing the magnesium by infusion often helps restore potassium equilibrium and diminishes the tendency to arrhythmia. ACE inhibitors may need to be withheld during the acute phases of treating congestion because of their known effects of vasodilating the efferent glomerular arterioles, thereby probably diminishing GFR. ACE inhibition is (re)started as congestion resolves, unless the patient is hyponatraemic. Perindopril, because of its less first-dose hypotensive action, is the drug of first choice. After a test dose of 2mg the patient is kept under observation; blood pressure is checked every 15 minutes for 90 minutes, and 2-hourly thereafter. Provided that profound or symptomatic hypotension does not occur, maintenance perindopril or another ACE inhibitor is started and blood pressure checks revert to the routine 4-hourly observations. Uptitration of dose to the highest tolerated may be undertaken prior to discharge. Digoxin still has a role even in sinus rhythm; digitalization by 0.75-1 .Omg in the first 24 hours is followed by an appropriate maintenance regimen (125-250 g/day) monitored by plasma levels. Renal function should be closely monitored; overintense diuresis may decrease preload (and hence cardiac output and tissue perfusion) so much that prerenal failure is precipitated. It is worth emphasizing that medical

management of heart failure is almost impossible without functioning kidneys. Preservation of renal function should therefore be a high priority. Maintenance of systolic blood pressure above lOOmmHg is important and helps to avoid serious derangement of renal function. During the bed-rest stage anticoagulation with low molecular weight heparin by subcutaneous injection should be used. Once a stable balance is achieved, oral diuretics replace i.v. treatment. ACE inhibitors have a diuretic 'sparing' action, so that diuretic doses may well have to be reduced some days or weeks after starting ACE inhibitors, otherwise the patient may become dehydrated and develop fatigue, postural hypotension and worsening renal impairment.

CHRONIC HEART FAILURE IN OUTPATIENTS Management - general measures The cause of heart failure must be established and remediable causes sought and treated. These might include the replacement or repair of valves, stripping of the pericardium in constriction, or the closure of significant intracardiac shunts. The essential investigations are outlined in Summary 4. The overall aim of treatment in chronic (non-surgical) heart failure is to relieve symptoms, to improve quality of life and functional capacity, and lastly to improve longevity. The general measures should include stopping smoking, reducing alcohol intake, fluid restriction, and possibly regular exercise training. Bed rest is helpful only in acute myocardial infarction, active myocarditis, infective endocarditis and intractable congestion unresponsive to conventional doses of diuretics (see above). The disadvantages of bed rest (skeletal and myocardial deconditioning, muscular atrophy and weakness, bedsores, deep vein thrombosis, autonomic maladjustment) outweigh the benefits. Patients who are asymptomatic at rest should be encouraged to mobilize as soon as possible, as exercise rehabilitation may improve functional capacity even in severe chronic heart failure. Drug therapy to improve prognosis is now possible, and (3-blockers, ACE inhibitors, spironolactone and nitratehydralazine combinations are the most effective, although the prolongation of survival only averages 6 months. About

SUMMARY 4 Essential investigations in chronic heart failure • ECG • Chest X-ray

• Urea, electrolytes, creatinine, thyroid function? BNP • Echocardiogram • Direct/indirect measures of cardiac functional reserve.

20-50% of patients in heart failure die suddenly. No antiarrhythmic agent has been shown to be effective in reducing this incidence, although the use of amiodarone appears promising. Patients with life-threatening arrhythmia may need AICD.

12

Drugs used in chronic heart failure Diuretics This class of drug is still the first-line treatment for symptomatic congestive cardiac failure because it is the most effective at relieving the symptoms of congestion, thereby producing the greatest impact on quality of life. They are all natriuretic and aquauretic; some of them conserve K+ and Mg2+, whereas others induce losses. Injudicious use of diuretics may lead to severe electrolyte imbalance (resulting in arrhythmias and sudden death) and hypovolaemia (resulting in tissue hypoperfusion and renal failure). Thiazide diuretics These are used principally to treat hypertension, but may occasionally have a role in controlling mild oedema and, when added to loop diuretics, may induce a diuresis when resistance to loop diuretics has arisen. Hypokalaemia and hyperuricaemia are two of the more commonly encountered problems with these drugs. To combat the former, combination with potassium-sparing agents is advised. Metolazone is a curious thiazide-like diuretic with a very powerful ability to induce diuresis when combined with loop agents in patients with renal impairment or resistance to loop diuretics. However, it is also remarkably potent in inducing electrolyte disturbances, and the most difficult one to correct is severe hyponatraemia. It should therefore be used sparingly, and stopped before hyponatraemia occurs. Bendroflumethiazide (bendrofluazide) Hydrochlorothiazide Metolazone

2.5-5.0mg per day 25-100 mg per day 5-10 mg per day

Potassium-sparing diuretics It is almost always necessary to add these agents to the treatment of patients taking diuretics alone. ACE inhibitors usually diminish the need for potassium-sparing agents and when used in this combination close monitoring of serum potassium (aimed between 4.5 and 5.5mmol/L) is essential because severe hyperkalaemia may occur. Spironolactone, an aldosterone receptor antagonist, at a dose of 25 ing/day, has been shown to be safe and extend the lives of patients with heart failure treated with loop diuretics and ACEI. Its antiandrogenic effects (e.g. gynaecomastia) may not be tolerated by some male patients, in which case amiloride may be used instead. Amiloride Spironolactone Triamterene

5-20mg/day 25-200 mg/day 50-250 mg/day

495

Loop diuretics These are the cornerstone of diuretic therapy in all but the mildest degrees of heart failure. They share with thiazides the tendency to produce metabolic disturbance and need the addition of potassium-sparing agents to avoid these pitfalls. Bumetanide Furosemide (frusemide) Torasemide

l-10mg/day 40-1000 mg/day 2.5-40 mg/day

Combination diuretics Combinations of diuretics, incorporating fixed doses of potassium-sparing agents, are very widely prescribed. It is accepted that combinations of diuretic with potassium supplements are likely to be less effective at maintaining potassium homeostasis, and unhelpful for magnesium homeostasis. Coamilofruse (Frumil) Frusene

1-2 tablets/ day 1-2

Dyazide

1-2

furosemide (frusemide) 40 + amiloride 5 furosemide (frusemide) 40 + triamterene 50 triamterene 50 + hydrochlorothiazide 25

A suggested management algorithm for chronic heart failure is illustrated in Figure 12.26. Vasodilators Vasodilator drugs are used to decrease both preload and afterload. Excessive preload reduction may result in loss of Starling's effect and compromise ventricular output. Excessive afterload reduction can cause hypotension, reduce coronary perfusion, and exacerbate myocardial ischaemia and cerebral hypoperfusion, leading to dizziness and syncope. ACE inhibitors ACE inhibitors are the vasodilators of first choice in the treatment of chronic moderate and severe heart failure. There are many ACE inhibitors available: Captopril Enalapril

6.25-50 mg t.d.s. 2.5-20mg b.d.

Lisinopril Perindopril

5-40mg o.d. 2-4 mg o.d.

Ramipril

1.25-5 mg b.d.

Short-acting Intermediate duration of action Long duration of action Long-acting, least firstdose hypotension Long-acting

It is still unclear whether the different properties exhibited by the different types make any significant difference to their clinical efficacy.

496

• Up to 10% of all patients do not tolerate ACE inhibitor therapy owing to unacceptable coughing. • First-dose hypotension at the start of ACE inhibitor therapy can occur unexpectedly in patients with chronic heart failure, especially in those with unsuspected renal artery stenosis. Certain precautions should be heeded.

ACE introduction should be avoided if the patient is relatively hypovolaemic (as judged clinically by the absence of oedema, low JVP, dehydration, postural hypotension, elevated serum urea with normal serum creatinine), in which case the diuretics should be withheld for a day or two before introduction. Patients with significant renal artery stenosis can present with the syndrome of heart failure, and ACE inhibition is contraindicated unless medical autonephrectomy is indicated. Patients with markedly elevated blood pressure (especially acutely) despite features of heart failure should be investigated to rule out renal artery stenosis before the introduction of ACE inhibitors. Sometimes hospital admission for close BP monitoring may be required. A suitable regimen might start with 2mg of perindopril, as this has been shown to be least likely to cause first-dose hypotension, and the dose subsequently increased to 4 nig o.d. or switched to another ACE inhibitor. • The most problematic adverse effect is impairment of renal function. In a significant minority inhibition of angiotensin-induced constriction of the efferent glomerular arteriole and overall renal function is affected. Rising creatinine (>200 mol/L) is a reason to withdraw this therapy. Nitrates In patients with heart failure secondary to coronary heart disease nitrates have a distinct role, particularly as ACE inhibitors do not have any specific antianginal activity. Long-acting nitrates are used, leaving a 6-8-hour nitratefree period to avoid tolerance to the haemodynamic effects of nitrates. Isosorbide mononitrate modified release

40-120 mg o.d.

Other agents The combination of high-dose nitrate with hydralazine (50-75 mg q.d.s.) has been shown to improve function and prognosis for severe heart failure, but the drug combination can be difficult to tolerate. Calcium antagonists are not generally useful in heart failure. p-Blockers With indication for use in heart failure, p-blockers can now be claimed to be a ubiquitous cardiological drug. Trial data have shown that they can improve the prognosis, in terms of survival and hospitalization. There are no conclusive data to show that they improve functional capacity or quality of life. Clinician concerns (which led to these drugs being previously categorized as contraindicated for use in heart failure) are still valid, as inappropriate use in some patients may precipitate exacerbation of heart failure. It is therefore vital that -blockers should be introduced by trained personnel, starting with the lowest doses, which may be increased very gradually if tolerated. Carvidelol is a non-cardioselective -blocker with ocreceptor blocking activity, which produces vasodilatation. It was the first to be reported to provide remarkable survival improvements in heart failure.

12 Blood tests: Biochem FBC TFT Chest X-ray ECG Urine dipstix Cause known? eg. previous Ml previous myocarditis elderly patient with BP

Another cause for symptoms No

No

Yes

Echo Yes

Regional ?IHD

Surgical cause e.g. aortic stenosis pericardial constriction

Global 3 cardiomyopathy

Look for reversibilit; e.g. FDG nuclear medicine scan Cardiac catheter

Revascularization for CAD

Primarily systolic failure? e.g. MI/IHD alcohol myocarditis or diastolic failure? e.g. hypertension hypertrophy HCM

No

Don't know

Echo

Yes

Surgical candidate?

No

Yes

REFER

Caution with diuretic ACE

Diuretic ACE I

May still need antifailure treatment

Sinus rhythm

?p-Blocker verapamil

No

YES

If severe HF

AF

Bradycardia & heart block

Tachyarrhythmia

Dual-chamber pacemaker

Antiarrhythmic agent ±AICD

FIG. 12.26 Management of chronic heart failure

Not all -blockers are equally prognostically beneficial to heart failure patients. Bucindolol has recently been shown not to confer similar benefit. Bisoprolol and metoprolol have been shown to be beneficial in improving survival.

Inotropic agents Positive inotropic agents are not without potential dangers (e.g. arrhythmogenicity, and liability to induce worsening cardiac myocyte necrosis). In viral myocarditis they should

497

TABLE 12.8 Digoxin dosage Digoxin • Loading dose 15 g/kg in 3 doses every 6 hours • Maintenance doses 62.5-3.75 go.d. depending on age and renal function • Dose adjusted according to plasma levels (therapeutic range 1.0-2.6 mmol/L) Digitoxin Mainly non-renal metabolism. Very long half-life and rarely used

be avoided, unless as a bridge to transplantation. With the exception of digoxin (see below and Table 12.8), chronic oral inotropic therapy (with 2 stimulants and phosphodiesterase inhibitors) has been very disappointing. Digitalis preparations (digoxin, medigoxin, digitoxin, ouabain) are now well established as of value as mild positive inotropic agents in chronic heart failure. Digoxin is still the only oral positive inotropic agent available. Antiarrhythmic agents These drugs often have to be used in heart failure patients. Most intravenous antiarrhythmic agents are cardiodepressant to varying degrees. The most negatively inotropic are p-blockers, verapamil and disopyramide, and these are generally avoided in severe heart failure. In therapeutic doses the following intravenous agents exert little negative inotropism, and may be used in heart failure: lidocaine (lignocaine), mexiletine and procainamide. Orally, amiodarone, digoxin, mexiletine and procainamide are virtually devoid of negative inotropic effects. In heart failure the most effective and least proarrhythmic antiarrhythmic agents for supraventricular and ventricular tachycardia are -blockers and amiodarone. The oral preparation of amiodarone is virtually devoid of negative inotropism and is therefore suitable for use in patients with severe heart failure who are intolerant of (3blockers. Their concurrent use with amiodarone requires a reduction in dose of digoxin and warfarin. O

CARDIAC TRANSPLANTATION Since the first human heart transplant in 1967, the technique has become an accepted form of treatment for patients dying of heart failure and for whom no other form of treatment offers any help. The current annual rate of cardiac transplantation worldwide is approximately 3000. The number of patients requiring transplantation each year in the USA alone is estimated to be 35 000-70000. The shortfall is mainly due to the shortage of donor hearts, and

Q MCQ12.6

498

is responsible for the plateauing of the transplantation rate in the USA and Europe. Cardiac transplantation will not become the panacea of heart failure treatment, and selection of recipients will necessarily remain strict.

Indications The vast majority of patients who undergo cardiac transplantation have either terminal idiopathic dilated or ischaemic cardiomyopathy. Other indications include peripartum cardiomyopathy, congenital heart disease and cardiac tumours. The patient must have reached end-stage heart failure and have a very limited life expectancy. The best available method of measuring the prognosis of these patients is estimation of their cardiac reserve (see p. 494). Contraindications to transplantation of the heart alone include increased pulmonary vascular resistance, blood group incompatibility, or any coexisting systemic illness that may significantly limit life expectancy.

Surgery There are two main surgical procedures: • Orthotopic transplantation. The recipient's heart (except the venous attachment side of the atria) is excised and replaced by the donor's heart. This is the commonest procedure. • Heterotopic transplantation. The recipient's heart is not excised and the donor's heart is anastomosed side to side to the corresponding atria and great vessels of the recipient's heart. Contrary to popular belief, the surgery itself is simpler than most open heart surgical procedures. The main operative problem is the preservation of the donor heart during transit.

Management By far the most difficult part of cardiac transplantation is the postoperative care. The two major complications are graft rejection and infection, which account for most of the early mortality. There is as yet no early and accurate noninvasive method of detecting rejection. Apart from clinical suspicion and loss of R wave in the ECG, the most reliable method of diagnosing rejection is by percutaneous endomyocardial biopsy. Immunosuppressive regimens include combinations of ciclosporin, azathioprine, corticosteroids, cyclophosphamide, antithymocyte globulin (ATG), use of monoclonal antibody against the CD3 molecule on mature T cells, and occasionally vincristine or methotrexate. The need for immunosuppression means that patients are at increased risk from bacterial infection. Opportunistic fungal infections and reactivation of primary viral infection (notably cytomegalovirus) occur. Late complications include accelerated coronary disease in the donor heart, hypertension, and an increased incidence of malignancies (especially

lymphoproliferative), and bone diseases (osteoporosis, a vascular necrosis).

Prognosis Current survival rates of cardiac transplantation are as good as those of renal transplantation, with overall 1-

and 5-year survival rates of about 80% and 70%, respectively, for orthotopic transplantation. The survival rate for heterotopic transplantation is less good, with a 5-year survival of about 50%. Factors that negatively influence prognosis include major HLA mismatch, prolonged cold preservation of the donor heart, the presence of preformed circulating antibodies and advanced donor age.

12

CASE STUDY 12.1 ACUTE ONSET OF BREATHLESSNESS AND ARRHYTHMIA A 66-year-old man presented to his GP feeling generally unwell and breathless. He felt he had 'flu'. Initially he had felt feverish with mild nausea and diarrhoea. This had been some weeks before his current presentation, which consisted of significant limitation in exercise tolerance. Anything other than minor exertion produced breathlessness without chest pain. His sleep was very poor. He would fall asleep then wake in the middle of the night with breathlessness. His abdomen felt bloated and he was unable to eat comfortably. He was aware that his heartbeat was much more forceful, fast and irregular. In the past he had had no significant medical history except for being a heavy cigarette smoker which he had managed to stop 10 years prior to presentation. Twenty years previously he had had a period of excess alcohol consumption, but this had been resolved and he was now teetotal. He was taking no medication. The GP found him to be dyspnoeic on minimal exertion. He had a pulse rate of 104 which was irregularly irregular. Blood pressure was 110/70. The jugular venous pressure (JVP) was elevated to the angle of the jaw and he noted mild ankle swelling and some abdominal distension. His heart sounds were rapid on auscultation and there was a suspicion of a systolic murmur. The chest revealed bilateral absent breath sounds and basal dullness above which there were a few fine crepitations. The GP referred him up to the cardiology clinic with a diagnosis of cardiac failure and atrial fibrillation.

Questions 1. Is this diagnosis appropriate? What differential diagnoses should be considered? 2. What investigations should be arranged? 3. What would be the principles of treatment for this patient assuming that there were no surgically remediable cause for his heart failure?

Discussion Heart failure is a clinical syndrome which was diagnosed confidently and correctly by the GP. The features in support were the symptoms of exercise limitation. His nocturnal symptoms could represent orthopnoea with some features of paroxysmal nocturnal dyspnoea (PND). The GP noted the findings consistent with right heart failure in that the JVP was elevated and he had peripheral oedema. The abdominal distension could represent ascites and possible enlargement of his liver. He had dull lung bases which could be bilateral pleural effusions and the crepitations heard above this area of dullness could be evidence of left heart failure producing pulmonary oedema. He was noted to have a fast irregular pulse most likely due to atrial fibrillation complicating the heart failure or indeed contributing to it. There was a possible systolic murmur. Having detected the clinical syndrome, it is important for the doctor to establish remediable causes. These might include non-cardiac causes of heart failure such as thyrotoxicosis, anaemia and excessive

alcohol intake. Cardiac causes of heart failure that need to be excluded would be ischaemic heart disease, valvular disease or pericardial disease. Investigations are designed to establish a diagnosis behind the presenting clinical syndrome of heart failure. Blood tests will exclude anaemia and thyrotoxicosis. The plasma biochemical profile should include renal and liver function to give baseline values because treatment may affect renal function adversely. When the presentation is relatively acute it is important to include cardiac enzymes in the investigation to exclude silent myocardial infarctions as a possible cause of heart failure. An ECG will confirm the presence of atrial fibrillation and also give clues to the aetiology of heart failure particularly if there are any changes consistent with either acute or chronic ischaemic heart disease. Chest X-ray would confirm the presence of pleural effusions and the clinical suggestion of pulmonary oedema. Calcification around the pericardial sac, or elsewhere in the heart, might be seen. The size of the cardiac silhouette and any contribution from coexistent pulmonary disease will be established. An echocardiogram can be used to establish whether the heart failure relates to valvular disease or is a primary problem of the heart muscle. In this patient the problem is likely to be due to viral myocarditis leading to the appearance of thin-walled, very poorly functioning left and right ventricles. A degree of mitral regurgitation was noted secondary to

499

CASE STUDY 12.1

CONTINUED

left ventricular dilatation and this accounted for the systolic murmur. In a man of this age, exclusion of myocardial ischaemia is important even in the absence of any other pointers such as chest pain or ECG change. This is because occasional silent presentations of the severe ischaemia can be adequately treated by revascularization, usually by cardiac bypass surgery. Consideration therefore would be given, in this age group, to a myocardial perfusion scan using thallium and to cardiac catheterization with coronary angiography. The treatment would begin by management of the acute symptoms produced by his presumed post-viral myocarditis. Fluid retention would be

treated by the use of loop diuretics such as furosemide (frusemide) with the addition of the potassium-sparing diuretic spironolactone. His atrial fibrillation should be treated. The heart rate should be controlled with digoxin and anticoagulation with warfarin is indicated. After adequate anticoagulation subsequent DC cardioversion to sinus rhythm would be an option if mitral regurgitation was not severe and if the left atrium was not severely dilated. ACE inhibitors should be introduced as they improve survival from myocardial failure and will allow the use of lower doses of diuretics in the long term. In addition -blockers should be introduced into treatment, although careful titration starting at

Heart-lung transplantation Patients with end-stage pulmonary vascular hypertension (primary or Eisenmenger's syndrome) and with parenchymal lung disease (mostly cystic fibrosis) have the option of heart and lung transplantation. The two classes of disease each constitute about half of the transplanted cases. The 1- and 3-year survival rates are currently about 65% and 50%. A major difficulty in management is that the diagnosis of rejection is more difficult in the lung than in the heart. FURTHER READING ON TRANSPLANTATION O'Connell J B, Bourge R C, Costanzo-Nordin M R, Driscoll D J, Morgan J P, Rose E A, Uretsky B F 1992 Cardiac transplantation: recipient selection, donor procurement, and medical follow-up. A statement for health professionals from the Committee on Cardiac Transplantation of the Council on Clinical Cardiology, America Heart Association. Circulation 86:1061-1079.

ARRHYTHMIA Arrhythmia is a general term for any cardiac rhythm or conduction abnormality. Anatomically the arrhythmia

O Fig. 12.13

500

very low doses is necessary in this context. The particular -blockers of use in this situation are carvedilol and propranolol. In randomized trials both of these drugs have shown improvement in prognosis when given in heart failure. Cardiac transplantation is an option although in this age group the success rate is greatly diminished. For this reason some centres do not accept patients in this age group for transplantation. Acute post-viral cardiomyopathy has been treated successfully by temporary mechanical ventricular assistance devices. This patient's chronic and slightly less severe problem is unlikely to be suitable for this management.

may originate supraventricularly (atria or AV junction) or ventricularly. The most compelling clinical feature is whether the arrhythmia causes haemodynamic disturbance; however, its implications depend more on the type of arrhythmia, so that ventricular tachycardias have a poorer prognosis than supraventricular tachychardias.

INVESTIGATIVE TECHNIQUES The principal method of analysis of arrhythmia is the 12-lead ECG. The correct interpretation of the ECG may depend on comparison with previously acquired recordings.

24-hour ECG (Holter monitor) Recorders can monitor two channels of an ECG in ambulant patients over a continuous 24-hour period (Fig. 12.27). Some rhythm disturbances may occur very infrequently and require alternative equipment. Event recorders can be applied by the patient during symptoms and record a few seconds of ECG in a solid-state memory. This can subsequently be replayed or transmitted by telephone to a central station, where a doctor can advise on the nature of the event. For those patients who receive no warning of impending rhythm disturbance, there are newer implantable automatic solid-state recorders (ILR, implantable loop recorders) that sample the ECG over long periods but only memorize significant events, or events triggered by the patient.

but not the ventricular, activity; this is most commonly encountered in complete heart block (p. 503 and Fig. 12.31). Sinus arrhythmia is the quickening of the heart rate that occurs cyclically with each inspiration. It is most marked in individuals (often young) with high vagal tone.

12

EGTOPIC OR EXTRASYSTOLIC BEATS These are defined as beats that arise outside the normal heart pacemaker. They may be generated in the atria, the AV node, the bundle of His or the ventricles. They may also arise as 'escape' beats in sinus bradycardia.

Atrial ectopics FIG. 12.27 24-hour continuous ambulatory ECG monitoring At 19.20 hours the sinus rhythm is interrupted by ventricular ectopic beats and a short run of ventricular tachycardia. O

Exercise EGG The exercise test has a role in managing certain patients with arrhythmia. Those with recurrent ventricular tachycardia, or Wolff-Parkinson-White syndrome, may undergo exercise testing to check the efficacy of their drug treatment. However, not all these patients have exerciseinduced arrhythmia and may instead require more sophisticated electrophysiological (EP) studies.

Intracardiac electrophysiological testing The techniques of cardiac catheterization enable localized intracardiac ECGs to be recorded, using catheters similar to pacing catheters but with electrodes mounted distally. The activity of the atria, AV node, bundle of His and ventricles can be recorded independently. By combining recording with artificial stimulation from intracardiac pacemakers, a detailed analysis of intracardiac conduction can be made. This has proved useful in studying patients with tachycardia, in whom the abnormal rhythm may be induced and its origin (ventricular or supraventricular) confirmed. The site, number and potential danger from accessory AV bypass tracts (e.g. in Wolff-Parkinson-White syndrome) can be mapped and the tracts ablated. The effectiveness of drug regimens and the potential for using antitachycardia pacemakers are also tested using EP studies.

NORMAL SINUS RHYTHM Each ventricular (QRS) complex is preceded by a P wave with a normal P-wave axis (i.e. from the sinus node, P-wave axis = 0° to +90°) and normal PR interval (0.11-0.20s). In some circumstances sinus rhythm may control the atrial,

Atrial ectopics or atrial premature beats (Fig. 12.28A) are usually preceded by a P wave of different shape, different PR interval and, frequently, a different axis to the sinus P wave. The following QRS complex is normal. Atrial ectopics are common even in normal hearts, although, like all extrasystoles, they may be felt as uncomfortable fluttering, palpitation or 'skipped beats' in the chest. Stimulants, such as alcohol or caffeine-containing drinks and cigarettes, may be precipitants.

Junctional ectopics Junctional ectopics may arise from the AV node or proximal bundle of His, although this differentiation cannot normally be made, hence the less specific term 'Junctional' rather than the older 'nodal beat'. Atrial activation occurs retrogradely, so that the P wave is positive in aVR and negative in II, III and aVF. The delay between ectopic activation and ventricular activity depends on its site of origin and will vary the timing of the P wave. This can precede, be buried in, or follow the QRS complex, which has a configuration identical to that in sinus rhythm.

Ventricular ectopics Ventricular ectopic (VE) beats, by definition, are not preceded by a P wave and have a broad QRS configuration (Fig. 12.28B). There is no prognostic significance attached to whether the ectopic beat has a left or right bundle branch block pattern. However, when frequent or multiform VEs are found in association with underlying coronary heart disease or cardiomyopathy, there is an increased risk of sudden death. Control of VEs with drugs may improve symptoms, but may not improve the prognosis; the severity of the underlying heart disease is probably a more important influence on survival. In addition, the drugs used to suppress VEs may themselves precipitate serious ventricular arrhythmias.

Clinical features Ectopic beats, whether from atria or ventricles, are commonly found in normal individuals and generally not per-

501

FIG. 12.28 ECGs showing ectopic beats fAl An atrial ectopic beat, demonstrated by the varying shape of the P wave. B Coupled ventricular ectopics (bigeminy).

ceived by the patient. In some patients the frequency of ectopics is increased and they may produce the feelings of a 'missed beat' or even atypical chest pain (see p. 467). Ectopics are frequently more obvious to the patient at rest, presumably owing to the relative bradycardia allowing an ectopic focus to activate. Suppression with exercise that increases the heart rate is common and a good prognostic sign. Explanation of the cause of the symptom and avoidance of certain common precipitants of ectopics, such as alcohol, caffeine and cigarette smoke, may be all that is needed to treat the patient. Some patients improve when given (3blockers (e.g. metoprolol 25-100mg b.d., sotalol 20-80mg b.d. or bisoprolol 2.5-5.0mg o.d.)

BRADYCARDIA AND HEART BLOCK Bradycardia Defined as a heart rate less than 60bpm, bradycardia may be due to vagal tone in athletes or, more commonly, be an unwanted effect of -adrenoreceptor-blocking drugs. It generally requires no treatment, unless severe and producing symptoms. In an emergency (e.g. severe bradycardia complicating a myocardial infarction), it responds to atropine (600 g i.v. repeated to a maximum of 2.4 mg in 2 hours). If persistent, a temporary artificial pacemaker should be implanted. Chronic bradycardia may be a manifestation of the sick sinus syndrome (p. 516) and may benefit from permanent pacemaker implantation.

Sinus node block and sinus arrest Both sinus node block and sinus arrest are associated with a pause in the normal train of sinus rhythm, leading to a delay in the appearance of the P wave. In sinus node block the resultant pause between P waves is approximately double the preceding sinus (P-P) interval. The essential pacemaker rate remains constant but a normal discharge does not occur. In sinus arrest the pause is not a Fig. 12.14

502

FIG. 12.29 ECG showing first-degree heart block Note prolonged PR interval (0.28s).

multiple of the sinus interval, implying a transient change in pacemaker automaticity which has led to the loss of the P wave. Chronotropic incompetence is a poorly defined condition affecting some patients whereby they are unable to increase their heart rate in response to exercise. Their maximum heart rate may reach only 75-80% of the agepredicted maximum (220/age). Atrial rate-responsive pacemakers are indicated in this relatively uncommon condition. Many cardiac drugs may produce sinus node dysfunction, particularly those used as antiarrhythmics. (3Adrenoceptor blockers, the calcium antagonists verapamil and diltiazem, as well as amiodarone, propafenone and flecainide, are all potential causes. Occasionally cimetidine, lithium and phenytoin may interfere with sinus node function. Although idiopathic degeneration of the sinus node, particularly with ageing, may be the commonest cause apart from these drugs, hypertension, myocardial ischaemia and rarer conditions such as sarcoid may account for sinus node dysfunction.

First-degree heart block In first-degree heart block (Fig. 12.29) the PR interval is prolonged (>0.20s). This may occur in otherwise normal patients and is caused by delayed propagation within the AV node. A modestly prolonged PR interval may also accompany vagally induced bradycardia. However, long PR intervals in the presence of a normal or fast heart rate imply an intrinsic defect in the AV node. This is generally benign. The development of first-degree heart block in the context of infective endocarditis usually means aortic root abscess formation. Similarly, widespread conducting tissue disease is suggested by the combination of firstdegree block, left axis deviation and bundle branch block. Symptomatic patients may require a pacemaker. Firstdegree block may also occur following myocardial infarction (p. 562).

Second-degree heart block Second-degree heart block (Fig. 12.30) is defined as intermittent failure to conduct an impulse from the atria to the ventricles, and may be temporary or permanent. It occurs in two major forms. The Wenckebach phenomenon or Mobitz type I block The Mobitz type I block (Wenckebach phenomenon)

12

FIG. 12.30 ECGs showing second-degree heart block S3 Mobitz type I showing progressive lengthening of the PR interval until the P wave is not conducted to the ventricles (arrow). B Mobitz type II. Four sinus beats are followed by cessation of conduction following the P wave (arrowed).

produces successively increasing PR intervals until a P wave is not conducted; the cycle then repeats itself. There are usually 3-5 beats in a cycle. This block occurs within the AV node and is probably relatively benign. It may complicate acute inferior myocardial infarction and generally does not require specific treatment. It may also be seen in patients with generalized conducting tissue disease, and as part of the sick sinus syndrome (p. 516).

FIG. 12.31 Three rhythm strips demonstrating complete heart block A Atrioventricular dissociation with a relatively rapid ventricular rate and narrow QRS complex (with acute inferior infarction). B A slow idioventricular rate with broad QRS complexes, [c] The unusual combination of complete heart block and atrial fibrillation (no P waves).

Mobitz type II block In the Mobitz type II block an unheralded failure to conduct a P wave occurs. The site of the block is below the AV node in the proximal bundle of His. This condition has a tendency to progress unpredictably to complete block, which may cause Stokes-Adams attacks or even death. This conduction disturbance is much less common than type I block, and because of its implications usually requires treatment with an artificial pacemaker.

may be within the AV node, when the escape ventricular rhythm will be within the lower regions of the node or bundle of His, giving narrow, normal configuration QRS complexes and heart rates of about 40bpm. Complete heart block may also occur at the level of the bundle of His, so that ventricular or distal bundle pacemakers control the heart rate, producing a much slower pulse (about 30bpm) and wide bundle branch block QRS complexes. O

Other special forms of second-degree block During atrial tachycardia (e.g. atrial flutter) there may often be conduction of alternate beats to the ventricles in a ratio of 2 P waves to 1 QRS complex. When a tachycardia is present this block constitutes the normal physiological function of the AV node, and the block can often be increased further (to 3:1 or 4:1 or more) by manoeuvres that stimulate vagal activity, such as carotid sinus massage. In the absence of an atrial tachycardia, established 2:1 conduction represents a form of second-degree block. Highgrade block is present when the ratio of P waves to QRS is 3:1 or greater. In either case, artificial permanent pacemaker implantation is the treatment of choice when the arrhythmia persists.

Clinical features

Complete or third-degree heart block In complete or third-degree block (Fig. 12.31) the atrial impulses fail to be conducted to the ventricles. The block

Symptoms First-degree heart block is asymptomatic, as are many cases of second-degree block, but patients may be aware of the dropped beat and the stronger post-pause beat in Mobitz types I and II. In complete heart block the most dramatic symptoms are related to recurrent asystole or the ventricular arrhythmias that complicate severe bradycardia. However, complete heart block may present without symptoms, despite a marked bradycardia; with heart failure secondary to a low cardiac output; and in the elderly as a confusional state secondary to poor cerebral perfusion and multiple syncopal episodes. Stokes-Adams attack Stokes-Adams attacks can be caused by transient asystole or a transient burst of tachycardia. They are usually unheralded and the patient may first be aware of lying on the ground or falling. Recovery is usually rapid and complete.

503

Skin pallor may be followed by flushing. If the anoxia lasts more than a few seconds a convulsion may occur and, on occasion, incontinence. Incomplete attacks are common, presenting with transient giddiness (without true vertigo) or near fainting (presyncope). Attacks are sporadic and random. There may be months to years between symptoms in some patients with intermittent block, so that 24-hour ambulatory ECG recordings may have to be made repeatedly for the diagnosis to be made. Occasionally, treatment with a pacemaker is warranted on the basis of the history alone, although intracardiac electrophysiological studies may help in these patients. Signs There are few clinical signs in heart block. The first heart sound is quiet in first-degree block, and intermittent dropped beats will be felt at the pulse in second-degree block. In complete heart block a slow pulse and intermittent cannon waves will be seen in the JVP, caused by occasional simultaneous contraction of the atria and ventricles. The large stroke volume and collapsing pulse with a systolic murmur may mimic aortic valve disease.

Management In congenital complete heart block there is a surprising degree of tolerance to the slow heart rate, which may show modest rate responsiveness with exercise. Nevertheless, adults with this condition may risk sudden death, and permanent pacemaker implantation is warranted. In acute inferior myocardial infarction complete heart block is due to transient ischaemia of the SA and AV nodes, and may not need a temporary pacemaker unless symptoms of heart failure occur (p. 562). In extensive anterior myocardial infarction complete heart block is related to extensive muscle damage and carries a poor prognosis. Drug treatment of complete heart block with isoprenaline slow-release capsules is now hardly ever justified, and permanent pacemakers should be offered to virtually all patients. Like sinus node disease, idiopathic fibrotic degeneration of the AV node is the most frequent cause of heart block. It may be precipitated after ischaemic damage or secondary to calcific degenerative aortic or mitral valve disease. Aortic valve surgery and congenital heart defect repairs may cause heart block, acutely or after delays of months to years. O

THE PERMANENT ARTIFICIAL PACEMAKER The basic pacemaker consists of a small battery container which can be implanted under the skin and connected by O MCQ12.7

504

an insulated electrode to the heart. Small electrical impulses can be transmitted through the electrode to stimulate the heart to contract. Since 1960 there has been a great increase in their use, and 1 million people worldwide now live with an implanted pacemaker. The initial sole indication for pacing was the treatment of chronic complete heart block, but now a major indication is the treatment of the sick sinus syndrome (p. 516). More recently, units able to treat tachycardias and administer internal DC cardioversion have been developed. Pacemakers that respond to the patient's activity by changing the pacing rate are now in common use. Early pacemakers were fixed rate. Modern pacemakers, however, are almost exclusively demand (or standby) units, providing a pulse only when the heart rate falls below certain predetermined limits, and have battery lives of 8-15 years. They may offer considerable adjustment and analysis via transcutaneous telemetry. The complexity of cardiac pacemakers has increased for a number of reasons: • To increase battery life by reducing unnecessary pacing and by reducing output to the lowest safe limit; • To allow fine control of the pacing heart rate and, with dual-chamber devices, to restore the atrial contribution to filling, which may be very important to individuals with impaired ventricular function or valve disease; • To detect when the patient needs an increase in heart rate (e.g. with exercise); • To detect tachycardias and abort them by programmed bursts of pacing (antitachycardia). A letter coding system, used internationally, describes the type of pacemaker used. In some pacemakers a facility allows the interval between the last natural beat and the paced beat to be longer than the pacing interval. Thus, only when a pause corresponding to a very low heart rate is exceeded will the pacemaker step in and gradually accelerate to its programmed paced rate. This stops the pacemaker inserting beats for trivial slowing of the heart (e.g. during sleep).

Dual-chamber pacemakers Dual-chamber pacemakers have been developed to achieve a more physiological situation than is possible with ventricular pacing alone. These require the implantation of electrodes in the right atrium and right ventricle. They use more complex, and therefore more expensive, generators. In addition, the capacity to 'fine-tune' the multiplicity of programmable functions to each individual patient requires both time and the availability of skilled pacemaker technicians or doctors in pacing clinics. Nevertheless, these units are justified in certain groups: • Patients with AV block but normal SA activity, in whom a true physiological situation can be achieved, with normal rate increase with activity, and maintained AV sequential contraction. This is usually reserved for young or active elderly patients.

• Patients with poor ventricular function or valve disease, where the atrial contribution to filling is of prime importance. • Patients with the 'pacemaker syndrome': a small number of patients develop severe symptomatic hypotension with pacing of the right ventricle alone. • Patients with carotid sinus sensitivity. Dual-chamber pacemakers should be avoided in patients with unstable SA activity or atrial fibrillation.

Dual-ventricular pacemakers It has been recognized that improvements in cardiac output and hence patients' symptoms may be gained by pacing the left ventricle in some individuals with conduction disturbance (LBBB). Electrodes are placed in a branch of the great cardiac vein via the coronary sinus in conjunction with the usual leads in RA and RV. The pacemaker thus provides synchronized LV and RV contraction, with claimed benefits in patient exercise capacity.

Pacemaker implantation and complications Permanent pacemakers are almost always inserted by a cardiologist using local anaesthesia and full surgical sterile technique. Like temporary pacemakers (p. 562), the pacing electrode is passed down a central vein (usually the cephalic or subclavian) to the right heart. The electrode and generator box are then buried subcutaneously, above the pectoralis major on the high anterior chest wall. Complications of implantation are similar to those of temporary pacemakers. Infection is rare, but when it occurs it requires major revision of the entire system and accounts for considerable morbidity and a small mortality. Lead displacement and fracture are very rare with modem pacing systems. Specific pacemaker malfunction is now very rare. With current telemetry the pacemaker (and lead) function can be assessed transcutaneously, so that premature battery failure can be predicted, problems with lead impedance

TABLE 12.9 Pacemaker description Letter position Function Pacemaker types

1

2

3

4

Chamber paced A = atrium V = ventricle D = dual A+V

Chamber sensed A V D

Response to sensing

Rate responsiveness

T = triggered I = inhibited D

R = Rate responsive

An entry of '0' in one position denotes the absence of that function, e.g. VOO would be an 'old-fashioned' ventricular pacemaker with no sensing and hence no response to sensing - the old fixed-rate pacemaker.

(electrical resistance), current leak and failure to sense can often be diagnosed, and the unit reprogrammed to compensate for problems until replacement is required. Failure to sense the patient's electrical activity and intermittent failure to capture may sometimes cause problems. Dualchamber pacemakers, although potentially much closer to normal physiology, carry their own problems, the most serious being the capacity to generate pacemaker-induced arrhythmia. Fortunately, these problems are rare. Interference with pacemaker function by external sources of high magnetic and electromagnetic fields is less of a problem than might be expected; few problems arise in the normal environment of most patients. However, magnetic resonance imaging is contraindicated, and cautery during surgery has to be performed with care, avoiding placing the ground electrode near the generator and altering the unit to a fixed-rate mode.

12

THE TACHYCARDIAS Tachycardia is defined as a heart rate above 100 bpm, and is divided according to the origin of the pacemaker impulse. A classification is given in Table 12.10 and a guide to the use of antiarrhythmic agents in Table 12.11. A decision tree to aid in diagnosis is shown in Figure 12.32.

Mechanisms of production of tachycardias Three basic mechanisms have been established: abnormalities of automaticity; early after-depolarizations with triggered activity; and abnormal impulse conduction or re-entry. Abnormalities of automaticity Under normal circumstances only SA and AV node cells display automaticity. Abnormal automaticity may be displayed by diseased atrial and ventricular muscle. Stretch, alkalosis, temperature increases and hypoxia all increase the rate of repolarization and reduce the resting membrane potential (Vm), thereby accentuating the tendency to abnormal automaticity. Cardioactive drugs also influence heart rate and rhythm via these mechanisms. Therapeutic doses of some antiarrhythmics achieve their beneficial effects by decreasing the rate of depolarization (Table 12.11). Toxic doses may cause atrial and ventricular tachyarrhythmia by increasing the diastolic depolarization slope and lowering Vm. Early after-depolarizations ana triggered activity Following an action potential there may be a further spontaneous depolarization, which can occur early during repolarization or late. Single 'after-depolarizations' can precipitate further spontaneous beats, to produce a rapid cycle of spontaneous activity. Factors that encourage their occurrence include hypoxia, hypercapnia, high circulating

505

CASE STUDY 12.2 AN EPISODE OF COLLAPSE A 78-year-old woman presented to casualty having collapsed in the street whilst walking to a restaurant. The collapse was witnessed by her niece. They were walking at a good pace along a normal pavement on the flat. The patient was talking when she became silent and then slumped to the ground without warning. She struck her forehead on the pavement causing a small graze. She looked very pale at this moment but recovered within seconds. She was not incontinent and had no limb twitching. On recovery she was distressed but completely orientated. They attended casualty where she was found to be well orientated with no significant injuries. Physical examination was normal. She was normotensive and had a regular heart rate of 86 beats per minute with occasional pauses noted at the radial pulse. An ECG confirmed sinus rhythm, the PR interval was 334 msec and the QRS complexes were narrow and normal. During the 12-lead ECG, a 1.8 second pause in heart rhythm was detected. This appeared to be sinus arrest and the patient was not aware of this pause. During history taking she admitted that for the previous 2 weeks she had had waves of 'panic' attacks. These consisted of short-lived episodes when she felt very uncomfortable and was aware of a rapid fluttering in her chest associated with feeling faint. She was clear that the fall precipitating this admission was not preceded by any fluttering in her chest. She had seen her GP for these 'panic attacks' and

he had ordered a 24-hour halter monitor ECG. That revealed a background of sinus rhythm with a long PR interval and occasional pauses up to 2.3 seconds. Several episodes of narrow-complex, irregular tachycardia of up to 150 beats per minute were noted during the 24 hours. One of these which lasted 4 minutes was associated with the patient's symptoms in the chest.

Question What is the diagnosis and how should this patient be managed? Discussion This very fit 78-year-old woman has had a single syncopal fall which had the characteristics of a Stokes-Adams attack. This did not appear to be preceded by symptomatic arrhythmia on this occasion. Her 24-hour ECG revealed that she has episodes of tachycardia which punctuate normal sinus rhythm which itself was interrupted by occasional pauses of up to 2.3 seconds on this 24-hour recording. Her baseline ECG confirmed sinus rhythm but revealed a very long PR interval with no other conduction abnormality. This patient has the tachycardia-bradycardia or 'tachy-brady' syndrome, otherwise known as sick sinus syndrome. As is often the case, it is occurring in an otherwise healthy person with no other predisposing illnesses. Management must include the control of the tachycardia as well as

catecholamines, non-specific muscle injury and stretch. There is evidence that this activity occurs frequently and is important in producing tachycardia, particularly in the long QT syndrome (p. 517) and with drugs of the VaughanWilliams classes IA and III (Table 12.11).

Abnormal impulse conduction and re-entry

506

A wave of depolarization may sometimes circle endlessly

prevention of further syncopal episodes. It can be assumed that her collapse leading to the casualty admission represented a prolonged pause with sino-atrial arrest. Even without such a presentation, the current resting ECG indicates that it would be difficult to give her prophylactic treatment against paroxysmal atrial fibrillation (AF) because of the very long PR interval and occasional pauses. The risk would be that standard treatment of AF, such as a -blocker such as sotalol, or a class one agent such as flecainide, or a class three agent such as amiodarone, may all exacerbate the pauses and lead to the development of more profound arrhythmia. This patient received a permanent pacemaker. A dual chamber system was used to maintain atrioventricular synchrony. It is a demand pacemaker which will be inhibited during normal sinus rhythm but provide synchronized atrial and ventricular pacing if it detects a pause of greater that 1 second's duration. In order to prevent further bouts of paroxysmal AF she has been successfully treated with sotalol 40 mg b.d. Modern dual pacemakers have algorithms which may also help with the prevention of paroxysmal AF that has been generated by bradycardia pauses and atrial ectopics. These are known as DDRP pacemakers and their use is increasing although it is not yet determined whether their use will be an advantage over standard dual chamber pacemakers combined with drug therapy in this clinical setting.

around a loop of cardiac tissue. The mechanism of re-entry is shown in Figure 12.33. The loops of myocardial tissue may be found as a combination of alterations in anatomy (congenital or acquired), and abnormal function of parts of the heart. The best-known is that due to the presence of anomalous pathways connecting the atria with ventricular myocardium. This is the basis of the pre-excitation syndromes, so called because of the tendency of the aberrant pathways to excite

12

TABLE 12.10 Features and classification of tachycardias Clinical

ECG

Type Supraventricular Sinus Atrial Paroxysmal atrial tachycardia Atrial fibrillation Atrial flutter

Rate (bpm)

>100

120-250

>140

150

Response to carotid sinus pressure

Symptoms and signs

P waves

QRS complex

Normal Abnormal shape

Normal Regular and normal aberrant conduction produces broad QRS of RBBB or, less commonly, LBBB type. Broad QRS also if BBB already present Irregular Very regular QRS complexes

Transient minor slowing May terminate tachycardia

Palpitation Palpitation, dizziness, breathlessness, chest pain may occur Diuresis on cessation

Transient slowing Conduction may decrease from usual 2:1 to 4:1 or less

As above As above

Absent Flutter waves; sawtooth pattern in lead V1

Junctional AVNRT* AVRT**

180-240 180-240

None seen (buried in QRS) Inverted P waves after QRS; QP interval < PR

Normal and regular Normal and regular

None, or abrupt cessation None, or abrupt cessation

As above As above

Ventricular Ventricular tachycardia

150-250

Dissociated P waves diagnostic, but can be difficult to see None seen

Regular broad QRS > 120ms sometimes >140ms. QRS look similar in V-leads 'concordance' Loss of regular discernible QRS features

No response

May cause collapse; sometimes well tolerated

Inappropriate

Collapse with rapid loss of consciousness

Ventricular fibrillation Ventricular flutter

150-300

* AVNRT = AV nodal tachycardia **AVRT = AV re-entrant tachycardia

areas of adjacent ventricle and produce the delta wave on the ECG (Fig. 12.34); the clinical syndrome that results is the Wolff-Parkinson-White (WPW) syndrome (p. 516). Slow conduction in the anomalous pathways is the basis for the arrhythmia in WPW, whereas in ischaemia segments of diseased ventricular muscle with slow conduction or unidirectional block are the cause.

SUPRAVENTRICULAR TACHYCARDIAS Supraventricular tachycardias (SVT) arise in the atria or AV node. They are characteristically associated with narrow QRS complexes, but there are important exceptions: • When there is pre-existing bundle branch block - a situation impossible to diagnose with certainty unless there are ECGs available in normal sinus rhythm - before or after the arrhythmia; • When there is rate-dependent bundle branch block; • In the pre-excitation syndromes (although the delta waves often do not show during tachycardia, as the anterograde conduction may be normal), orthodromic tachycardia.

Clinical features Common symptoms of SVT are an unpleasant awareness of rapid heartbeats or palpitations in the chest. Associated cardiac disease may dominate the picture, causing the arrhythmias to lower cardiac output, producing complaints of breathlessness, dizziness, fatigue or chest pain. The tachycardia may precipitate acute LV failure. Patients who suffer paroxysmal SVT, particularly those with structurally normal hearts, may notice polyuria during, or after, the attack.

Diagnosis Differentiating between the many forms of SVT may occasionally be difficult. The ECG is the basic diagnostic tool, although aspects of the history and, occasionally, physical examination can give clues. Very fast heart rates may produce ST and T-wave changes on the ECG which persist for some hours after cessation of the arrhythmia.

Sinus tachycardia Sinus tachycardia is a persistent sinus rate over 100 bpm. It is regarded as pathological when inappropriate and

507

TABLE 12.11 Guide to use of antiarrhythmic drugs VaughanWilliams classification

Electrophysiological action On action potential

Increased threshold for activation Slow rate of rise of action potential? (Inhibit fast sodium channel)

QRS and QT prolongation, especially quinidine

Disopyramide

Comment

Toxicity now limits use. Effective in VT, and may convert AF or A flutter to sinus rhythm

Gl disturbance Proarrhythmogenic Interacts with digoxin and numerous other drugs

Now rarely used in UK practice

Not used long term, owing to side-effects. Acute i.v. for ventricular and supraventricular arrhythmia Ventricular and supraventricular arrhythmia Prophylaxis in paroxysmal AF Ventricular arrhythmia during and after myocardial infarction Phenytoin useful for digoxin toxicity arrhythmia

Lupus syndrome Other autoimmune disorders Negative inotropy Anticholinergic side-effects

IB Lidocaine • (lignocaine) Mexiletine Phenytoin .

Inhibit fast sodium channel Shorten action potential

Usually none

1C Flecainide 1 Propafenone

Inhibit fast sodium channel Inhibit His-Purkinje network with QRS widening

QRS duration prolongation and QT lengthening

Mainly ventricular arrhythmia

Bradycardia

SVT Exercise-induced VT Rate control of AF, as adjunct to digoxin For i.v. use, ultra short acting

Transient i LV function

Pneumonitis, proarrhythmia, interference with thyroid function, photosensitivity Profound hypotension p-blocker side-effects

Class II: p-Adrenoreceptor blockers e.g. Propanolol Slow diastolic depolarization Atenolol

Esmolol Class III: Amiodarone

CNS side-effects

Apart from lidocaine (lignocaine) in the acute situation little use in UK

Major proarrhythmic potential with flecainide

Have largely taken over class 1 treatment

Hypotension, wheeze Negative inotropy Claudication worsens

Prolongs action potential

QT prolongation

Prolongs action potential Additional p-blocking effects Prolongs AP

QT prolongation QT prolongation

Extremely wide-spectrum against SVT, VT and VF For life-threatening arrhythmia i.v. for resistant VF SVT and VT

QT prolongation

Acute termination AF

Proarrhythmia

Class IV: Calcium antagonists e.g. Verapamil Inhibit slow calcium channel

None

SVT, except acute AF

Hypotension Profound AV nodal block

Unclassified Digoxin

Inhibits Na/K ATPase

None

Control of AF rate

Adenosine

AV nodal block

None

SVT. Differentiation of SVT andVT

Digoxin toxicity Proarrhythmic Transient block; angor animi

Bretylium tosylate Sotalol Ibutilide

chronic, e.g. with diseases such as thyrotoxicosis, or when secondary to heart failure. Under these circumstances treatment consists of control of the underlying disorder. A number of diseases of sinus node function produce a chronic sinus tachycardia, such as the dysautonomia of diabetes, cardiac infiltration with amyloid, and extrapyramidal disease of the Shy-Drager syndrome.

Atria I tachycardia 508

Toxic effects

On ECG

Class I: Membrane-stabilizing drugs IA Quinidine

Procainamide

Indications

In atrial tachycardia (Fig. 12.35) the ECG shows abnormal

Not in general use in UK

P waves at rates from 120 to 250 bpm. The mechanism of the arrhythmia is usually enhanced, or abnormal, automaticity in the SA node or other site in the atria. Causes include coronary heart disease, cardiomyopathy, rheumatic heart disease and the sick sinus syndrome. The normal AV node may not conduct at rates much greater than 200 bpm, so that very fast atrial rates may be associated with intermittent failure of contraction. This may lead to an atrial tachycardia with 2:1 conduction or higher degrees of block. Conduction through the AV node may be further impaired by disease or drugs such as digoxin. Indeed, one

12

509

FIG. 12.33 Mechanism of re-entry tachycardias Q] Normal conduction. B An area of abnormal conduction affects one limb of the circuit. This interrupts the antegrade conduction (large grey arrow) but is able to conduct retrogradely some moments later (black dashed line). This combination allows re-entry of the impulse (black arrow) and a tachycardia ensues.

FIG. 12.35 ECG showing atrial tachycardia with 2:1 conduction Dots indicate position of P waves. (Atrial rate is 300/min.)

FIG. 12.34 ECG characteristics of the Wolff-Parkinson-White syndrome Note the short PR interval and pre-excitation of the ventricle, producing the slurred upstroke of the R wave (delta wave, arrowed in V4). O

of the manifestations of digoxin toxicity is atrial tachycardia (at a relatively slow rate) with block.

Management Withdrawal of digoxin is important if toxicity is suspected. Otherwise, measures to control ventricular rate include intravenous verapamil (slow bolus of 5mg repeated to a maximum of 10-20mg in 20min) when there has been no prior -blocker therapy, and intravenous digoxin when glycoside toxicity is not the cause. Patients who have received -blocker therapy should be given verapamil with great caution, for fear of precipitating severe bradycardia or high-grade AV block. Intravenous -blockade (e.g. atenolol 2.5 mg i.v. over 2-5 min, repeated 2-3 times over 30min; esmolol 500 g/kg for 1min then 50 g/kg/min for up to 10 min) may also terminate the tachycardia, as will DC cardioversion or atrial overdrive pacing.

Atrial fibrillation Atrial fibrillation (AF) is probably the most common arrhythmia, and in many patients may be permanent. It is characterized by a chaotic, rapid and low-amplitude wave-

Q Fig. 12.15

510

Fig. 12.16

FIG. 12.36 Rhythm strips showing atrial fibrillation and flutter H Atrial fibrillation. Note the irregular ventricular rate. B Atrial flutter with typical sawtooth P waves and a regular ventricular response (4:1). C Atrial flutter with a variable ventricular response. ©

form on the ECG (Fig. 12.36). The ventricular response is characteristically completely irregular in rhythm, and usually shows a narrow QRS complex unless there is preexisting disease or rate-dependent bundle branch block. Despite an atrial activity of 350-600 impulses/min, the AV node conducts only intermittently and randomly. The nodal rate is usually less than 200 impulses/min, or even less in the presence of AV nodal disease or drugs suppressing nodal activity (e.g. digoxin). An exception to the rule of irregular ventricular activity in AF is when it coexists with complete heart block, when a subnodal pacemaker provides a slow, regular ventricular complex (Fig. 12.31). The rhythm is believed to be caused by localized re-entry circuits within the atria, and is a frequent consequence of chronic atrial distension from any cause. The causes of AF are shown in Table 12.12. The most common underlying condition is probably hypertension. Some patients develop

TABLE 12.12 Causes of atrial fibrillation With structural heart disease • Rheumatic mitral valve disease • Hypertension • Cardiomyopathy Dilated Hypertrophic • Atrial septal defect • Coronary heart disease

Without structural heart disease Alcohol Thyrotoxicosis Acute pericarditis Constrictive pericarditis Pulmonary embolism Sick sinus syndrome Myocarditis Coronary heart disease

there is associated rheumatic mitral valve disease, especially mitral stenosis. Stroke due to embolism is nearly five times more common in patients with AF than in those in sinus rhythm, and this relative risk rises to 17-fold if there is associated mitral valve disease.

12

Clinical features Clinically, the symptoms are determined by the rate of the arrhythmia and the underlying cardiac state, but they do not differ significantly from those of any SVT. The pulse is of variable rhythm and volume. Very early beats, especially with fast AF, may not produce sufficient stroke volume to produce a pulse at the wrist, producing an apex to radial pulse deficit in 'uncontrolled' AF. The venous pulse shows only a single waveform, corresponding to ventricular systole (the v wave).

RECENT ADVANCES ATRIAL DEFIBRILLATOR

Patients with intractable and poorly tolerated paroxysmal atrial fibrillation, especially those with left ventricular dysfunction, are being considered for nonpharmacological interventional therapy. These may include AV nodal ablation followed by dual-chamber pacemaker insertion, or operative procedures (e.g. 'maze or corridor' operation) to restore sinus rhythm while preserving atrial transport. An alternative currently being explored is the insertion of an implantable atrial defibrillator, similar to the implantable ventricular cardioverter/defibrillator. This device detects atrial fibrillation with satisfactory sensitivity and specificity. With improvements in electrode design, and by keeping the shock duration short and energy low (<1.5 J), intraatrial shocks are reasonably well tolerated by patients. The risk of proarrhythmia (inducing ventricular tachyarrhythmias) in humans has yet to be evaluated and kept below 1% per year (as in acceptable antiarrhythmic drug therapy). The final hurdle is cost, as these devices are significantly more expensive than permanent pacemakers.

the rhythm for no discernible reason and are said to be suffering lone atrial fibrillation; they appear to have a normal prognosis. For other patients the immediate consequence of the arrhythmia and its long-term prognosis are determined by the underlying condition. AF may be paroxysmal or permanent. Even when structural heart disease is absent paroxysms of AF can sometimes be symptomatically disabling, and although drug therapy to control the ventricular rate during an attack is reasonably effective, there may be considerable difficulty in providing adequate prophylactic therapy. The loss of a mechanical atrial systole can be very important in patients with severe heart disease, and in some is sufficient to precipitate heart failure. The loss of atrial contraction predisposes to stasis of blood in the atria and atrial appendages. Thrombi form, which pose the threat of systemic embolization. This tendency is accentuated where

/Vlancfgfemenf

The aim of treatment is control of the ventricular rate, both at rest and during exercise. In individuals with paroxysmal AF, prevention of recurrences is attempted. Consideration should be given to converting all patients to sinus rhythm, unless there are grounds to believe that successful cardioversion is highly unlikely or reversion to AF almost inevitable. Patients falling into the latter category might be those with long-standing mitral valve disease and hugely dilated atria, or those discovered to have been in the rhythm for many years. The risk from cardioversion lies in the possibility of systemic embolization from the previously fibrillating left atrium. The risk of having intra-atrial clot due to AF rises rapidly with the time that the patient is exposed to the rhythm, so that cardioversion in anything other than emergency situations should be deferred in those who have been in AF for more than 48 hours without adequate anticoagulation. Acute AF Severely haemo dynamically compromised patient. Central line cannulation is needed for the infusion of amiodarone: 5mg/kg diluted in 250 mL of 5% dextrose and infused over 20-30 minutes. Significant electrolyte disturbances are excluded or treated and oxygenation maximized by treating pulmonary oedema, if present (p. 492). Arrangements are made to undertake DC cardioversion under a general anaesthetic. The initial slow bolus of amiodarone should be followed by an infusion of 1200 mg over the succeeding 24 hours, even if DC cardioversion is successful. Anticoagulation is not an issue in these patients. An alternative and older protocol is to use intravenous digitalization prior to DC cardioversion, since degeneration to ventricular fibrillation is unlikely in the absence of digoxin toxicity. Digoxin is given as 500 g in l00 mL 5% dextrose infused over 30 minutes. This may be repeated with a second dose of 500 g infused over 1 hour, or ouabain for much faster digitalization: Img in 100 mL 5% dextrose over 1 hour. In a compromised patient, negative inotropic agents such as verapamil, flecainide and -blockers should be avoided.

511

Haemodynamically stable patient with AF > 130bpm. The aim is to control ventricular rate to achieve a resting AF rate of 90bpm or less, and produce a 'chemical cardioversion' within 48 hours. If this conversion is not achieved the patient is discharged on maintenance drugs designed to control the ventricular rate, and anticoagulants prior to elective DC cardioversion. Oral digoxin is used to initiate ventricular rate control: digoxin 15 g/kg divided into three doses over first 24 hours. The maintenance dose of digoxin is established with reference to the level of this drug in the blood (see p. 518). In patients with good left ventricular function improved control of the heart rate may be achieved by using oral verapamil 40-80 mg 8-hourly. Caution should be exercised with combinations of digoxin and verapamil, because verapamil increases the plasma levels of digoxin. In the absence of structural heart disease or ischaemia, flecainide by i.v. bolus (slow injection of 2mg/kg up to 150 mg) is gaining popularity. Anticoagulation. All patients should be anticoagulated with heparin unless there are specific contraindications, such as very recent surgery. Heparin is given as a bolus of 5000U followed by 24000-48000U/24h adjusted according to the clotting times, aiming for a 2.5-3.0-fold increase over control clotting times. If cardioversion is not achieved the patient should be converted to oral anticoagulation with warfarin (aiming for an INR = 3.0) with a view to elective cardioversion after 3-6 weeks' adequate anticoagulation. Chemical cardioversion. This means using drugs to achieve what is usually undertaken with DC shock. It may be achieved with oral amiodarone or intravenous flecainide. The regimens are: a. Amiodarone 200 mg oral t.d.s. for 1 week, 200 mg b.d. for the second week, maintenance 100-200 mg daily. (NB: It is necessary to reduce the digoxin and warfarin doses if they are used concurrently.) Amiodarone is effective but may take several days to restore sinus rhythm. It may be combined with digitalization, but there is an interaction between these drugs, so digoxin doses should be reduced and plasma levels checked. b. Flecainide: 2mg/kg i.v. slow bolus over 10 minutes minimum. Maintenance with oral flecainide 50 mg 12hourly, increasing to a maximum 300mg/day. A recent report suggests that oral loading with 300 mg flecainide, followed by standard maintenance regimens, is as effective as i.v. treatment. In patients with good ventricular function (3-blockade with esmolol, the ultra-short-acting agent, has been successfully used (see p. 518 for treatment regimen). Maintaining sinus rhythm. If cardioversion is immediately successful, the patient with very recent-onset AF in

Fig.

512

12.16

an otherwise normal heart does not require anticoagulation. Digoxin is not continued, and if there has been an obvious precipitating event to the AF, such as excess alcohol, the appropriate advice is given and no preventative drugs are prescribed. If the AF, albeit of recent onset, was poorly tolerated, owing, for instance, to the coexistence of ventricular hypertrophy and poor LV compliance, treatment designed to prevent further attacks is advisable (see below under Paroxysmal AF). Patients who required anticoagulation or were electively cardioverted after some weeks in AF should continue to receive anticoagulants for up to 3 months, as well as prophylactic antiarrhythmic treatment continued for that time. If they remain in sinus rhythm and do not have severe cardiac abnormalities as the substrate for the rhythm, all antiarrhythmics may be discontinued at the 3-month review. Patients with severe cardiac abnormalities, or those in whom AF was poorly tolerated, should probably continue treatment in the long term. Chronic sustained AF Treatment aims to reduce the resting ventricular rate to about 90 bpm, with a modest increase on exertion. Digoxin alone, although effectively controlling the resting heart rate, is insufficient to control the response to exercise in a large proportion of patients. A combination of digoxin with verapamil (or digoxin with a -blocker) is superior to digoxin alone in these circumstances. Anticoagulation with warfarin reduces the risk of emboli and is essential when rheumatic valve disease is present, or if the heart (particularly the left atrium) is enlarged. It is indicated in patients with a previous history of systemic emboli, and where the cause of AF is thyrotoxicosis. Paroxysmal AF can be disabling. Commonly used prophylactic treatments are shown in Table 12.13. Digoxin is less effective at preventing attacks. The symptoms may be

SUMMARY 5 Digoxin points

Rapid loading

I.V.

Oral

500 g diluted over 30 min Second 500 g over 60 min

15|ig/kg in 3 divided doses in 24 h

Slow loading

125-250 g oral b.d. for 1 week

Maintenance

62.5-375 g o.d. (therapeutic range 1.0-2.6 mmol/L)

Reduce all dosages in: • elderly patients • renal impairment (by 30-70%, according to GFR) • combination with verapamil • combination with quinidine • combination with amiodarone

12

TABLE 12.13 Drug treatment of paroxysmal atria I fibrillation Amiodarone • Loading dose - 200 mg t.d.s. (to a total of 1.2-1.6 g) Maintenance dose -100-400 mg/day Flecainide • Upto150mgb.d. Propafenone • Up to 900 mg/day in divided doses Disopyramide • 250-800 mg/day in divided doses • Must be used with digoxin • Valuable in bradycardia-dependent AF

helped by verapamil, which lowers the peak heart rate during a paroxysm of AF. Ablation of the AV node and pacemaker implantation are used in intractable cases; this does, however, render the patient pacemaker dependent. Atrial flutter Atrial flutter is an arrhythmia caused by multiple re-entry circuits producing atrial contraction at 300 ± 50 bpm. Conduction block at the AV node (most commonly 2:1) is almost always present, producing a regular pulse of 150 bpm. This pulse rate should always lead to the clinical suspicion of atrial flutter. Vagal slowing of AV conduction produced by carotid sinus massage will temporarily increase the degree of AV block, and the sudden halving or more of the ventricular rate can easily be detected at the bedside. On the ECG monitor, P waves previously concealed within T waves become visible as a sawtooth pattern, thus establishing the diagnosis (Fig. 12.36C). O Physical examination may show flutter waves in the neck. The causes of atrial flutter are the same as those of AF (Table 12.12), but flutter is rarely permanent. Drugs are less effective in flutter compared with fibrillation, and DC cardioversion is the therapy of choice. Energies of 25-50 J are frequently sufficient to achieve successful cardioversion to sinus rhythm. Intravenous verapamil will successfully convert many patients to sinus rhythm, as will pacing the atria at or below the tachycardia rate (underdrive pacing), or rapidly for a few minutes (overdrive pacing). Chaotic atrial tachycardia and wandering atrial pacemaker Chaotic atrial tachycardia is characterized by multiple wave shapes and PR intervals. It is quite common in elderly patients, whereas in younger patients it is usually associated with severe respiratory disorders. It requires no specific intervention, except for control of any associated illness and withdrawal of digoxin if this has previously been prescribed. Conversely, a sustained rhythm may be slowed by digoxin or verapamil.

FIG. 12.37 AV nodal re-entry tachycardia Note rapid narrow QRS complex tachycardia (rate 185 bpm).

The wandering atrial pacemaker is a slower arrhythmia often associated with variations in vagal tone, allowing various subsidiary atrial pacemakers to become established. It requires no special treatment.

JUNGTIONAL TACHYCARDIAS AV nodal and AV re-entry tachycardia or paroxysmal supraventricular tachycardia The mechanism involved in these tachycardias is re-entry through a functionally distinct area of the AV node or tissues surrounding it. They are common, particularly in (young) patients with structurally normal hearts. There is a 1:1 relationship between atrial and ventricular activity. P waves generally cannot be identified; if they can be seen, they will be buried within the QRS or immediately follow it (Fig. 12.37). The onset is abrupt, usually precipitated by an atrial extrasystole, and an equally abrupt cessation is the rule, although the latter may not be appreciated by the patient as a sinus tachycardia may follow the arrhythmia. The usual rates are about 180-240 bpm, with narrow QRS complexes, except if there is pre-existing or rate-dependent bundle branch block. If manoeuvres or drugs induce a lower degree of conduction than 1:1, AVNRT is excluded and a paroxysmal atrial tachycardia or flutter is the more likely diagnosis. Management Treatment may not be necessary if the attacks are transient and infrequent. Several vagotonic manoeuvres may be helpful (Table 12.14), all of which can be taught to the patient who suffers paroxysmal attacks. Failing these simple approaches, drug therapy is usually successful. Intravenous adenosine is the drug of choice, although intravenous injections of verapamil, -blocker, disopyramide and amiodarone adenosine may also prove useful. Atrial and

513

TABLE 12,14 Simple procedures to terminate paroxysmal SVT

!

• Carotid sinus massage. If effective, the rhythm is abruptly stopped; occasionally only moderate slowing occurs • Cold water splash on face (to mimic the diving reflex). This may be effective, but is often difficult to administer • Performance of Valsalva's manoeuvre (often effective) • Swallowing cold drinks • Inducing vomiting (a few patients discover this themselves)

ventricular pacing may also be used, the former following a similar procedure to that used to convert atrial flutter. A more elegant approach is to interrupt the re-entry circuit with critically timed ventricular or atrial ectopic impulses using chronically implanted antitachycardia pacemakers. Long-term prevention can often be achieved with (3blockers, amiodarone, flecainide and propafenone. Transcatheter ablation of accessory pathways or the AV node can be undertaken. These latter techniques need a prior definitive intracardiac electrophysiological study (p. 501), but with the advent of radiofrequency catheter ablation are the treatment of choice in significantly symptomatic patients.

TABLE 12.15 Differentiation of VT from SVT with aberration or bundle branch block (BBB) Feature

Value in differentiation

Broad QRS complex (>120ms)

Very broad (>160ms) - almost always VT Broader than 140ms favours VT, but may be due to pre-existing BBB

Bizarre QRS morphology, similar in V1-V6 and not resembling any BBB pattern - concordance

Characteristic of VT when present

Independent P-wave activity

Diagnostic of VT, when found

Fusion beat with QRS between sinus and tachycardia morphology

Diagnostic of VT, when found

Variability of rate from 10 to 20ms; except for fusion/capture beats

Not good

Hypotension, shock and collapse

No value

When V-A conduction occurs with 2:1 block

When ventricular rate is shown to be greater than atrial, the rhythm is VT

Non-paroxysmal AV nodal or functional tachycardia Non-paroxysmal AV nodal tachycardia is a narrowcomplex tachycardia associated with inverted P waves in II, III and aVF. Characteristically, it is caused by ischaemia and digoxin toxicity, although junctional rhythms may be precipitated by most forms of heart disease; the mechanism is enhanced automaticity. Because the atria and ventricles are stimulated virtually simultaneously, the P waves may precede the QRS complexes with a short PR interval. Alternatively, they may be buried within the QRS complex or follow after it. The onset of the tachycardia is gradual and the established rates, between 60 and 130bpm, considerably slower than the paroxysmal forms of tachycardia. Therapy is along the same lines as for AV nodal and reentry tachycardias, although lignocaine may also be useful. Withdrawal of digoxin may be all that is required when toxicity is implicated.

BROAD-COMPLEX TACHYCARDIAS The hallmark of ventricular rhythm is the broad QRS complex. Ventricular fibrillation and flutter do not cause diagnostic difficulties, but ventricular tachycardia (VT) is often wrongly diagnosed as SVT with rate-related aberra-

FIG. 12.38 Ventricular tachycardia (monomorphic) A rapid, broad QRS complex tachycardia illustrating the similarity of appearance in complexes from V1 to V6 (concordance). O

tion, or with pre-existing bundle branch block. There is often a reluctance to diagnose VT, particularly on the part of less experienced physicians, probably because the implications for the patient are so much worse than for SVT. Guidelines for their differentiation are given in Table 12.15. In the context of acute infarction, or when there is severe underlying myocardial disease, the observation of a broad-complex tachycardia should always be assumed to be ventricular in origin until proved otherwise.

Ventricular tachycardia Q Fig. 12.17

514

VT (Fig. 12.38) is defined as the appearance of three or more ventricular ectopic beats in a sequence. When sustained, the rate is between 150 and 250 bpm and regular.

Unsustained VT can be seen in people with normal hearts, but organic heart disease is suspected if the rhythm is sustained or produces symptoms. Causes are coronary heart disease, cardiomyopathy or myocarditis. VT carries the risk of sudden death, usually due to degeneration of the rhythm to ventricular fibrillation (VF). Even with the guidelines to diagnosis given in Table 12.15, VT can in some circumstances be difficult to diagnose, and prolonged examination of the ECG or the use of an oesophageal ECG lead may be necessary. In about half of patients the sinus node continues to fire independently, so that atrial P waves dissociated from ventricular activity can sometimes be observed, if not buried within the QRS complex. O In a few patients, retrograde activation of the atria occurs; this may be recognized by finding inverted P waves. In paroxysmal VT (150-250bpm) the onset is abrupt, usually following a critically timed ectopic beat. A slower form of VT, accelerated idioventricular rhythm (rate 60130bpm), occurs as an enhanced automatic rhythm during periods of suppressed sinoatrial activity. It is usually seen in acute infarction and its prognosis - unlike that of paroxysmal VT - is generally good. Management Acute In the presence of acute haemodynamic disturbance or acute myocardial infarction, immediate DC cardioversion with a synchronized shock is the treatment of choice; this is followed by an intravenous lidocaine (lignocaine) bolus dose of 100 mg (or 1 mg/kg), followed by an infusion, starting at 4 mg/min for 30 min, reducing to 2 mg/min for 2 hours and continued at 1 mg/min thereafter, for approximately 24 hours. This approach is generally successful, particularly in acute infarction, when there is rarely any indication to continue to chronic oral antiarrhythmic therapy. In resistant or recurrent cases an alternative therapy is amiodarone. Mexiletine, flecainide or bretylium may be required in certain instances, although the acute use of many different antiarrhythmic drugs should be avoided. The serum potassium must be maintained at or above 4.5mmol, and acidosis and hypoxaemia reversed. Chronic Episodic or paroxysmal VT complicating organic heart disease (e.g. hypertrophic cardiomyopathy, chronic coronary heart disease and cardiomyopathy) requires prophylactic therapy. In choosing the drug, information on its

SUMMARY 6 Treatment of tachycardia where doubt persists as to SVT or VT

EQ

• DC cardioversion if haemodynamics are compromised. • Do not use verapamil. • Try i.v. adenosine (0.05-0.25 mg/kg) - it will terminate an SVT but will not affect VT or haemodynamics. • Try i.v. lidocaine (lignocaine) (1 mg/kg bolus) - it may terminate VT.

efficacy in the individual patient should be taken into account. The response to exercise testing and monitoring the effectiveness of therapy with Holter 24-hour ECG recordings is helpful. In those patients in whom VT is inducible by exercise, -blocking drugs may be useful; sotalol, with its class III effects (Table 12.11), may be the rational first choice. Patients with poor ventricular function are probably most at risk from VT. The choice of drug for these patients is very limited, owing to the negative inotropy of most antiarrhythmic agents. Amiodarone or mexiletine are the most useful in these patients. A new form of therapy for those patients in whom drug therapy fails is the use of small implantable automatic defibrillators (AICD) which detect the arrhythmia (VT or VF) and, after an appropriate pause, deliver a DC shock. Surgery and catheter ablation Certain patients have an anatomical basis for their ventricular arrhythmia - either a ventricular aneurysm or a localized area of ischaemically injured myocardium. Treatment options now include excision of the aneurysm or electrically abnormal tissue by surgery, or ablation using catheters capable of delivering radiofrequency ultrasound.

Torsade de pointes ventricular tachycardia Torsade de pointes VT is an uncommon but important arrhythmia, characterized by a VT, often of modest rate but associated with an undulating QRS height owing to a slow but continual variation in the QRS axis (Fig. 12.39). Its importance is owing to its precipitation by antiarrhythmic drugs, particularly if electrolyte disturbances coexist.

Management Acquired forms are treated by withdrawal of the offending agent and correction of electrolyte abnormalities. In

FIG. 12.39 Rhythm strip illustrating torsade de pointes, a broad-complex ventricular tachycardia Note the fluctuation in size of the complexes.

515

OTHER ARRHYTHMIC SYNDROMES Sick sinus syndrome

FIG. 12.40 Rhythm strip showing ventricular fibrillation

less stable situations an infusion of isoprenaline, or increasing the heart rate with a temporary pacemaker, is the appropriate remedy. Congenital syndromes can be caused by abnormalities in sympathetic supply to the heart. These are best treated with (3-blocking agents.

Ventricular fibrillation and ventricular flutter VF is characterized by rapid, irregular and uncoordinated electrical activity in the ventricles, probably due to re-entry circuits within localized areas of myocardium. The ECG shows more or less coarse, irregular waveforms, without discernible P, QRS or T waves (Fig. 12.40). In ventricular flutter the entire tracing appears as a rough sawtooth with a rate of 160-250 bpm, without normal QRS morphology. If either rhythm occurs effective contraction and cardiac output cease, leading to unconsciousness within seconds. These rhythms do not usually revert spontaneously and are often precipitated by an ectopic beat or runs of VT, particularly when they complicate acute myocardial infarction. VF is probably the arrhythmia responsible for most cases of sudden death in the community (p. 561). In the context of acute myocardial infarction, when VF occurs early, within 24-48 hours of onset of the illness (primary VF), and is successfully treated, the prognosis for the patient is not adversely affected. However, VF occurring late in the course of acute infarction (secondary VF) often follows extensive muscle damage and carries a grave prognosis. Management Witnessed collapse due to VF (usually seen only in coronary care units) requires immediate action; a precordial chest thump (p. 519) may convert up to 15% of cases without the need for electrical defibrillation. In all other circumstances basic cardiopulmonary resuscitation should begin until the treatment of choice is available, namely defibrillation by the application of an external DC shock (p. 519). Correction of acidosis, hypoxaemia and electrolyte imbalance improves the likelihood of conversion to a stable rhythm. Drug treatment and long-term prophylactic therapy are the same as for VT. Primary VF does not generally require subsequent prophylactic therapy. O O MCQ12.8

516

O

MCQ12.9

The sick sinus syndrome (SSS) is a relatively common condition, characterized by episodic arrhythmias combining episodes of profound bradycardia and heart block with episodes of tachycardia (it is sometimes called the tachycardia-bradycardia syndrome). The bradycardias may be paroxysmal or sustained sinus bradycardia, episodes of sinus arrest, or SA block producing long pauses in pulse rate and pauses after atrial ectopics. Immediately following a tachycardia there may be a prolonged pause or very slow sinus rhythm sufficient to cause dizziness or a Stokes-Adams attack. The tachycardias are mostly supraventricular, with AF, flutter, chaotic atrial rhythm and atrial tachycardia being the most common, although VT (especially torsades de pointes) is occasionally seen, as is chronic AF with a slow ventricular rate. The pathology is probably degeneration and fibrosis in the sinus node and conducting tissues. The AV node is often involved in the degenerative process. Ischaemic heart disease and amyloid infiltration of the heart may be the cause in some patients, and SSS has also been associated with many other pathologies, from cardiomyopathy to Friedreich's ataxia. Drugs such as digoxin, (3adrenoreceptor blockers or quinidine may mimic SSS. Apart from discomfort and Stokes-Adams episodes, the condition carries a risk of systemic embolization and is the cause of episodic heart failure. Sick sinus syndrome is often unmasked when drugs are given to treat tachycardia, and profound bradycardia or heart block is produced, necessitating the implantation of a pacemaker. Management and prognosis Treatment is with a combination of antiarrhythmic drugs and pacemaker implantation. The prognosis is generally good, but mainly dependent on any underlying condition.

Wolff-Parkinson-White syndrome The ECG abnormality of short PR interval (<0.12s) with abnormal QRS widening due to delta wave (Fig. 12.34) is relatively common. The clinical syndrome, WolffParkinson-White (WPW) syndrome, with paroxysmal tachycardias is much less common, although the precise incidence of arrhythmia in individuals with the ECG abnormality is not clearly established. The syndrome generally occurs in an otherwise normal heart, although it may complicate Ebstein's anomaly (p. 538) or mitral valve prolapse (p. 525). The abnormality is due to an accessory pathway between the atrium and ventricle allowing a reentry or reciprocating tachycardia to develop (p. 506). The tachycardia produces narrow QRS complexes without the delta wave, because conduction is anterograde through the AV node. Thus, the diagnosis of WPW may not be made until an ECG in sinus rhythm is obtained. There is an

increased incidence of AF. In a small proportion of patients who have a short refractory period in the accessory pathway AF may conduct very rapidly (1:1) to the ventricles; this may degenerate to VF and is the usual cause of sudden death in these patients. Investigation and management Electrophysiological studies establish the site, number and refractory period of the accessory pathway(s). Treatment may include antiarrhythmic drugs, but often requires ablation of localized accessory bundles during electrophysiological (EP) testing. Digoxin and verapamil should be avoided, as they can increase accessory pathway conduction. Lown-Ganong-Levine syndrome

Patients with Lown-Ganong-Levine syndrome (LGL) have short bypass tracts that connect the atria to various portions of the proximal bundle of His or distal AV node. The ECG manifestation is of a short PR interval (<0.12s) without pre-excitation (no delta waves). Such patients are subject to paroxysmal tachycardias, but are less commonly threatened by rapidly conducted AF. Treatment is the same as for WPW syndrome.

Long QT syndrome Several congenital disorders are described which are characterized by a long (>0.45 s) QT interval. They are caused by increased action potential duration and predispose to ventricular tachycardia, and are associated with sudden death, exceeding 60% at 15 years in untreated symptomatic patients. The Romano-Ward syndrome is autosomal dominant and more common than the autosomal recessive disorder Jowell-Lang-Nielsen syndrome. The nuclear basis for these disorders has been characterized and mutations on the genes coding four cardiac action potential ion channels have been identified: LQT1, LQT2,LQT3 and SCN5A. The clinical course of long QT syndromes is influenced by the associated mutations, so that one group (LQT1 patients) is susceptible to physical and emotional stresses as triggers for VT, and LQT3 patients are more at risk during sleep. Treatment has reduced mortality to 3-4% within 10 years of the first symptom. The mainstay of treatment consists of -blockers, cardiac sympathetic denervation, and the use of pacemakers whenever there is bradycardia or pause-induced syncope. Gene-specific therapies remain an exciting prospect for the future.

SYNOPSIS OF DRUG TREATMENT OF ARRHYTHMIAS A decision tree for the diagnosis of tachycardias is shown in Figure 12.32. Summaries of the treatment of, and drug doses used in, arrhythmias are given in Tables 12.11,12.16 and 12.17. O

12

TABLE 12.16 Emergency drug treatment of common arrhythmias Arrhythmia

First-line therapy

Second-line therapy

Atrial fibrillation

i.v. flecainide bolus plus infusion. DC cardioversion if BP low DC cardioversion (50-1 00 J)

3-Blockers Verapamil DC cardioversion Verapamil Vagotonic manoeuvres Atrial pacing DC cardioversion Digoxin (if not digoxin toxicity) 3-Blocker Disopyramide Flecainide Amiodarone DC cardioversion DC cardioversion

Atrial flutter Atrial tachycardia and junctional tachycardia SVT and AVNRT or AVRT type

Vagotonic manoeuvres Adenosine Verapamil Vagotonic manoeuvres Adenosine Verapamil

SVT in WPW with

Adenosine

narrow QRS SVT in WPW with broad (delta wave) QRS VT and VF

Bradycardia Complete heart block

Amiodarone Flecainide Adenosine DC cardioversion Lidocaine (lignocaine)

DC cardioversion Bretylium Amiodarone Flecainide Mexiletine

Atropine Isoprenaline Pacemaker Isoprenaline

RESUSCITATION

Within 20 seconds of cardiac inactivity the victim falls unconscious, breathing soon ceases and the pupils dilate. Sudden death affects about 50000 people a year in the UK. Half of these are not previously known to have coronary artery disease. About 60% of deaths due to myocardial infarction occur within an hour of the first symptom (see Fig. 12.62), the majority occurring outside hospitals. The commonest rhythm disturbance (over 60% of cases) is VF. Cardiopulmonary resuscitation (CPR) may involve: • Basic life support, requiring training but no special equipment; • Advanced life support, which requires equipment and specialized skills. Many cities in the UK and other industrialized countries are involved in training the public in basic life support skills, and provide specialized ambulances with very rapid response times and facilities for advanced life support and DC cardioversion. Examples are Belfast and Brighton in the UK, and Seattle in the USA. With such basic life support available about 20% of victims may reach hospi-

517

TABLE 12.17 Doses and methods of administration of drugs commonly used in acute arrhythmia Drug

First dose

Method

Maintenance

Side-effects

Amiodarone

2-5mg/kg

i.v. slow bolus

Usually required 600-1 200 mg/day i.v. infusion

iBP Need central venous access

Atropine

0.6-1. 2 mg

i.v. bolus

Not used

Anticholinergic side-effects

p-Blockers Atenolol Metoprolol

5-10 mg 5-1 5 mg

i.v. bolus i.v. bolus

50-1 00 mg oral/day 1 00-400 mg oral/day

4 Heart rate;

Very short acting

Esmolol

500ng/kg

i.v. bolus 1 min

50u.g/kg/minfor4min

Bretylium

5-10mg/kg

i.v. rapid in emergency

Optional 1-2mg/min (i.v. infusion)

Digoxin

0.5 mg

i.v. slow bolus

Optional 0.75-1 .Omg for first 24 h, then 0.25 mg/day (oral)

Flecainide

1-2mg/kg

i.v. bolus over 10-30 min

Optional 0.15-0.25mg/kg/h (i.v. infusion)

Isoprenaline

50 g

i.v. rapid in emergency

Optional 0.02-0.1 8 g/kg/min

Lidocaine (lignocaine)

100mg

i.v. bolus

Essential 4mg/min reducing to 2mg/min for 24 h (i.v. infusion)

Verapamil

5mg, repeat with 10mg after 10 min

i.v. slow bolus

Optional 0.005 mg/kg/min for 1h (i.v.) or 80-1 20 mg/8-hourly (oral)

Adenosine

0.05-0.25 mg/kg (3-1 8 mg)

i.v. bolus (fast)

Not usual

tal and 11% live to be discharged. These numbers are halved when basic life support is not available. Survival is worse in older patients, and when arrest occurs at home or in the street. In hospital, when cardiac arrest due to VF occurs within 4 hours of symptoms and is successfully treated, 80% of patients are alive 3 years later.

BP; bronchospasm

BP; nausea Toxicity: lower dose in renal impairment and in elderly BP; cardiac failure Heart rate CNS toxicity

BP;

Transient

heart rate; AV block

BP; chest pain; angor animi

TABLE 12.18 Basic cardiac arrest procedures A Assessment Patient unconscious or asleep? Call for help if unconscious. Is airway clear? Clear it, pull chin up. B Breathing

BASIC LIFE SUPPORT

518

The basic cardiac arrest procedures are given in Table 12.18. Initial assessment is rapidly followed by establishing a clear, open airway. This is done by head tilt with chin lift and jaw thrust. In the absence of airway obstruction, rescue breathing begins with the mouth-to-mouth technique. Recommendations vary, but in the UK, four quick breaths are advocated. In the USA, two slow breaths (1-1.5s each) have been suggested, in order to diminish the likelihood of gastric distension with air, and subsequent vomiting (risking inhalation of vomit into the unprotected airway). Protective masks and airway equipment should be provided for personnel most often exposed (ambulance and paramedical staff), even though the risk of infection (such as from hepatitis B or HIV) is likely to be very small. Bag and mask techniques are taught to nursing and

If patient is breathing: turn to recovery position. If patient is not breathing, but is choking: Heimlich manoeuvre. If patient is not breathing, but is not choking: start mouth-to-mouth resuscitation. Call for assistance. C Circulation If no pulse: start external cardiac massage call for assistance continue ventilation

medical staff, but are frequently more difficult to use and less effective in inexperienced hands than is mouth-tomouth ventilation. Closed cardiac compression follows at a rate of 60-80 compressions/min. Training in the technique is essential. Midsternal compression is carried out with the heel of one hand to depress the sternum by 2-4 cm. The patient should

be on a rigid, flat surface with the resuscitator kneeling alongside. The arms are kept straight and compression is produced primarily by body-weight movement. Closed chest compression achieves a cardiac output by virtue of cyclical changes in intrathoracic pressure, rather than by squeezing the heart. A single-handed resuscitator should provide 15 chest compressions for every 2 mouth-to-mouth breaths. If two people are available, the ratio should be 5:1. CPR must be continuous, and pauses to reassess the situation or allow for advanced life-support techniques should not be greater than 10 seconds. The first response to a witnessed arrest should be a forceful precordial thump. This may be sufficient to convert VF or VT to a more stable rhythm by localized myocardial depolarization, or even to initiate a rhythm in cases of acute asystole.

ADVANCED LIFE SUPPORT Advanced life support (summarized in Fig. 12.41) consists of defibrillation with DC shock, endotracheal intubation, and the use of drugs and pacemakers.

DC defibrillation In VF, the earlier DC shock is given, the more successful the outcome. When defibrillators are immediately avail-

able, such as in coronary care units, the first procedure following the precordial thump should be defibrillation at 200 J. This will be sufficient to revert 85% of cases to a more stable rhythm. In 10% of patients a further 200 J shock is required and is successful. If these approaches are unsuccessful, life support must be begun immediately. Outside coronary care units, basic life support must be begun and interrupted only when the defibrillator is ready to produce the first shock of 200 J. If no monitor is available, major pulses are checked and, if absent after 3s, 15 chest compressions are given with the appropriate mouth-to-mouth breaths. A second shock follows the same pattern of events and, if VF persists, intravenous lidocaine (lignocaine) (100 mg i.v. bolus) is given before the third shock at 400 J. Adrenaline may be given before the fourth shock at the highest energy and, if an acidosis is demonstrated by arterial gases, bicarbonate (Immol/kg) is given before the fifth high-energy shock (Fig. 12.41). The technique of DC shock administration can be taught to nursing and paramedical personnel. Early defibrillation administered out of hospital by ambulance crews has been shown to improve survival from VF. Defibrillation paddles are placed on the chest with firm pressure, one under the right clavicle and one just lateral to the usual cardiac apex. Skin impedance is reduced with conducting gel or pads, and the administrator of the shock ensures that no-one is touching the patient before the shock is delivered. Complications are remarkably few, although very high-energy shocks or many repeated shocks carry the risk of some myocardial damage.

12

Endotracheal intubation If there is no recovery after a few minutes of basic life support, endotracheal intubation is necessary. This is a skilled technique which can be mastered by medical and paramedical staff alike, but which requires frequent practice. After the cardiac arrest it may be necessary to continue ventilation, via a mechanical ventilator, of seriously ill patients, those with a damaged ribcage from cardiac massage and those who sustain cerebral damage.

Drugs in advanced life support The number of drugs of proven benefit is small, despite the large number available to treat ventricular arrhythmia.

FIG. 12.41 Flow diagram of resuscitation procedures

(Based on recommendations of the UK Resuscitation Council.)

• Adrenaline (epinephrine) is the inotropic and chronotropic agent of choice. It is given intravenously in a dose of Img (ImL of 1/1000) or 2mg via the endotracheal tube. • Lidocaine (lignocaine) is given after VF and during attempts to treat VF. A bolus of Img/kg (or 100 mg) is followed by an infusion (4mg/min, reducing to 2mg/min over 24 hours). A second half-dose bolus can be given 10-20 minutes after the first. • Sodium bicarbonate. Bicarbonate infusion is only used in established acidosis, which is a relatively late event.

519

An average dose is Immol/kg of 4.2% sodium bicarbonate solution. • Calcium chloride is useful in specific circumstances, such as resuscitation after cardiopulmonary bypass and in calcium antagonist overdose. The usual dose is 10 mL of 10% calcium chloride. • Other drugs. The treatment of resistant VF or ventricular tachyarrhythmias may require the use of one or other of the alternative drugs available (p. 517). None, however, has a proven role as 'standard' therapy in CPR, and their indications have to be determined individually at the time of the resuscitation attempt.

RESUSCITATION FROM CHOKING Collapse due to acute airway obstruction is often caused by a lump of food. In the UK it is still recommended that active coughing with gravity drainage be encouraged while the subject is still conscious. This is supplemented by back blows and, finally, the abdominal thrusts known as the Heimlich manoeuvre. The latter depends upon gripping the subject from behind and, with clenched fist, producing a rapid, powerful upward thrust in the upper abdomen. This produces a rapid airflow, which may be sufficient to dislodge the obstruction.

Special techniques Cardiac pacing Cardiac pacing can be used to treat bradycardia, heart block and, sometimes, asystole. Asystole, unless it has been caused by drugs, rarely responds to cardiac pacing. Intra-aortic balloon counterpulsation Highly specialized techniques such as intra-aortic balloon counterpulsation have a limited role in specialized centres, where they can be used to support the circulation until a definitive procedure can be performed. Intracardiac injection Intracardiac injection has been used to administer calcium or adrenaline during cardiac arrest with asystole. Central venous cannulation delivers drugs to the heart with much less hazard.

When to stop Rigid rules cannot be given, but guidelines are possible. Resuscitation is almost never successful in patients who have a cardiac arrest complicating another serious illness, such as pneumonia or malignancy. It is important, therefore, that the staff likely to be first to witness a collapse should have had clear guidance concerning the desirability of resuscitation in individual patients. In general, the outlook for a useful neurological recovery is poor if the arrest occurred out of hospital and was not witnessed, or if more than 4 minutes passed without even basic life support being attempted. If no spontaneous respiration or cardiac output is achieved after 30 minutes of adequate advanced life support, then further attempts at resuscitation are questionable. The state of the pupils, whether widely dilated and reacting or not, is unreliable with respect to outcome. Age should not be used as the sole determinant of policy on resuscitation. O

O MCQ 12.10 520

VALVULAR DISEASE OF THE HEART AETIOLOGY AND NATURAL HISTORY In western countries, rheumatic heart disease was, until recently, the commonest cause of valve disease, and in most developing countries this is still the case. Various forms of degenerative valvular disease are increasingly being seen, some based on congenital abnormalities of the valve, others apparently arising de novo in middle age or later. For both rheumatic and congenital heart disease, it is often difficult to predict the progression of what initially may seem a trivial lesion of no consequence, to one that may be life-threatening 30 or more years later. What dictates progression of the valve lesion? Mechanical forces must be involved; pressure and shear forces on the valve probably make it thicken and even calcify. The tolerance of the myocardium and the advent of myocardial disease add to the difficulty in prognosis.

THE EFFECTS OF STENOSIS AND REGURGITATION Heart valve disease can occur either because of a narrowing of the valve (stenosis), causing obstruction to blood flow through the valve, or when the valve leaks, when the lesion is termed regurgitant, incompetent or insufficient. A combined or mixed lesion is common.

Valve stenosis Mild degrees of stenosis may produce an audible murmur and also a risk of endocarditis, but place an insignificant physiological burden on the heart. For significant symptoms to result, stenosis has to progress until the valve is narrowed to less than 30% of its normal area. Flow through the valve is maintained at close to normal levels by a compensatory increase in pressure in the chamber of the heart upstream of the stenotic valve. This chamber therefore hypertrophies.

TABLE 12.19 Relationship between valve area, blood velocity and pressure gradient Area (cm2)

Condition of valve Stenosis alone Normal Mild stenosis Severe stenosis Stenosis and regurgitation Normal area, no regurgitation Normal area, 50% regurgitation Mild stenosis, 50% regurgitation Mild stenosis, 66% regurgitation

Velocity (m/s)

Pressure gradient (mmHg)

3

1

4

1.5 0.6

2

16 100

5

3

1

4

3

2

1.5

4

16 64

1.5

6

144

2

The Bernoulli equation (P = 4V ), relating blood flow velocity (V) to pressure gradient (P) across a valve, shows the effects of regurgitation on the pressure gradient if the forward stroke volume is maintained. The figures are chosen for simplicity. With 50% regurgitation, the forward stroke volume and (theoretically) the flow velocity doubles; with 66% regurgitation, they increase threefold. In practice, the flow time also increases.

The severity of the stenosis is quantified by the pressure drop across the valve. In a normal valve this drop is not measurable by conventional techniques. However, in severe aortic stenosis the left ventricular pressure may rise to as much as 200 mmHg more than the aortic pressure, and the valve area may be less than 25% of normal. In mitral stenosis the left atrial pressure cannot rise above 40 mmHg because the pulmonary capillaries are unable to withstand a higher pressure. Thus, the mitral valve gradient is usually much less than the aortic valve gradient in significant stenosis, and the residual valve area accordingly greater. It follows that there are three ways of assessing the severity of a valve stenosis: • The pressure gradient across the valve. This indicates the additional work required to push blood through the valve. However, this is flow dependent (no flow, no gradient); if there is heart failure, with low flow, the valve gradient is smaller. • The valve area. Formerly, this could only be calculated from cardiac catheterization data, but direct measurement is now possible from ultrasound. There is, however, controversy over the accuracy of the measurements, because of the beam spread and lateral resolution of the ultrasound. • The peak blood velocity across the valve. Measured by Doppler ultrasound, this bears a very simple relationship to the pressure gradient (Table 12.19).

Valvular regurgitation Normal human valves are highly efficient. The aortic valve closes as the flow decelerates, and is fully closed with about 97% efficiency. However, severe mitral or aortic valve

regurgitation can be associated with up to 80% regurgitation (efficiency 20%). The forward flow then has to be five times greater than normal to allow for this leakage, and the chamber supplying the blood dilates accordingly. In aortic regurgitation the left ventricle enlarges; in mitral regurgitation the left atrium and left ventricle both enlarge greatly. Acute regurgitation (such as mitral valve chordal rupture, or aortic regurgitation from acute endocarditis) is badly tolerated, as the left ventricle and left atrium have not had time to increase in volume. The rapid increase in left ventricular diastolic pressure or left atrial pressure has severe consequences for left ventricular perfusion and pulmonary capillary pressure.

12

Combined valve lesions The combined effect of even mild stenosis with regurgitation is potentially catastrophic. Because the forward flow with as little as 50% regurgitation is twice normal, even a 50% stenosis becomes highly significant. This is shown in Table 12.19, where it can be seen that, with a combined lesion, the regurgitation greatly potentiates the effect of the stenosis. This is particularly true of the mitral valve, where pressures need to rise much less before producing symptoms.

MITRAL STENOSIS Aetiology Although a rare congenital form exists, mitral stenosis is almost always a consequence of rheumatic heart disease, and is the commonest valvular manifestation of the disease. The two valve cusps become adherent along their commissures, producing progressive stenosis with a 'fish mouth' orifice. Initially the cusps remain fairly pliable, but become increasingly rigid with time and eventually calcify. The increasing degree of stenosis produces a rise in left atrial pressure and left atrial dilatation. The raised pulmonary venous pressure can induce pulmonary oedema and lead to progressive pulmonary hypertension, with secondary incompetence of the tricuspid valve. Clinical features Mitral stenosis, which is commoner in women, may remain asymptomatic for many years with little limitation of exercise tolerance. However, an additional load on the heart, such as pregnancy, exertion, emotional stress or intercurrent chest infection (which is more common in patients with mitral valve disease because of the pulmonary vascular congestion), may produce the first symptoms of left heart failure. The onset of fast AF in middle age is a common presentation, often leading to left heart failure, and thrombus formation with the risk of embolism. Right heart failure eventually develops as a consequence of pulmonary hypertension.

521

The typical facial appearance of a patient with mitral valve disease - ruddy complexion, with red or vaguely cyanosed cheeks (mitral fades] - is now more commonly encountered in severe non-rheumatic causes of heart failure. Symptoms Of left heart failure The symptoms of left heart failure (p. 488) are shortness of breath on exertion, orthopnoea, paroxysmal nocturnal dyspnoea, pulmonary oedema, and haemoptysis of fresh blood occurring from rupture of a congested bronchial or pulmonary vein (uncommon with other causes of left heart failure). Haemoptysis may also be due to pulmonary infarction, which is more common in mitral valve disease. Due to atrial fibrillation The raised left atrial pressure may lead to AF, sometimes paroxysmal at first, but then sustained. The loss of atrial contraction and the shortened diastolic filling period can change an asymptomatic patient into one severely disabled by breathlessness, and may convert a quiescent stenosis into a life-threatening condition. Systemic embolism. Thrombus can form in the enlarged left atrium and give rise to systemic emboli. A hemiplegia may occur and embolism to other sites is frequent. The patient in sinus rhythm is also at risk from embolism, but to a lesser extent. The greatest risk is in the patient who, having been in AF, reverts to sinus rhythm, as the return of coordinated atrial contraction can dislodge clot. Due to a raised left atrial pressure Long-standing raised left atrial pressure often leads to constriction of pulmonary arterioles and thickening of the capillary basement membranes. There is increased resistance to flow through the lungs, and hence pulmonary hypertension. Right ventricular hypertrophy compensates for the increased load on the right ventricle. When pulmonary hypertension is severe, right ventricular failure follows, usually (but not always) associated with left heart failure. Right ventricular dilatation may lead to tricuspid regurgitation, with its attendant symptoms and signs (p. 531). Less frequently, tricuspid valve disease (regurgitation more frequently than stenosis) occurs through direct involvement in rheumatic fever. Massive enlargement of the left atrium can lead to hoarseness from pressure on the recurrent laryngeal nerve, dysphagia from oesophageal compression, and bronchiectasis from distortion of the left main bronchus.

522

Signs The main signs of mitral stenosis are listed in Summary box 7. Pure mitral stenosis in sinus rhythm without pulmonary hypertension is relatively uncommon. The signs are then purely auscultatory, as cardiac output is maintained and there is no significant right ventricular hypertrophy. The auscultatory signs are:

• An opening snap (OS, Fig. 12.42B) best heard halfway between the apex and the left sternal edge. It occurs close to the aortic second sound in severe stenosis, but is more separated from it in mild stenosis. • A rumbling diastolic murmur. This commences with the opening snap and fades rapidly in mild stenosis, but is full length and runs into the presystolic murmur in severe stenosis. • The presystolic murmur coincides with atrial systolic flow through the valve. It is of the same character as the mid-diastolic murmur and is sometimes heard when the latter cannot be heard. It is absent in AF. • A loud first heart sound. In the mobile stenotic valve the cusps are still fully open as systole starts, and so they close with a louder sound than normal. This is an unreliable sign. Later signs are AF and pulmonary hypertension. These patients also have signs of a low cardiac output, mitral fades, small stroke volume with small-volume pulse, a raised JVP, right ventricular hypertrophy, and a middiastolic murmur following the opening snap. When the cardiac output is very low the mitral murmur may be virtually inaudible, especially if the patient is in heart failure - so-called silent mitral stenosis.

Investigation In sinus rhythm the ECG may show P mitrale (Fig. 12.12, p. 477). Severe pulmonary hypertension may produce right ventricular hypertrophy or 'strain', but this aspect of the ECG is often unhelpful.

SUMMARY 7 Symptoms and signs of mitral stenosis Symptoms Of left heart failure

Shortness of breath, orthopnoea, paroxysmal nocturnal dyspnoea, haemoptysis

Of atrial fibrillation

Palpitations and syncope (at onset) May precipitate left heart failure Systemic embolism to CNS, limbs, gut and kidneys

Of pulmonary hypertension

Weight loss and cachexia Right heart failure

Signs Uncomplicated (auscultatory only)

Opening snap Mid-diastolic murmur Presystolic murmur

With atrial fibrillation

Irregular pulse

With pulmonary hypertension

Parasternal lift (RVH) Loud P2 Tapping apex (loud S1)

If very severe

Mitral facies Cardiac cachexia

12

FIG. 12.42 Features of mitral stenosis A Chest X-ray showing cardiomegaly with gross left atrial enlargement. B Diagrammatic illustration of the pressures in left atrium (blue) and left ventricle (black), showing a diastolic pressure gradient (shaded light blue). The timing of an opening snap (OS) and diastolic murmur is shown. S1 and S2 = first and second heart sounds.C Typical M mode. The mitral valve (arrowed) movement is abnormal (compare with Fig. 12.18). D 2D echocardiographic appearance in mitral stenosis. Thickened mitral valve (MV), dilated left atrium (LA) and the LV cavity (LV) are shown.

Chest X-ray. The chest X-ray may show enlargement of the left atrium (Fig. 12.42A), redistribution of flow to the upper zones, left heart failure and enlargement of the right side of the heart if there is coincident pulmonary hypertension. A penetrated lateral X-ray will show a calcified valve in most patients over the age of 50. Echocardiography. The echocardiogram shows thickening of the cusps (Fig. 12.42D), which are fused together so that the posterior cusp moves anteriorly in diastole with the anterior cusp. The severity of the stenosis is assessed by the rate of ventricular filling and direct assessment of the valve area from 2D echo. The echo is extremely useful for examining the size of the left atrium and hence the risk of thrombus formation and embolism. Occasionally, clot can be demonstrated in the left atrium with 2D echo, but is more reliably shown by transoesophageal echo. Doppler ultrasound demonstrates the blood velocity and, by calculation, can give a fair approximation of the pressure gradient across the valve non-invasively. It can also detect mitral regurgitation. Middle-aged adults in the developed countries will almost always be investigated by cardiac catheterization, mainly coronary angiography, prior to consideration for surgery.

Management

who does not require any treatment, with an eventual decision on anticoagulation; • Medical management of the symptomatic patient with a decision on surgery; • Surgical management of the patient with more severe disease.

Medical In the asymptomatic patient in sinus rhythm no treatment is necessary, but the patient should be advised about prophylaxis against infective endocarditis (p. 578) and carefully assessed during pregnancy (p. 582). The onset of AF may lead to the development of symptoms. Even asymptomatic patients in sinus rhythm with moderate stenosis are at risk from thromboembolism from the left atrium, so the relative risks of anticoagulation have to be weighed against the benefits in each case. The indications for anticoagulation in mitral stenosis are listed in Table 12.20. In the symptomatic patient the aim of medical management should be to control the symptoms of left heart failure with diuretics. In AF the ventricular rate should be controlled with digoxin, either alone or in combination with (or substituted by) a -blocker or verapamil (p. 511). The risk of thromboembolic episodes should be reduced by anticoagulation.

Management of mitral stenosis consists of three phases: • Monitoring the asymptomatic patient with mild disease

523

TABLE 12.20 Indications for anticoagulation in mitral stenosis • • • • •

Aetiology

Moderate or severe stenosis Enlarged left atrium Atrial fibrillation - paroxysmal or sustained Sinus rhythm with recurrent palpitations Any symptoms suggestive of arterial embolism

The percentage of valves coming to surgery which are rheumatic is now below 50% in many series. Regurgitation can result from:

Surgical Closed and balloon mitral valvotomy Closed surgical mitral valvotomy is an extremely effective operation, which has been performed successfully even during the second trimester of pregnancy without much hazard to mother or fetus. The operation uses a mechanical dilator to tear the fused commissures. It is introduced via a left thoracotomy, through the apex of the left ventricle. The surgeon also amputates the atrial appendage, thereby removing the source of most systemic emboli. Cardiopulmonary bypass is not required. In recent years the need for this operation has diminished considerably with the development of catheter-based techniques, where a balloon is passed via the femoral vein and across the atrial septum. It is positioned across the mitral orifice and inflated to the appropriate size until the stenosis is relieved. O The balloon technique requires only 1-2 days' hospitalization and avoids the need for a thoracotomy scar, but leaves behind the atrial appendage. The results obtained are comparable to those of the previous surgical series. Symptomatic improvements can be expected to last 4-12 years and sometimes much longer before repeat valvotomy or, more commonly, valve replacement is required. Open valvotomy Open valvotomy through a midline sternal incision, with the patient on cardiac bypass, is necessary if there is doubt about the applicability of valvotomy. It allows direct inspection of the valve and, should regurgitation result from splitting the valve, it can be replaced or even repaired. Valve replacement is essential if there is significant regurgitation or if the valve is heavily calcified. Prognosis

Mitral valvotomy can dramatically improve the symptoms of the patient with mitral stenosis. With time, however, the valve tends to become more rigid and calcify, becoming restenotic even if there is no fusion of the commissures. Often, a degree of regurgitation develops and, occasionally, other valves affected by rheumatic heart disease start to cause problems. OO O Fig. 12.18

524

GMCQ10.11

MITRAL REGURGITATION

Case 12.2

• Dilatation of the valve ring • Damage, retraction or perforation of the valve cusps • Damage to the subvalvular apparatus (chordae, papillary muscle, or the ventricular muscle to which the papillary muscles attach). Non-rheumatic causes of mitral regurgitation are: • Heart failure. Dilatation of the left ventricle will produce 'functional mitral regurgitation' owing to dilatation of the valve ring and a shift in the attachments of the chordae tendinae. This may disappear as the ventricle shrinks with treatment, but may become a permanent feature. • Myocardial infarction. Papillary muscle dysfunction can arise as a result of inferior myocardial infarction, producing a characteristic type of mitral regurgitation which is usually not severe. In contrast, infarction complicated by rupture of the papillary muscle produces catastrophic regurgitation, and is usually fatal unless repaired urgently. • Degenerative changes. These are often seen in mitral valves at surgery. The pathogenesis is not fully understood, as degenerative changes can occur in rheumatic valves or others known to be mildly incompetent for a long period. Rupture of the smaller chordae tendinae can occur slowly and progressively, causing increasing mitral regurgitation. Acute chordal rupture of a major trunk causes sudden severe mitral regurgitation. Other causes In mitral valve prolapse there is prolapse of a mitral valve cusp late in systole, often from a degree of 'redundancy' in the length of the chordae tendinae. It may be a feature of Marfan's syndrome and osteogenesis imperfecta. Elderly patients often develop a minor degree of mitral regurgitation from calcification of the valve ring. This produces a characteristic wheezy systolic murmur. Endocarditis may make even a trivial degree of regurgitation more severe, and acute endocarditis can rapidly destroy a normal valve. There is a degree of mitral regurgitation in most cases of hypertrophic obstructive cardiomyopathy (p. 571). Congenital forms of mitral regurgitation include cleft cusps in AV canal defects and endocardial cushion defects in association with ostium primum atrial septal defects (p. 540). Sharp (stabbing) and blunt (steering-wheel injury) trauma to the precordium, acromegaly and Libman-Sacks endocarditis in systemic lupus erythematosus (SLE) are other rare causes.

IB

Clinical features Symptoms and pathophysiology Chronic The symptoms of chronic mitral regurgitation are essentially the same as those of mitral stenosis, but severe pulmonary hypertension is less common with pure mitral regurgitation. The slowly progressive regurgitation is accommodated by a gradual increase in size of the left ventricle, which ejects the blood into a slowly enlarging and compliant left atrium and pulmonary venous vasculature. Up to 80% of the left ventricular stroke volume may be regurgitant in severe cases, and the left ventricular stroke volume can exceed 250 mL. Symptoms in chronic mitral regurgitation usually develop when the left ventricle begins to suffer deterioration and end-diastolic pressure rises. Usually, systolic function and ejection fraction are well maintained, because the left ventricle is ejecting largely into a low impedance outflow. The symptoms of left heart failure, AF and pulmonary hypertension are the same as for mitral stenosis (p. 522). Acute In acute mitral regurgitation, e.g. complicating acute MI or due to rupture of chordae tendinae, a small left ventricle is regurgitating into a small uncompliant left atrium. This produces a very high pressure pulse or v wave in the left atrium, which may reach 60 mmHg. Left atrial pressure rapidly exceeds that needed to produce pulmonary oedema. The low forward output may precipitate tachycardia, which makes the regurgitation worse. Acute mitral regurgitation thus usually presents with severe pulmonary oedema. Signs Uncomplicated pure mitral regurgitation, if trivial, has only auscultatory signs (Fig. 12.43). However, if the regurgitation is more severe, the left ventricular volume load becomes clinically evident by a hyperactive left ventricle with displacement of the cardiac apex down and laterally. Enlargement of the left atrium in systole may be felt as a parasternal heave. In severe regurgitation the pulse may become small and rather jerky from a shortened ejection time. Severe regurgitation is evident on auscultation by the presence of a third heart sound and short mid-diastolic murmur, produced by the greatly enhanced forward flow through the mitral valve during diastole. The pansystolic murmur of rheumatic mitral regurgitation, or other types of regurgitation with a central jet, is best heard at the apex of the heart and may be louder with the patient tipped to the left. Other types of mitral regurgitation - including the late systolic murmur of the floppy valve syndrome - may be best heard at the left sternal edge, and are often not pansystolic in character. Differential diagnosis from trivial aortic stenosis and ventricular septal defect can therefore be difficult, and requires non-

FIG. 12.43 Features of mitral regurgitation A Illustrates diversion of blood into the left atrium. B Shows the pressures in the left ventricle and left atrium in both mild and severe cases and the accompanying murmurs. PSM = pansystolic murmur; MDM = mid-diastolic murmur. C M-mode echocardiogram through the left ventricle (LV) in a case of severe mitral regurgitation due to valvular disease. The ventricle is dilated and hyperkinetic. D M-mode echocardiogram at the same level in a patient with mitral regurgitation due to poor LV function. The LV is dilated and hypokinetic. IVS = interventricular septum; PW = posterior wall.

invasive tests such as echocardiography and Doppler ultrasound, or even cardiac catheterization.

Mitral valve prolapse Mitral valve prolapse (also known as floppy mitral valve, or Barlow's syndrome) varies in severity from a trivial abnormality picked up on echocardiography, which has little or no consequence to the patient, to severe mitral regurgitation and heart failure. In its mild form there are single or multiple systolic clicks in the mitral area, which may be associated with a late systolic murmur of mild mitral regurgitation. There may be a tendency to supraven-

525

Investigation of mitral regurgitation The ECG is of little value in indicating the severity of the regurgitation. There may be P mitrale and left ventricular hypertrophy, but very severe regurgitation can exist in the presence of an almost normal ECG. Chest X-ray may show enlargement of the left ventricle as well as the left atrium. There may be evidence of raised pulmonary venous pressure. M-mode echocardiography is useful in demonstrating the volume load on the left ventricle and the size of the left atrium. The important distinction of mitral regurgitation due to an abnormal valve from that secondary to poor LV function may be revealed (Fig. 12.44). This has important management implications, perhaps emphasizing a surgical approach for the former and medical antifailure treatment for the latter. The technique will in most cases be able to show a rheumatic thickened mitral valve, mitral valve prolapse or flail cusp. Doppler ultrasound demonstrates mitral regurgitation and distinguishes it from aortic stenosis or a ventricular septal defect. Cardiac catheterization gives a direct measurement of the left ventricular filling pressure and left atrial pressure. Angiography shows the severity of the leak and often its mechanism; it also allows associated coronary disease to be defined. FIG. 12.44 Features of aortic stenosis A Pressures in the left ventricle (black) and aorta (blue) showing a systolic pressure gradient due to aortic stenosis. The ejection systolic murmur is illustrated. B 2D echocardiogram showing severe aortic valve (AV) thickening. Ao = aorta; LV = left ventricle; LA = left atrium; RV = right ventricle. DC] Position of probe and direction of Doppler beam. D Continuous-wave Doppler record showing highvelocity signals. Peak = 4.5 m/s (dottled line), predicting a gradient of 82mmHg.

tricular arrhythmia and atypical chest pain and associated mild skeletal abnormalities, commonly a straight back (loss of thoracic kyphosis). The underlying abnormality consists of redundancy in the chordae tendinae, allowing posterior prolapse of one or other mitral cusp in late systole. More severe degeneration of the subvalvular apparatus is seen in certain patients, particularly those with Marfan's syndrome (Ch. 3, p. 71), although severely floppy mitral valves may occur without obvious cause. Isolated mitral regurgitation, even if severe, can be tolerated for long periods if its development is gradual. A number of patients with the mild 'click-murmur' syndrome suffer significant complications, such as systemic embolization, arrhythmia and endocarditis. Treatment of severe regurgitation is surgical.

O MCQ12.12

526

Management Trivial regurgitation is very well tolerated and requires only endocarditis prophylaxis (p. 578). Moderately severe mitral regurgitation can be well tolerated for years, but the patient may slowly come to accept a diminishing exercise tolerance. The timing of valve replacement can be difficult. Worsening left ventricular function is an indication, as the procedure carries more risk when left ventricular function is poor. Severe regurgitation and acute regurgitation merit early valve replacement. Repair of regurgitant valves is increasingly advocated in anatomically suitable cases. O

MITRAL REGURGITATION WITH STENOSIS The combination of mitral regurgitation with stenosis is almost invariably a result of rheumatic heart disease, but occasionally arises in other conditions, such as LibmanSacks endocarditis in SLE. Mixed mitral valve disease presents later than pure stenosis, and is very common after mitral valvotomy. The physical signs are shown in Figures 12.42 and 12.43. AF is more likely than with pure stenosis. Even a mild degree of regurgitation makes the effective degree of mitral stenosis much greater (p. 521). The predominant lesion can be difficult to assess in the presence of hypertension or aortic valve disease, which will also produce left ventricular hypertrophy. A pansystolic murmur alone may arise from coincident tricuspid regurgitation with pure

mitral stenosis. The degree of mitral regurgitation can vary with the heart rate and the degree of left ventricular dilatation. What may appear to be severe regurgitation when the patient is in heart failure may settle considerably as the patient improves. Management is as for mitral stenosis alone (p. 523). Assessment of the mitral valve with Doppler ultrasound shows that minor degrees of mitral regurgitation are extremely common, even when these cannot be detected clinically.

AORTIC STENOSIS Aetiology Outflow obstruction to the left ventricle can be at the valve itself (valvular), above it (supravalvular) or below (subvalvular). About 1% of cases of 'aortic stenosis' are supraor subvalvular. Supravalvular stenosis is a congenital form of atresia of the ascending aorta, and is associated with a characteristic 'elfin' facies. Subvalvular stenosis is either a congenital ring lesion (which is amenable to surgery) or a variable muscular obstruction produced by hypertrophic obstructive cardiomyopathy (p. 571). Valvular aortic stenosis can result from: • Congenital aortic stenosis, presenting in childhood, the valve having one, two or three cusps; • Bicuspid aortic valve presenting in adulthood. About 1% of the population are born with two, instead of three, cusps. About one-third of these develop calcific aortic stenosis, one-third mixed aortic valve disease, and one-third have no significant valvular abnormality. The valve itself seems to be mechanically unsatisfactory and suffers mechanical stresses, which lead to progressive damage; • Rheumatic heart disease. This involves the aortic valve in about 50% of cases of rheumatic valvular disease, most frequently producing aortic regurgitation, often a mixed lesion but sometimes pure stenosis; • Senile aortic calcification. Although calcification of the valve ring is commonly benign, it can invade the cusps and produce a calcified tricuspid valve as it renders the cusps immobile. This aetiology is more common over the age of 70. The valve cusps are not fused. Pathophysiology

Clinical features

12

Aortic stenosis may remain asymptomatic until it is very severe, and is then life-threatening. The symptoms are: • Left heart failure. Shortness of breath on exertion may precede paroxysmal nocturnal dyspnoea and orthopnoea, or the patient may present with acute pulmonary oedema. • Angina. The thickened myocardium is perfused by a lower than normal perfusion gradient owing to low aortic diastolic pressure and a high left ventricular enddiastolic pressure. For these reasons, angina can occur without any coronary narrowing, although coronary artery disease often coexists in older patients. • Syncope on exertion. This is a sinister symptom which indicates a fixed cardiac output and presages sudden death. The mechanisms may be: skeletal muscle vasodilatation causing a drop in blood pressure; a rise in ventricular pressure producing reflex bradycardia and vasodilatation; myocardial ischaemia triggering ventricular arrhythmia. • Sudden death, probably resulting from myocardial ischaemia and ventricular irritability, leading to a fatal arrhythmia. This may be provoked by exertion. • Right heart failure. Rarely the presentation is right heart failure supervening on left. • Emboli from a heavily calcified valve. These are usually fairly small and cause small strokes or visual disturbances.

The signs of pure aortic stenosis are shown in Figure 12.44. Although the systolic ejection murmur suggests the possible diagnosis, the most important physical sign is the slowrising poor volume pulse, best felt at the carotid, where a systolic thrill may also be evident. The heart is not enlarged unless the patient has been in failure, but the left ventricular apex beat is powerful and 'pressure loaded'. Differential diagnosis from a 'flow murmur' is based on the slow-rising pulse, the presence of left ventricular hypertrophy and the absence of the aortic component of the second heart sound. The loudness of the murmur is a poor index of severity and the murmur may even be absent in heart failure. In congenital aortic stenosis there may be an ejection click as the valve cusps tent as they tension.

Outflow obstruction to the left ventricle leads to: • Concentric left ventricular hypertrophy • Compensatory enlargement of the coronary arterial circulation as the metabolic demands of the hypertrophied left ventricle increase; • A relatively fixed cardiac output. A resting pressure gradient of l00mmHg would have to become 400mmHg for the flow velocity through the valve to double. The heart is incapable of generating such a pressure, and is therefore near its limit at rest.

investigation Electrocardiography A normal ECG virtually rules out significant aortic stenosis, except in congenital disease when the trace may remain remarkably normal despite a large gradient. A number of ECG changes are seen in aortic stenosis: • Left ventricular hypertrophy. Often widening of QRS and T wave changes of 'strain' predominate over voltage

527

SUMMARY 8 Symptoms of aortic stenosis • • • • •

Left heart failure - dyspnoea, orthopnoea, pulmonary oedema Angina Exertional syncope or near syncope Small cerebral emboli Right heart failure

change. T-wave inversion over LV leads may be the only feature. Loss of anterior R waves may occur. • Left axis deviation and, eventually, left bundle branch block may occur either from severe hypertrophy or invasion of the conduction system by calcium from the valve. • Left atrial hypertrophy or 'P mitrale\ A large negative component to the P wave in VI indicates raised left ventricular filling pressure. • Complete heart block.

Medical Treatment of heart failure follows conventional lines (p. 494). Vasodilator agents are contraindicated, as a reduction in systemic resistance can lead to a severe drop in blood pressure, and syncope.

AORTIC REGURGITATION

Chest X-ray The chest X-ray can be normal but may show a poststenotic dilatation of the ascending aorta, prominence of the left ventricle and a calcined aortic valve on lateral chest X-ray. The latter in a patient under 55 suggests tight stenosis; above this age the valve may calcify even with an insignificant gradient. Significant enlargement of the left ventricle occurs with heart failure or coincident aortic regurgitation.

Aetiology

M-mode echocardiography M-mode echocardiography can display thickening and calcification of the valve cusps, but is less useful in congenital stenosis. The compensatory hypertrophy of the left ventricular wall is well shown. 2D echocardiography may give a good view of the aortic valve orifice.

Dilatation of the valve ring occurs classically in syphilitic aortic regurgitation, but can also occur with hypertension and aortic dissection; less common causes include cystic medial necrosis in Marfan's syndrome, and osteogenesis imperfecta. Damage to the valve cusps is a feature of rheumatic aortic regurgitation, infective endocarditis, and less common conditions such as Libman-Sacks endocarditis, mucopolysaccharidoses and pseudoxanthoma elasticum. The seronegative arthritides, ankylosing spondylitis, Reiter's disease and psoriatic arthritis can be associated with an aortitis and cusp damage, leading to aortic regurgitation. Rarely, rheumatoid arthritis is associated with a nodular cusp damage leading to aortic regurgitation. Bicuspid aortic valve (found in 1% of the population) can lead to predominant aortic regurgitation (p. 527). There are other congenital forms of aortic regurgitation, e.g. that associated with a supracristal ventricular septal defect.

Doppler ultrasound Doppler ultrasound measures the jet velocity and predicts the valve gradient. A high blood velocity (>4m/s) indicates a surgically significant gradient, but as jet velocities can be underestimated, a low jet velocity record is a less reliable finding. Cardiac catheterization Measurement of the valve gradient by cardiac catheterization is not essential if the diagnosis is well established. Catheterization is usually indicated where there is doubt about the site or degree of stenosis, a not-infrequent occurrence in older patients, who may be poor ultrasound subjects. It also allows analysis of ventricular function. Coronary angiography is now considered essential prior to operation in all but the very elderly.

Management

528

tomy in children may gain sufficient time for them to reach a reasonable size before having aortic valve replacement. For adults, valve replacement is the only treatment. If the ventricle has already been badly damaged the results are less good, but in the patient with a normal ventricle the left ventricular hypertrophy regresses fairly rapidly. Lesser degrees of stenosis need serial follow-up, as the stenosis may become more significant with time at a rate averaging about 8mmHg/yr, although this may be very variable.

Surgical Tight aortic stenosis has such a poor prognosis that surgery should be offered to all patients. Valvuloplasty or valvo-

The aortic valve cusps normally overlap on closure to the extent of about 25% of the cusp area. Aortic regurgitation can arise from: • Dilatation of the valve ring so that the cusps no longer meet adequately to prevent leakage • Damage to the cusps themselves • In some diseases, a combination of these factors.

Clinical features Symptoms and pathophysiology Aortic regurgitation can be well tolerated for years if it develops slowly. Up to 80% regurgitation can be found in severe chronic aortic regurgitation, but acute regurgitation (such as follows aortic dissection or acute infective endocarditis) is poorly tolerated. The severity of the regurgitation is normally indicated

12

by a low diastolic blood pressure and wide pulse pressure. These may appear more 'normal' in heart failure as the stroke volume falls and the end-diastolic pressure of the left ventricle rises. As aortic regurgitation progresses, the increasing leakage requires a larger and larger forward flow. This is usually achieved not by a tachycardia but by the left ventricular end-diastolic size increasing with an increased stroke volume. Initially, the ejection fraction of the left ventricle is well maintained and exercise tolerance is excellent. However, after an unpredictable time there is a steady or sudden deterioration in left ventricular function, usually with a great increase in heart size and the development of symptoms of left ventricular failure. Eventually, right heart failure follows, with the development of pulmonary hypertension. The development of heart failure indicates serious dilatation of the left ventricle and the need for urgent consideration of valve surgery. Heart failure in aortic regurgitation is associated with rapidly worsening left ventricular function, which may never recover even with valve replacement. Angina can develop because the dilated left ventricle has increased oxygen requirements (as in congestive cardiomyopathy), but is usually associated with coronary artery disease. AF occurs in about 15% of cases, usually those with long-standing failure. Physical signs The physical signs of aortic regurgitation (Fig. 12.45) are related to the large stroke volume, the peripheral vasodilatation and the compensatory increase in size of the left ventricle. The pulse may be collapsing or bisferiens (p. 471), or feel normal if there is heart failure. The blood pressure indicates the large pulse pressure and low diastolic pressure. If the diastolic pressure is well maintained in the presence of severe aortic regurgitation, coincident hypertension should be suspected. The left ventricle is very active, and in severe cases the apex beat is displaced. The increased forward flow is often accompanied by a systolic flow murmur, which, of itself, does not indicate coincident stenosis. The early diastolic murmur is notoriously difficult to hear. The murmur is best heard with the diaphragm of the stethoscope, with the patient sitting forward having breathed out. It may be best heard at the left sternal edge, nearer the apex or in the aortic area, depending on the direction of the jet. Typically, valve ring dilatation regurgitation is better heard in the third right interspace rather than the third left. An Austin Flint murmur may be associated with aortic regurgitation. This is an apical diastolic murmur, similar to that of mitral stenosis, arising from the anterior cusp of the mitral valve, which vibrates in the jet of aortic regurgitation. Because aortic regurgitation may be both difficult to hear and is a frequent lesion in infective endocarditis, a patient with a fever and steadily widening pulse pressure is regarded as having aortic regurgitation and endocarditis until proved otherwise.

FIG. 12.45 Features of aortic regurgitation [A] Pressures in the aorta (blue) and left ventricle (black) are illustrated with the accompanying murmurs (3 = third sound). B Direction of regurgitant flow (arrow) and orientation of M-mode and Doppler beams (dotted lines), C Continuous-wave Doppler signal showing a regurgitant jet throughout diastole with a maximal velocity of 4m/s (upwards signal). The small downward signal excludes aortic stenosis. D M-mode echocardiogram showing flutter of the anterior mitral valve leaf (arrowed), which may close prematurely in severe aortic regurgitation, causing an early first heart sound.

Investigation ECG shows left ventricular hypertrophy of the diastolic overload type, in that, initially at least, voltage changes with prominent Q waves over the lateral leads predominate over T-wave changes. Later, the T waves invert. Chest X-ray shows increasing size of the left ventricle and often some dilatation of the proximal aorta. Generalized cardiac enlargement may follow and, eventually, changes of raised pulmonary venous pressure. Echocardiography. M mode shows the dilated left ventricle and its wall thickness. Calculation can be made of the stroke volume and ejection fraction, which are useful for following progress. In many cases vibration of the anterior cusp of the mitral valve or the septum can be seen, confirming the diagnosis. Doppler ultrasound will confirm the diagnosis but cannot quantify the lesion. Cardiac catheterization is necessary to examine left ventricular function, the severity of the aortic regurgitation and its anatomy, as well as looking for other pathology, such as mitral regurgitation and coronary artery disease.

529

SUMMARY 9 Clinical features of aortic regurgitation Symptoms Left heart failure: breathlessness Myocardial ischaemia, angina Atrial fibrillation (less frequently) Physical signs Large stroke volume with vasodilatation Collapsing pulse Wide pulse pressure Head nodding, capillary pulsation Enlarged left ventricle with dynamic displaced apex Auscultatory High-pitched early diastolic murmur Forward flow systolic murmur Premature first heart sound (in severe cases) Austin Flint murmur

Management Surgery Those in heart failure have been left too late to obtain the maximum benefit from surgery, as a dilated 'myopathic' heart never returns to normal and is associated with a high risk of sudden death. The currently used criteria for surgery are based on echocardiographic dimensions; others are now being developed based on the use of isotope left ventricular angiography. A fall in the ejection fraction of the left ventricle on exercise has been suggested as a criterion for surgery, as the ejection fraction is normally well maintained or even increases. Acute aortic regurgitation requires very close attention and urgent valve replacement at the first sign of heart failure. Delay can result in catastrophic heart failure and death, as the degree of regurgitation and size of the heart rapidly increase.

Medical Medical management of chronic regurgitation consists of treatment of heart failure and control of other problems, such as arrhythmias and endocarditis prophylaxis (p. 578). There is some evidence that using vasodilators such as nifedipine can delay the need for valve replacement when begun in moderate regurgitation. O

Investigation and management Assessment of combined lesions has been greatly helped by Doppler ultrasound and echocardiography. The gradient measured by Doppler ultrasound may appear much greater than the pull-back pressure gradient measured at cardiac catheterization. Surgery for combined lesions is based on the same criteria as for single lesions. Onset of symptoms means that there is severe left ventricular hypertrophy and probably fibrosis, with consequent poor left ventricular function. Surgery is therefore ideally timed to anticipate symptoms and allow the left ventricular hypertrophy to regress. A gradient of over 50mmHg, the development of a left ventricular strain pattern on ECG, left ventricular dilatation and any evidence of heart failure are all considered indications for aortic valve replacement.

COMBINED MITRAL AND AORTIC VALVE DISEASE Over 50% of patients with rheumatic mitral valve disease will eventually develop evidence of aortic valve disease, most commonly aortic regurgitation, but sometimes stenosis as well. The presence of disease in both valves makes assessment of the severity of the individual lesions more difficult. Mitral valve disease in particular tends to reduce cardiac output and stroke volume, which may make the physical signs of aortic valve disease, such as the carotid pulse, far less evident. Aortic regurgitation and mitral regurgitation both cause a volume load on the left ventricle, so the contribution of each lesion to this is difficult to assess. The auscultatory signs may also be far more difficult to elicit, particularly if the patient is in heart failure.

MIXED AORTIC VALVE DISEASE

Investigation and management

About one-third of patients with aortic valve disease have a combined lesion with significant regurgitation and stenosis. As previously described, regurgitation makes a great difference to the valve gradient (p. 521, Table 12.19), and a stenotic valve with a gradient of 16mmHg in systole would have a gradient of over 60mmHg if twice the flow

The presence of aortic valve disease should be suspected in every patient with rheumatic mitral valve disease, particularly if they have ECG evidence of left ventricular hypertrophy, or a slightly sustained carotid pulse. Fortunately, echocardiography and, in particular, Doppler ultrasound enable the valves to be individually examined for evidence of stenosis and regurgitation. This is far more accurate than inspection of the valves at surgery. If a patient is being considered for mitral valve replacement in the presence of aortic valve disease, it is unwise for the surgeon to leave anything but trivial aortic valve disease.

O MCQ 12.13 530

were accommodated, because of 50% regurgitation. The physical signs of the combined lesion are distinctive, with a large-volume sustained or bisferiens pulse, a markedly hypertrophied left ventricle which is both pressure and volume loaded, and both systolic ejection and early diastolic murmurs. The second sound is single. A combined lesion cannot be caused by syphilis or other pathologies that primarily dilate the aortic root.

12 FIG. 12.46 Features of tricuspid regurgitation fAl Illustration of regurgitant flow and direction of Doppler beam (dotted line). PA = pulmonary artery; SVC and IVC = superior and inferior vena cava. Q] and [C] Continuous-wave Doppler signals, both showing tricuspid regurgitation (downwards signal). In B there is severe (organic) regurgitation with normal pulmonary artery pressure and increased tricuspid forward flow. In C there is pulmonary hypertension with a high-velocity regurgitant jet through the tricuspid valve.

What may seem a trivial degree of aortic regurgitation or stenosis at presentation may become clinically important with replacement of the mitral valve.

TRICUSPID REGURGITATION Tricuspid regurgitation most commonly arises as a result of dilatation of the valve ring secondary to right heart failure with pulmonary hypertension. Less common causes are actual rheumatic involvement of the tricuspid valve leaflets producing 'organic tricuspid disease', and rightsided endocarditis, which is rare in anyone other than intravenous drug addicts. Floppy tricuspid valve may be associated with floppy mitral valve, and some congenital heart disease is associated with tricuspid regurgitation either as a primary phenomenon or secondary to right heart enlargement (e.g. atrial septal defect). Endomyocardial fibrosis, a disease occurring in the tropics, especially Africa, causes mitral and tricuspid regurgitation. Doppler ultrasound has shown that a right-sided cardiac enlargement in rheumatic heart disease is almost invariably associated with a regurgitant jet through the tricuspid valve, very often in the absence of any physical signs. Similarly, it may be the cause of an 'innocent systolic murmur'.

Clinical features Pathophysiology and symptoms The leakage of a small percentage of the right ventricular stroke volume is easily accommodated by the right atrium. As the volume of regurgitation increases, there is increasing enlargement of the right atrium and right ventricle, and the movement of the interventricular septum becomes dominated by the right ventricle and so 'paradoxical'. The large systolic pressure pulse in the right atrium leads to

stagnation and, eventually, reversal of blood flow in the great veins; the liver becomes distended and pulsatile (Fig. 12.46). The high hepatic venous pressure leads to cardiac cirrhosis, and this may occasionally lead to secondary portal hypertension and splenomegaly. The poor venous return, high venous pressures and poor hepatic function are associated, in most patients, with a degree of cardiac cachexia, poor wound healing and difficulties with haemostasis. Tricuspid regurgitation due to disease of the valve is much better tolerated, as the regurgitation is at a much lower pressure than when it is secondary to pulmonary hypertension. The principal symptoms are those of right heart failure, with low cardiac output, fatigue, oedema and pain from hepatic congestion which may arise on exercise - so-called 'hepatic angina', i.e. epigastric or right-sided subcostal pain arising on exercise and gradually (more gradually than 'true' angina) disappearing with cessation of exercise. Signs The patient may have AF and is often slightly jaundiced. In moderate to severe cases there is a highly pulsatile jugular vein with a large v wave and rapid y descent. The height of the v wave peak may be as much as 15cm above the sternal angle, and in these circumstances the jugular vein is palpably pulsatile and can be confused with an arterial pulse. The right ventricle is very active and gives rise to a right parasternal impulse. A pansystolic murmur is heard at the left sternal edge, or even at the cardiac apex, which may be formed by the right ventricle under these circumstances. The murmur is often lower in pitch than a mitral murmur. There can be a right ventricular third heart sound. Abdominal examination reveals a pulsatile liver, which is often visible but is best felt bimanually.

531

Investigation ECG may be unhelpful and is usually dominated by other factors associated with the underlying cardiac disease. Chest X-ray shows right-sided cardiac enlargement, and there may be a fullness in the region of the distended superior vena cava. O Echocardiography shows an enlarged right ventricle with paradoxical movement of the interventricular septum suggesting 'volume load' of the right ventricle. The tricuspid valve leaflets are more easily seen with an enlarged right heart. Doppler ultrasound will confirm a regurgitant jet through the tricuspid valve. The jet velocity will give an indication of the gradient across the valve, and hence the right ventricular pressure and the severity of the pulmonary hypertension (if any).

Management As tricuspid regurgitation is often secondary to pulmonary hypertension, the treatment is frequently that of the underlying cause. Bed rest and diuretics will be effective in many cases, with the murmur and pulsatile jugular veins disappearing. Intravenous diuretics may be required, as the oral drugs may be poorly absorbed. Organic tricuspid valve disease may require surgical treatment by replacement or annuloplasty at the same time as other cardiac surgery.

TRICUSPID STENOSIS Tricuspid stenosis is a very uncommon manifestation of rheumatic heart disease. It mimics mitral stenosis, but produces right heart failure rather than left. It almost always complicates rheumatic involvement of the mitral and aortic valves, and should be suspected if there is a high venous pressure with a slow y descent in the JVP. The diastolic murmur is reputedly enhanced on inspiration and is best heard at the left sternal edge. There can be a tricuspid opening snap, and often coincident tricuspid regurgitation.

Investigation and management The diagnosis can be confirmed by echocardiography, which will show the thickened cusps. Doppler ultrasound can be used to measure the jet velocity and thus calculate the diastolic gradient. Cardiac catheterization with a double-lumen catheter is needed to measure the rather small valve gradient. Treatment is by valve replacement.

O Fig. 12.19

532

PULMONARY REGURGITATION Pulmonary regurgitation is usually caused by dilatation of the pulmonary valve ring in pulmonary hypertension. It may also follow pulmonary valvotomy or valvuloplasty. In severe cases the right ventricle is hyperactive from volume loading, and the murmur is difficult to distinguish from that of aortic regurgitation, being an early diastolic murmur at the left sternal edge starting immediately after the pulmonary component of the second sound. It may have a slightly lower pitch and be more obvious on inspiration, but the two can easily be confused. The Graham Steel murmur of pulmonary regurgitation in mitral valve disease is far more frequently due to aortic regurgitation than was formerly believed. Pulmonary regurgitation is usually well tolerated but can be one reason why the heart remains enlarged after surgical correction of Fallot's tetralogy.

PULMONARY STENOSIS Pulmonary stenosis is the second most common form of congenital heart disease in adults; it is commoner in women. Mild cases, with gradients of less than SOmmHg, are common. Stenosis may occur at valvular or subvalvular level, or even in the pulmonary arteries. There is a raised pressure in the right ventricle and, in severe cases, reduced lung blood flow and poor effort tolerance.

Clinical features Most cases are asymptomatic. With gradients of over SOmmHg, the most common symptom is dyspnoea on exertion. In very severe stenosis symptoms of exertional syncope and angina become prominent, similar to those associated with aortic stenosis. In children and young adults, pulmonary stenosis may present with severe right heart failure. The physical signs are of a small volume pulse, a large a wave in the JVP, a right ventricular substernal heave and a harsh systolic murmur in the pulmonary area. The latter appears very long and drowns the heart sounds in severe cases of infundibular stenosis, but can be a short ejection murmur in mild cases. The timing of pulmonary valve closure, which is delayed in severe cases, is a useful indication of severity.

Investigation and management ECG shows right ventricular hypertrophy. The chest X-ray shows oligaemic lung fields; a large poststenotic dilatation with dilatation of the pulmonary conus is common in valvular, but not infundibular, stenosis. The right ventricle is not significantly enlarged unless it has failed. Doppler ultrasound assesses the valve gradient. Echocardiography will show the hypertrophy of the right ventricle and the site of the obstruction.

Surgical valvotomy has been largely replaced as a treatment for pulmonary stenosis by balloon valvuloplasty (see below).

SURGICAL MANAGEMENT OF ACQUIRED HEART VALVE DISEASE The choices in surgical management of heart valve disease have recently become more complex, with closed valvotomy, open valvotomy and balloon valvuloplasty all being possible approaches.

Cardiac bypass In the technique of cardiac bypass the circulation is arrested and there is extracorporeal circulation from a roller pump and external oxygenator. The brain is perfused at relatively low pressure, and the oxygenator and perfusion system introduce microemboli into the arterial circulation. There is therefore a morbidity associated with prolonged cardiac bypass. Most of the morbidity seems to be reversible (e.g. intellectual impairment) but patients with cerebrovascular disease, such as carotid stenosis, are at particularly high risk from cerebral infarction.

Balloon valvuloplasty In balloon valvuloplasty a balloon catheter is inserted through the valve orifice and then dilated to split open the valve. The aortic valve can be dilated in children, and balloon valvuloplasty has been especially successful for pulmonary and mitral stenosis. If torrential regurgitation is produced during mitral valvuloplasty, valve replacement is essential.

Valve repair and replacement Plastic surgical repair to a damaged valve uses pericardial or other tissue to repair defects in the valve cusps, or suturing of the valve ring to a prosthetic former to correct a grossly dilated valve ring. The latter is commonly performed in severe tricuspid regurgitation. The procedures are time-consuming and technically demanding and have had a history of variable results. Prosthetic and biological valves The type of valve used for replacement depends very often on local preference and expertise. Mechanical valves require permanent oral anticoagulation with warfarin. Biological valves do not, but are prone to early failure and calcification. The original prosthetic Starr-Edwards valve essentially a ball in a cage - was built for durability, but early models formed thrombus on the struts and were also rather noisy. Numerous attempts have been made to improve on this, but many have developed mechanical problems and the Starr valve remains the most durable prosthesis currently available. In order to avoid the anticoagulation essential with

mechanical valves, various attempts have been made to use biological ones. Fresh human homografts have been used very successfully in the aorta, but availability has been a problem. The most widely used biological valves are xenografts. Porcine or bovine tissue is usually mounted on a ring or stent; such valves have the advantage of not requiring long-term anticoagulation, nor do they produce the loud prosthetic clicks of mechanical valves. Their disadvantages are that they have an unpredictable lifespan and can calcify and become stenosed very rapidly in young patients. However, most of them last at least 5 years. They may be considered in women of childbearing age who wish to have a family without the teratogenic risks of anticoagulation. Such patients must be prepared to face the subsequent reoperation for valve replacement when the biological valve fails. Under current UK legislation a heavy-goods or publicservice vehicle licence holder cannot drive if he or she is on anticoagulants, and so these patients may have to opt for biological valves if they wish to maintain their occupation.

12

FURTHER READING ON VALVULAR DISEASE Guidelines on the investigation and management of valvular heart disease July 1996 Prendergast B, Banning A P, Hall R J C on behalf of a working group of the British Cardiac Society and the Royal College of Physicians, London

CONGENITAL HEART DISEASE

Incidence and aetiology Congenital heart abnormalities are present in about 8-12 per 1000 live births (excluding bicuspid aortic valve and floppy mitral valve). The simpler abnormalities are commonest, with ventricular septal defect (VSD) accounting for 25-30% of cases and patent ductus arteriosus 10% (Table 12.21). Pulmonary stenosis, atrial septal defect (ASD), coarctation of the aorta, aortic stenosis and tetralogy of Fallot together account for another 35% of cases. Of the complex lesions, transposition of the arteries, AV septal defects and the hypoplastic left heart syndrome are the most frequently seen. The fetal circulation is shown in Figure 12.47. Only a minority of these patients (10-15%) would survive to adolescence and adulthood without the help of cardiac surgery and interventional cardiology. Over the last 15-20 years babies have survived who would previously have died, so that cardiologists and physicians are now more likely to see adult patients with medically or surgically corrected lesions than to diagnose a congenital defect for the first time in adult life. The relative frequency of different lesions at different ages depends on the compatibility of the lesion with longevity. ASD, for example, can present in the patient's

533

TABLE 12.21 Relative frequency of types of congenital heart disease Congenital defect

Frequency (%)

Live births Atrial septal defect Pulmonary stenosis Tetralogy of Fallot Ventricular septal defect Eisenmenger syndrome Patent ductus arteriosus Coarctation of the aorta Other

30 25 10 9 7 7 5 7

Stillbirths Ventricular septal defect Univentricular heart (with tricuspid and mitral atresia) Atrial septal defect Coarctation of the aorta Complete transposition Atrioventricular septal defects Hypoplastic left heart Tetralogy of Fallot Other

35 13 10 9 9 9 6 3

TABLE 12.22 Causes of congenital heart disease Causes

Defect

Rubella

Patent ductus, atrial septal defect, pulmonary stenosis

Drugs Alcohol

Ebstein's anomaly Tricuspid atresia

Thalidomide

Multiple defects

Inherited disease Connective tissue disorders Marfan's Ehlers-Danlos Inborn metabolic errors Mucopolysaccharidoses Homocystinuria Pompe's disease Chromosomal abnormalities Down's syndrome Trisomy 13 and 18

6

Turner's syndrome Other inherited syndromes Di George's Friedreich's ataxia Holt-Oram Kartagener's Tuberous sclerosis

FIG. 12.47 Diagram of the fetal circulation SVC and IVC = superior and inferior vena cava.

OMCQ 12.14 534

Ventricular septal defect

Lithium

Aortic dilatation and incompetence; mitral incompetence Mitral regurgitation; arterial dilatation Valve disease, cardiomyopathy Aortic and pulmonary dilatation Glycogen deposition in myocardium Atrial and ventricular septal defect; Fallot's tetralogy; endocardial cushion defect Ventricular septal defect; right ventricular anomalies; pulmonary stenosis Coarctation of aorta, bicuspid aortic valve; pulmonary stenosis Fallot's tetralogy; aortic arch anomalies Cardiomyopathy Atrial septal defect Dextrocardia Cardiomyopathy

70s, but congenital aortic stenosis or Coarctation of the aorta presents much earlier, either to the physician or as a cause of sudden death. As the cohort of patients with surgically palliated or corrected hearts reaches maturity, so special issues arise involving counselling, the advisability of pregnancy, the likelihood of inheritance in the next generation, and the timing of heart or heart-lung transplantation in the disabled or deteriorating patient. The commonest forms of congenital heart disease seen after birth are shown in Table 12.21. Spontaneously aborted fetuses have a much higher incidence of severe congenital heart disease than do normal-term infants. Congenital heart defects may occur as part of more complex syndromes; there are often associated minor thoracic abnormalities. Most cases of congenital heart disease remain unexplained. Some causes are listed in Table 12.22. Most lesions develop in the first trimester of pregnancy, many at about the 7th week. Many involve 'mistakes' in the complex

IB FIG. 12.48 Diagrammatic representation of blood flow in the systemic and pulmonary circulation A Normal. B Left-to-right shunt. El Right-to-left shunt; the pulmonary blood flow is reduced as a result of development of high resistance. RH = right heart; LH = left heart.

embryological processes of flexing, twisting and formation of septa and the formation and closure of foramina (Fig. 12.48). The formation of a patent ductus arteriosus (more correctly a persistent ductus arteriosus) and coarctation of the aorta appear to be due to physiological mechanisms that go wrong after birth.

CHROMOSOMAL ABNORMALITIES Chromosomal abnormalities can be accompanied by congenital heart disease. Down's syndrome (mongolism or trisomy 21, see Ch. 3, p. 66) is associated with endocardial cushion or AV canal defects (ASD mitral and tricuspid regurgitation), but VSDs and pulmonary stenosis also have an increased incidence. Turner's syndrome (XO) - characterized by stunted growth, web neck, female sexual stunting and cubitus valgus - has a greatly increased incidence of coarctation of the aorta and pulmonary stenosis. Maternal viral illnesses Viral infection in early pregnancy, typically rubella, appears to be the cause of some congenital heart defects. Rubella produces a characteristic association of defects, including deafness, cataract, and often persistent ductus arteriosus, although other lesions may occur. Although rubella is the best-documented viral agent, mumps and influenza epidemics have been shown to increase the risk of congenital defects if the mother has the illness early in pregnancy. Other agents, e.g. cytomegalovirus, herpes simplex and coxsackie virus, have also been implicated. About 5% of cases of congenital heart disease can be explained by viral illness. Maternal illness and drug effects Maternal illness, especially SLE, may predispose to congenital AV block. Drugs taken by the mother can be teratogenic, and drugs now associated with congenital heart disease include alcohol, lithium, amphetamines, phenytoin, oestrogen/progesterone and trimethadione. Counselling, pregnancy and contraception A mother who has congenital heart disease herself, or who has one affected child, can usually be counselled that sub-

sequent pregnancy carries a low risk (1-5%) of congenital heart disease. Concordance rates for identical twins are between 15 and 20%, and for non-identical twins between 3 and 5%, suggesting environmental as well as genetic factor(s) in the development of congenital heart disease. In families with affected first-degree relatives the risk to the fetus increases to 50% when there are three affected relatives. When both spouses have congenital heart disease the risk may be up to 15%. In all these instances professional help from clinical geneticists is invaluable. Fetal echocardiography can be used to screen high-risk pregnancies. Mendelian dominant inheritance is seen in Marfan's syndrome and some families with hypertrophic cardiomyopathy; both conditions present in adulthood. Pregnancy in cyanotic congenital heart disease carries a high risk of complications and accelerated deterioration, and frequently ends in spontaneous abortion. At particular risk are women with haemoglobin above 17g/dL and those with pulmonary hypertension or the Eisenmenger reaction (p. 536). Patients with lesser degrees of cyanosis and polycythaemia may occasionally be brought through successful pregnancies by a combination of rest and prevention of thrombosis, infection or prolonged labour. Sterilization is usually recommended for high-risk (Eisenmenger) women. Even this relatively minor surgical procedure can be dangerous, but is less so than pregnancy. High-oestrogen-containing contraceptive pills increase an already high risk of thrombosis and embolization; it is not clear if the low-oestrogen pills are any safer in this group of women. The potential associated infective risk prevents the use of intrauterine devices. Vasectomy for the male partner and barrier methods are alternatives, but are too unreliable. O

ACQUIRED COMPLICATIONS OF CONGENITAL HEART DISEASE Apart from the simplest conditions of ASD, PDA and VSD, where early surgical correction is associated with few long-term consequences, many operated and unoperated patients develop complications. Pulmonary vascular disease, myocardial failure, arrhythmia, systemic embolization, endocarditis and degeneration of prosthetic valves

535

and conduits are some of the complex problems posed by the survivors of congenital heart defects. Their optimal management calls upon many skills within cardiology and other disciplines. In the future long-term management of this group of patients is likely to be concentrated in a few centralized units containing all the expertise required and a large enough cohort of patients to further define treatment strategies - the so-called GUCHD units (grown-up congenital heart disease).

CYANOTIC CONGENITAL HE ART DISEASE Central cyanosis occurs with an arterial oxygen saturation of <85%. Milder degrees of cyanosis may not be evident at rest but are readily precipitated by mild exertion at the bedside. A degree of shunting through the ductus arteriosus occurs for several hours, or even days, after birth, and this may persist. More commonly, right-to-left shunting occurs through persisting embryonic defects or through the foramen ovale (Fig. 12.48). The persistence of central cyanosis after birth, or its later onset, depends on a left-toright communication and on the resistance to outflow from the right heart being higher than that of the left. High resistance to outflow of the right heart is usually caused by congenital pulmonary stenosis (valvular or muscular) or by acquired pulmonary hypertension due to constriction of the pulmonary resistance vessels. Acquired pulmonary hypertension occurs with high flow through the lungs, so a left-to-right shunt can eventually reverse and become right-to-left, with cyanosis (the Eisenmenger reaction). Cyanotic heart disease produces a characteristic appearance, with central cyanosis (which may vary with activity), clubbing, impaired growth and intellectual development (if severe) and marked polycythaemia after the first year of life. Complications of the right-to-left shunt include cerebral abscess and paradoxical emboli, in which embolism from clot in peripheral veins gains access to the systemic circulation through the shunt. The severe polycythaemia can lead to thrombosis and infarction. Polycythaemia is proportional to the degree of desaturation. Regular venesection may be needed to maintain haemoglobin levels of 16-18 g/dL. Pulmonary hypertension can be complicated by recurrent haemoptysis.

FALLOT'S TETRALOGY Pallet's tetralogy is the commonest form of cyanotic con-

Case 12.3 536

genital heart disease seen after the age of 1 year. The four main features of this abnormality (see Fig. 12.49) are: • A ventricular septal defect (through which the shunting occurs); • Pulmonary stenosis, which may be muscular or valvular, and acts as the obstruction that leads to the right-to-left flow; • An overriding aorta. The aorta is positioned above the VSD; • Right ventricular hypertrophy, due to the pressure and volume load to which the RV is subjected. Incomplete forms of the condition exist in which not all the features are present and the child may be acyanotic - 'pink tetralogy'.

Clinical features The child may be deeply cyanosed at rest and is subject to cyanotic and exertional syncopal episodes which appear to be related to changes in the degree of shunting. The physical signs are cyanosis, clubbing, right ventricular hypertrophy, a harsh systolic murmur at the upper left sternal edge (often with a thrill) and a single second heart sound. Children with tetralogy characteristically learn to squat, which increases systemic vascular resistance and so diminishes R-L shunting.

Investigation The characteristic features of Pallet's tetralogy on investigation are illustrated in Figure 12.49. Chest X-ray shows a characteristic cardiac silhouette of severe right ventricular hypertrophy which lifts the cardiac apex (coeur en sabot) and oligaemic lung fields. ECG shows right ventricular hypertrophy. Echocardiography shows the overriding aorta and VSD and, with Doppler, can demonstrate the pulmonary stenosis, the right ventricular pressure and, sometimes, the flow through the VSD. Cardiac catheterization can be used to measure the pressures and show the severity of the outflow tract obstruction.

Surgical treatment Total surgical correction is now often attempted at an early age, but palliative operations to increase the lung blood flow were common in the past and are sometimes required now to encourage growth of the pulmonary vasculature, prior to complete intracardiac repair. The palliative shunts included the Blalock-Taussig anastomosis (subclavian to pulmonary artery) and aortic to pulmonary shunts, the Waterston and Potts shunts.

Prognosis and clinical features after surgery Without correction, less than 5% survive to adult life; 90% survive with surgery. In future, adult patients will have had

LESS COMMON FORMS OF GYANOTIC CONGENITAL HEART DISEASE

12

Some of the less common forms of cyanotic congenital heart disease are important causes of neonatal and infant death; however, they can be diagnosed accurately with modern imaging techniques and treated surgically. Even in neonates, cardiac transplantation and heart-lung transplantation may make some formerly inoperable conditions amenable to surgery.

Transposition of the great arteries

FIG. 12.49 Features of Pallet's tetralogy A Diagram of abnormality. VSD = ventricular septal defect; PA = pulmonary artery; B = aorta. B Physical signs. C Chest X-ray showing right ventricular hypertrophy, typical 'coeur en sabot'. D EGG showing right atrial and ventricular hypertrophy and right axis deviation.

a total repair at an earlier age, probably without prior palliative surgery. Late mortality, 10-15 years after repair, is low, at 2-7%, and mainly comprises arrhythmia, including complete heart block. The vast majority of patients have normal exercise capacity and those that do not often have residual ventricular septal defects or pulmonary hypertension. These patients virtually all have a crescendodecrescendo murmur. Many have a decrescendo murmur of pulmonary incompetence; signs of a residual VSD are somewhat less common. The chest X-ray can be normal, but often shows cardiomegaly, and right bundle branch block on the ECG is almost always seen. After palliative surgery the radial pulse may be absent or weak on the side of the shunt and arterial collaterals may develop. The shunt may produce a continuous murmur. O

Complete transposition In this uncommon abnormality the atria and ventricles are in normal relationship to each other but the great arteries arise from the wrong ventricles (transposed). The pulmonary and systemic circulations are therefore operating in parallel rather than in series, which is incompatible with life unless there is some communication between the two circuits. About half the neonates have an intact ventricular septum, and death ensues as the patent ductus arteriosus closes naturally. Survival is dramatically improved by a catheter method of producing an ASD (Rashkind procedure). The other half of the group have large VSDs allowing bidirectional shunting and mixing of the bloodstream; 20% have a single ventricle, usually the right, now classified as double-outlet right ventricle (DORV). Double-outlet left ventricle (DOLV) is very rare. The natural history of this group is very variable, with some individuals surviving to adulthood. Children who survive thanks to the Rashkind procedure have few physical signs but remain cyanosed. Corrective surgery The 'Mustard' procedure produces an intra-atrial repair using a pericardial baffle to redirect flow in the atria to the appropriate ventricle. Results have been excellent, with a return to normal saturations, although arrhythmias may be a problem in later life. Concern over the ability of the right ventricle to function in later life has led to the development of an arterial switch operation and the Rastelli procedure, where a valved conduit connects the right ventricle to the pulmonary artery. It is not yet clear whether these two demanding operations will improve on the results achieved with intra-atrial repairs. Mustard baffles may develop stenoses or leaks in later life, requiring careful evaluation by echocardiography and catheterization. Relief of baffle obstruction may be required and achieved percutaneously using balloon dilatation techniques.

Congenitally corrected transposition Here the great arteries are transposed but blood flows

537

correctly, as the left and right ventricles are inverted. Thus, blood flows from the right atrium through a bicuspid mitral valve into a morphological left ventricle and into the pulmonary artery. From the left atrium a tricuspid valve leads to the right ventricle and the aortic valve. Survival without surgery is common and the clinical features are dominated by associated abnormalities and problems (these include VSD, single ventricle - DORV or DOLV - and pulmonary stenosis or pulmonary vascular disease).

Abnormalities of venous return The systemic veins may drain into the left atrium, producing cyanosis with exercise. Both situations are rare on their own and are associated with other serious congenital defects. Total anomalous pulmonary venous drainage (TAPVD) into the right atrium or left innominate vein is associated with right-to-left shunting across an ASD or patent foramen ovale. The physical features are those of an ASD with central cyanosis. Operations to reroute the veins and close the ASD produce excellent long-term results, although adults are susceptible to atrial arrhythmia.

Tricuspid atresia Unoperated adults with this condition do exist, although survival through childhood is unusual without surgical intervention. Blood flows from the right atrium to the left, through the mitral valve into the left ventricle and via a VSD to a not uncommonly stenosed right ventricular outflow tract. The clinical signs are dominated by pulmonary stenosis and VSD murmurs. Cyanosis is variable and depends on the size of the VSD and degree of pulmonary stenosis. Surgical correction is by a BlalockTaussig shunt or connection of the right atrium to the right ventricle, more commonly by Fontan's procedure, where the right atrium is connected to the pulmonary artery.

Ebstein's anomaly of the tricuspid valve This condition is often acyanotic. The septal and posterior leaflets of the tricuspid valve are attached to the right ventricular wall so that a large part of the right ventricle becomes atrialized. The right ventricle is small, but the right atrium becomes enormous. There is often some tricuspid regurgitation, and occasionally the abnormal valve obstructs filling, or even emptying, of the right ventricle. There is a high incidence of patent foramen ovale or ASD, and supraventricular arrhythmias are very common. The condition often presents in adult life. It may remain asymptomatic for years and present with an abnormal chest X-

Fig. 12.20

538

ray or ECG. The patient may, on the other hand, present with right heart failure, dyspnoea on exertion and, sometimes, mild cyanosis and clubbing. Clinically, the heart is large and inactive. There is often a murmur of tricuspid regurgitation and a cV wave in the neck. ECG shows small voltages, right atrial hypertrophy and right bundle branch block. A Wolff-Parkinson-White configuration (p. 516) is common. Chest X-ray shows a globular or flask-shaped heart with a prominent convex right atrial border. Echocardiogram and cardiac catheterization findings are characteristic. Cardiomegaly alone is not an indication for surgical treatment, as arrhythmia and heart failure respond well to conventional medical therapy.

ACYANOTIC CONGENITAL HEART DISEASE Conditions where blood is shunted from left to right may initially be without cyanosis but, if undetected, high flows through the pulmonary vascular bed will eventually lead to a reversal of the blood flow and cyanosis - the Eisenmenger reaction. Thus some acyanotic lesions may later produce cyanosis.

ATRIAL SEPTAL DEFECT ASDs can be of three types - secundum, primum or sinus venosus defect (Fig. 12.50). In adult cardiology they occur in the above order of decreasing frequency. The foramen ovale only acts as an ASD if right-sided pressures exceed left, e.g. in pulmonary embolism. ASD constitutes 10% of congenital heart disease.

Pathogenesis Blood flows from the left to the right atrium, and thence to the right ventricle and lungs. Because the lungs have a low impedance to flow, shunting is considerable. The patient has a limited exercise capacity, as maximal systemic cardiac output cannot be achieved on exercise. The right ventricle hypertrophies and the lung blood flow is increased, which seems to increase the likelihood of pulmonary infections. Symptoms are usually not prominent until middle age and include dyspnoea, fatigue (low exercise cardiac output) and, quite frequently, palpitations caused by atrial arrhythmias. Heart failure, when it occurs, is usually both right and left sided, as any rise in atrial pressure on one side of the septum will shunt the blood to the other side. After middle age the pulmonary vascular resistance may rise, with the development of massive pulmonary arteries (visible on chest X-ray) and, occasionally, reversal of the shunt.

Secundum a trial septal defect Clinical features Secundum ASD (Fig. 12.50A) is usually uncomplicated. The physical signs are the result of right ventricular hypertrophy and increased flow through the right side of the heart; this causes pulmonary and tricuspid flow murmurs and delayed closure of the pulmonary valve. The associated signs are wide, fixed splitting of the second sound and pulmonary systolic murmur; a tricuspid diastolic murmur is usually heard if shunting is significant (Fig. 12.51). Investigation and management Chest X-ray (Fig. 12.51) shows enlargement of the heart, prominent pulmonary arteries, pulmonary plethora and,

usually, a small aortic knuckle. ECG shows right bundle branch block in most cases, with right axis deviation. Echocardiography shows dilated RV with a 'volume load' of the RV. O The shunt can be demonstrated by contrast injection or with colour Doppler. Transoesophageal echocardiography can best delineate the types of ASD, and also the pulmonary venous drainage in relation to the ASD. An occasional finding associated with ostium secundum ASD is mitral stenosis (Lutembacher's syndrome). Cardiac catheterization can demonstrate the size of the shunt. Blood samples taken in the different cardiac chambers show a step-up in oxygen content of the blood occurring at the lower right atrium. The shunt size can be calculated from the oxygen content of the pulmonary artery blood and left-sided samples. Angiography to delin-

ED

FIG. 12.50 Sites of atrial septal defect A Ostium secundum. B Ostium primum (or atrioventricular canal defect). The septal leaflet of the mitral valve is often involved, allowing a communication between the left ventricle (LV) and both atria. C Sinus venosus defect, where the abnormality involves anomalous pulmonary venous drainage. SVC and IVC = superior and inferior vena cava. PV = pulmonary vein.

FIG. 12.51 Features of atrial septal defect A Chest X-ray showing enlarged pulmonary arteries, pulmonary plethora and relatively small aortic knuckle. B ECG showing a vertical axis (normal) and partial right bundle branch block, [c] Phonocardiogram demonstrating wide splitting of second heart sound. D M-mode echocardiogram at the ventricular level shows a volume-overload RV with paradoxical movement of the interventricular septum (IVS) which is out of phase with the posterior wall (PW) of the left ventricle.

539

eate ASD is no longer necessary because transoesophageal echocardiography and MRI are superior imaging techniques for this. Surgical closure is usually advised if the defect is associated with a shunt that gives a pulmonary flow greater than 1.5 times systemic flow even in adults, unless pulmonary vascular resistance and pressure have become so high as to cause a right-to-left shunt (Eisenmenger syndrome) and the patient turns cyanotic. In recent years a number of devices have been developed that allow the closure of ASD by percutaneous catheter-based occluders. For suitably placed and sized ASD these techniques may soon become the closure method of choice.

Primum atrial septal defect This is a much less common form, accounting for 10% of ASDs, but is relatively common in Down's syndrome. Clinical features, investigation and treatment The clinical features are similar to those of secundum ASD, but may be dominated by an associated mitral regurgitation, and the ECG usually shows a left axis deviation. A left ventricular angiogram shows the defect. Surgical correction is more complex than for a secundum defect, and may involve mitral valve repair or replacement.

Sinus venosus defect This is usually a high right atrial defect near the junction with the SVC. It is often associated with anomalous pulmonary venous drainage into the right atrium. O

VENTRICULAR SEPTAL DEFECT VSDs may occur as isolated lesions or be associated with other congenital defects of the heart. They are relatively common (Table 12.21) and have an equal sex incidence. About 50% close spontaneously during childhood. They may occur as muscular, membranous, infundibular or posterior (AV) defects and may be multiple. There is considerable shunting of blood from left to right unless a rise in pulmonary outflow resistance occurs owing to vascular disease caused by the high pulmonary flow. In extreme cases the flow reverses, causing cyanosis and the Eisenmenger syndrome. Small ventricular septal defects Small VSDs are common and are asymptomatic. There is

540

O MCQ 12.15

Fig. 12.21

Q MCQ 12.17

Case 12.4

MCQ 12.16

a loud systolic murmur with a thrill at the left sternal edge (maladie de Roger}, with no associated cardiomegaly and minimal, if any, changes of pulmonary plethora on chest X-ray. The ECG is usually normal. Differential diagnosis is from mitral regurgitation and is simply accomplished with Doppler ultrasound. The only complication of a small VSD is the risk of endocarditis, and antibiotic prophylaxis is essential. Although the surgical mortality and morbidity for closure of small defects are low, surgery is not usually necessary. Moderate and large ventricular septal defects With larger VSDs, systemic cardiac output is limited on exercise and can produce symptoms of fatigue and dyspnoea. As they are both volume loaded, both ventricles enlarge with displacement of the cardiac apex. The pansystolic murmur and thrill typical of maladie de Roger may be softer in these patients, as the jet velocity of the VSD and the pressure gradient between the ventricles becomes smaller the larger the defect. As pulmonary hypertension becomes prominent, the murmur may almost disappear as the pressures in the two ventricles balance. Investigation and management The features of VSDs on investigation are illustrated in Figure 12.52. ECG shows biventricular hypertrophy, and there is increasing right ventricular hypertrophy with pulmonary hypertension. Chest X-ray shows pulmonary plethora and cardiomegaly with both right and left ventricular hypertrophy. As pulmonary hypertension develops, the central pulmonary arteries increase in size and the peripheral vessels become pruned. 0 Echocardiography and Doppler ultrasound can show the location of the VSD, and can be used to calculate right ventricular pressure. Magnetic resonance scanning can demonstrate the defect and measure the shunt. Surgical closure is advised unless pulmonary artery pressure has risen too high. ©

COARCTATION OF THE AORTA Coarctation of the aorta is a narrowing, or even complete interruption, of the aorta distal to the aortic arch, with blood flow to the lower body augmented by flow carried through collateral vessels. Coarctation represents about 5% of adult congenital heart disease, and is seen in males three times as frequently as in females. The condition leads to upper body hypertension and an increased risk of aortic dissection. There is often a bicuspid aortic valve, which can become stenotic (or is congenitally stenosed). Other congenital defects are sometimes associated. There is also an associated risk of subarachnoid haemorrhage from berry aneurysm of the circle of Willis. These are presumed to be an associated congenital anomaly. The coarctation is nearly always at the aortic isthmus distal to the left subclavian artery, but can, rarely, occur

upper half being the left subclavian artery and the lower half the poststenotic dilatation of the aorta. Rib-notching is seen in most adults with coarctation, and is caused by the collateral arteries. Visualization of the coarctation is sometimes possible using 2D echocardiography, and the jet can often be located with Doppler ultrasound. The investigation of choice nowadays is MRI, which can also be used in serial scans to detect any sign of recurrence after surgery. Angiography of the aorta is not always needed, but cardiac catheterization is often performed to look for associated lesions and to confirm the length of the narrow segment.

12

Management and prognosis

FIG. 12.52 Features of ventricular septal defect A The characteristic systolic murmur associated with a widely split second heart sound with a premature aortic second sound (A2) and maintained mobility of P2 with respiration (dashed line on inspiration). B Doppler ultrasound showing highvelocity jet (6m/s). C and D 2D echocardiogram showing the septal defect (VSD).

Significant coarctation should always be corrected in children, either surgically or by balloon catheter. In adults, surgical correction is the definitive treatment, although it carries higher morbidity and mortality owing to the established hypertension, which often still requires treatment even after successful correction. The associated aortic valve lesions may progress and subarachnoid haemorrhage may occur. Endocarditis prophylaxis is necessary in uncorrected coarctation. OO

PERSISTENT DUCTUS ARTERIOSUS proximal to it. It may be due to ductal tissue in the wall of the aorta, leading to progressive stenosis of the aorta, as in ductal closure. The degree of stenosis becomes comparatively more severe as the child grows. The coarctation is followed by a poststenotic dilatation in most cases; however, if there is complete interruption of the aorta there may be some atresia of the descending aorta.

Clinical features Presenting complaints include left ventricular failure, subarachnoid haemorrhage and aortic dissection; however, most cases should be identified in childhood. The coarctation itself may become the site of an endocarditis, as may the bicuspid valve. The physical signs are hypertension, a large-volume pulse in the arms and carotids, and weak, delayed femoral pulses. Occasionally, the left subclavian is involved in the coarctation, producing weak left arm pulses. There is often palpable left ventricular hypertrophy, a systolic murmur over the precordium and in the back, and palpable collateral arteries running over the medial aspects of the scapulae.

Investigation Diagnosis is confirmed by Doppler ultrasound, demonstrating a lower systolic blood pressure in the ankles than in the brachial artery. The chest X-ray often shows the characteristic '3' sign over the aortic knuckle, with the

Persistent ductus arteriosus (PDA) is a congenital abnormality caused by failure of the ductus arteriosus to close and form the ligamentum arteriosum. The result is a shunt of blood from the aorta into the pulmonary circulation. The shunt is from the aorta just distal to the left subclavian artery, and so an increased flow occurs through the left heart and pulmonary circulation but not the right heart. The ductus arteriosus remains patent for a variable time after birth, and failure to close can be associated with congenital abnormalities other than PDA. The isolated lesion is usually diagnosed in childhood from the presence of a continuous or 'machinery murmur' at the upper left sternal edge.

Clinical features Overload of the left heart can lead to heart failure in older patients. The increased lung blood flow in infants can present with the consequences of increased pulmonary vascular resistance: extremely rarely, reversal of the flow through the ductus, with cyanotic blood passing to the distal aorta, occurs, producing clubbing of the toes but not the fingers. Rarely, the ductus may be the site of endocarditis.

Investigation and management Diagnosis is based on the physical findings of the characteristic murmur, a collapsing pulse and a chest X-ray which

541

usually shows a prominent pulmonary conus and increased lung blood flow. Echocardiography and Doppler ultrasound are extremely useful in demonstrating both the lesion and the volume load on the left ventricle. Cardiac catheterization with aortography is not usually necessary in uncomplicated cases, but it may be useful in differentiating the much less common aortopulmonary window, which requires much more complex surgery. Simple cases require ligation and transection of the PDA.

identification of anatomical lesions in the coronary arteries provides information about the feasibility of interventional procedures (bypass surgery or angioplasty), but it is the significance of the physiological effects of stenoses (e.g. angina, dyspnoea, arrhythmias, ECG changes or other investigational manifestations) which provides the indication whether the lesion(s) should be operated upon.

THE CORONARY CIRCULATION Flow in normal and abnormal coronary arteries

COR TRIATRIATUM In cor triatriatum the left atrium is divided by a diaphragm above the mitral valve, which gives rise to obstruction to pulmonary venous return to the left heart. There is a variable-size orifice (or none at all) which is easily opened surgically. The condition is thus eminently treatable.

DEXTROCARDIA Dextrocardia - in which there is a mirror image of the normal heart - is usually quite benign if associated with situs inversus (stomach bubble also on the right); however, it may be associated with other congenital abnormalities and can be confused with dextroversion, in which the heart is rotated rather than being a mirror image.

FURTHER READING ON ACYANOTIC CONGENITAL HEART DISEASE

Flow through the coronary arteries is regulated, with mean flow maintained through wide changes in mean arterial pressure (see Fig. 12.54). However, coronary vascular resistance is greatly affected by extrinsic compression during cardiac muscle contraction. Compression of the intramyocardial vessels during systole means that flow occurs in diastole (Fig. 12.53). This is particularly marked for the left coronary artery. Myocardial vessels are innervated and respond to neural, humoral and local metabolites. The heart has little capacity for anaerobic metabolism and cannot tolerate an oxygen debt. It is therefore dependent on increased perfusion to cope with an increased oxygen demand. Myocardial oxygen extraction is not increased with increasing demand, which must be met by increased coronary flow due to coronary vasodilatation. The coronary vascular resistance will fall to 20-25% of its basal state in response to maximal demand during exercise. This ability to lower the resistance to flow explains why it is possible to maintain normal flow at rest even when there is an 80% reduction in the diameter of a large coronary artery. The relationship between changes in perfusion pres-

Perloff J K 1995 Surgical closure of atrial septal defect in adults. N Engl J Med 333: 513-514

CORONARY HEART DISEASE (CHD) Coronary heart disease (CHD) refers to the consequences of coronary artery disease. Atheromatous disease outweighs all other causes, but coronary arteritis and embolism can cause ischaemia. CHD can still occur in the absence of demonstrable coronary angiographic stenosis, owing either to dynamic coronary artery spasm resulting in Prinzmetal angina, or to microvascular abnormalities too small to be detected by contrast angiography. Conceptually it is important to distinguish the anatomy of coronary artery disease from the physiology of the effects of the coronary stenoses on the myocardium. The QMCQ12.18

542

FIG. 12.53 Coronary arterial flow [A! Aortic pressure. IITI Flow in the left and right coronary arteries. Note the abrupt fall in left coronary flow with the onset of systole.

12 FIG. 12.55 Anatomy of the coronary circulation LAD = left anterior descending, LCx = left circumflex (dotted line represents position of LCx at the back of the heart), RCA = right coronary artery. Section across heart (A-A) shows the approximate distribution of blood supply.

FIG. 12.54 Effect of mean coronary arterial pressure on flow Coronary blood flow (blue line) is maintained constant despite variations in perfusion pressure. At perfusion levels below 60 mmHg coronary blood flow falls because the limit of coronary vasodilation is reached. A maximally vasodilated coronary artery has a linear pressure-flow relationship, as shown (light blue line). Stenoses of varying severity will alter the pressure-flow relationship of the dilated artery as shown by the grey and black straight lines. At rest an individual might have an artery flow at point A (a perfusion pressure approximately 100 mmHg); with exercise the demand rises (dashed line) and flow increases to point B, within the capacity of a normal vessel to dilate. If the individual's coronary artery had a moderate stenosis, only point C could have been reached, which is less than the flow required, so ischaemia (angina) is produced. A critical stenosis may interfere with the delivery of the necessary resting flow (D) so ischaemia at rest ensues.

sure and resistances of coronary arteries and flow is shown in Figure 12.54.

Coronary anatomy The anatomy of the coronary circulation is shown in Figure 12.55. Three major coronary arteries provide the heart with its blood supply. Major anatomical variations are rare. The main stems of the left coronary artery (LCA) and right coronary artery (RCA) arise from the aorta at the sinuses of Valsalva. The main stem of the LCA is a relatively short vessel. It divides into the left anterior descending artery (LAD) - which is responsible for the blood supply to the major part of the left ventricle and the interventricular septum - and the circumflex artery (Cx, usually the smaller branch of the left main stem), which supplies the posterior and variable amounts of the inferior walls of the left ventricle. The RCA supplies the right ventricle and inferior left ventricle. In some individuals variable amounts of the left ventricular apex, septum and posterior wall may be supplied by the RCA. This artery usually (60% of cases) provides the SA and AV nodes with their blood supply. The myocardium has a very rich capillary network, which appears to be regulated by precapillary resistance

vessels. In addition, there is the potential for a network of collateral vessels connecting coronary arterial networks to each other. When developed, these collaterals become important in coronary artery disease. O

ATHEROMA OF THE CORONARY ARTERIES The commonest cause of changes in the luminal diameter of the epicardial coronary arteries is the presence of atheromatous plaques (Fig. 12.56). The lesions may encircle an artery or be eccentric; they may be discrete and localized, or involve the greater length of the vessel. The eccentricity of some plaques may be important when vasospasm is the cause of angina; the normal arterial wall opposite the plaque maintains its ability to contract, producing a very marked reduction in cross-sectional area even if the smooth muscle contraction is modest in degree and the atheromatous plaque small. Acute changes in the plaques may account for unstable angina, myocardial infarction and sudden death. A fissure develops, breaching the thin endothelium that covers a lipid-rich plaque. The subsequent process is shown in Figure 12.56. These events appear to account for many of the more acute forms of coronary disease, and have renewed interest in the role of thrombosis in acute cardiac ischaemia. This has led to the successful use of fibrinolytic agents and aspirin in the treatment of infarction (p. 559). Under some circumstances atheromatous plaques may regress. This has been achieved by drugs that lower LDL cholesterol and triglyceride and increase HDL cholesterol.

Clinical features Symptoms The cardinal symptom is angina pectoris. This is a chest discomfort precipitated by exertion and relieved by rest. It is often described as a pressing or constricting feeling and, less often, as a pain. It remains remarkably constant within each individual, even though its precipitants may vary significantly from day to day. The patient may hold a clenched fist in front of the sternum to indicate the squeezing nature

543

A. Normal

B. Atheromatous plaque

C. Plaque dissection and acute expansion

FIG. 12.56 Progression of atheromatous lesion A and B Normal endothelium and structure of atheromatous plaque. C Acute expansion of the lesion occurs when the endothelial cells rupture, allowing haemorrhage into the plaque and platelet aggregation on the endothelial surface.

Crescendo angina and unstable angina Both crescendo and unstable angina may represent states of preinfarction. In crescendo angina a history of increasingly frequent attacks of angina with ever-diminishing levels of exertion is obtained. Unstable angina includes situations where episodes of pain are frequent, may occur without obvious cause and at rest, and where acute infarction has been excluded. Decubitus angina is angina occurring at rest in bed.

Signs There are few physical findings in uncomplicated CHD. An underlying cause, such as aortic stenosis, may give characteristic signs. There may be evidence of hypertension or hyperlipidaemia. Cardiac dilatation, hypertrophy and failure are all late features and are non-specific. Some individuals undoubtedly have significant ischaemic episodes without symptoms. This may be discovered by routine ECG, when pathological Q waves may be found with no history of any antecedent chest pains. Investigations of patients with ambulatory recorders have also shown some individuals to have changes in the ST segments usually associated with ischaemia, despite a total lack of symptoms. Even patients with known angina may have episodes of ST segment change which appear to be clinically silent. Patients with known coronary disease who have silent ischaemia fare worse than those without this finding.

Differential diagnosis of angina

Acute coronary syndromes The above is a description of typical stable atheromatous Fig. 12.22

544

coronary artery disease. Several other anginal syndromes occur which are common and account for many hospital admissions and much morbidity.

of the pain, and may describe radiating discomfort in the arm (left more often than right), the jaw or even a choking sensation in the throat. Dyspnoea frequently accompanies angina. There is usually rapid relief of the symptom with sublingual, or spray, glyceryl trinitrate. The pain of myocardial infarction may be similar, but generally begins at rest, does not respond so well to nitrates, lasts longer than 20 minutes, and is often associated with feelings of impending death, nausea, sweating and collapse. Unfortunately, in many patients with CHD the first manifestation of the disease may be sudden death. Prevention of atheroma is necessary to reduce this, although out-of-hospital resuscitation may improve the survival of the few fortunate enough to collapse within reach of a person competent in basic life support.

Vasospastic angina Some degree of arterial spasm is probably present in most episodes of angina, but spasm occurring in an otherwise normal coronary artery has also been observed. It can be severe enough to cause infarction, but is very rare. In Prinzmetal's syndrome rest pain is frequent and is associated with acute ST segment elevation, which resolves to normal with cessation of pain. This condition is rare and almost certainly involves coronary vasospasm.

Valve disease. Angina is a common symptom in aortic stenosis (p. 527) and aortic regurgitation (p. 529), particularly when the valve lesions are severe or acute. In cases of aortic regurgitation due to syphilis, chest pains may be caused by severe aortic reflux, but coronary ostial stenosis may also occur. In hypertrophic cardiomyopathy, hypertrophy may lead to angina. Aortic dissection and pulmonary embolism. These are discussed on pages 599 and 675. Pericarditis. Chest pain may be severe with acute pericarditis and is exacerbated by movement, deep respiration or coughing. It is often relieved by sitting forward and by non-steroidal anti-inflammatory drugs (NSAIDs). The clinical sign of pericardial friction rub may come and go. Oesophageal spasm. The pain of oesophageal spasm may be severe and, because it occurs at rest and can be associated with T-wave changes on the ECG, it may cause con-

siderable confusion and be diagnosed as unstable angina in some patients. Hiatus hernia with oesophagitis can cause chest pain. This can be demonstrated by infusions of mild acidic mixtures into the oesophagus of affected individuals. Radiological and endoscopic methods will demonstrate the abnormalities. CHD may coexist in the same individual. Acute cholecystitis and peptic ulceration can produce severe pains in the lower chest and xiphisternal areas, and may infrequently be mistaken for acute myocardial infarction. Non-specific T-wave changes on the ECG have been observed during these acute illnesses. Da Costa's syndrome (also known by several other names, e.g. neurocirculatory asthenia) has a long history and is characterized by sharp transient chest pains, which occur with relaxation, may disappear with exertion, and are associated with easy fatigue. Mitral valve prolapse. Atypical (for CHD) chest pains may occur in patients with mitral valve prolapse of mild degree. Hyperventilation. Patients who hyperventilate may produce symptoms indistinguishable from those of angina, and may have an abnormal resting and exercise ECG to accompany their symptoms. So-called syndrome X is characterized by typical angina in association with radiologically normal coronary arteries but a positive exercise test. It may be due to microvascular disease. Tietze's syndrome may cause diagnostic confusion until the painful costal cartilage is palpated.

Investigation The aim of investigations is to confirm that the clinical history is due to significant coronary artery disease, to determine at how much risk of infarction or progression the patient may be, and whether any remediable risk factors may be present. Once coronary artery disease has been confirmed using non-invasive tests, an assessment of ventricular function may be necessary and, finally, coronary angiography performed if surgery or angioplasty are being considered. (Investigation of myocardial infarction is considered on pp. 556-558.)

ECG The resting ECG may be normal, even when coronary disease is severe. Signs of ischaemia characteristically involve the ST segments, with depression in the leads corresponding to the myocardial territory involved (Fig. 12.57). The associated T waves may be flattened or biphasic. These ECG signs are not specific, however: they may be found in some normal individuals and may be produced by hyperventilation or a change in posture. Exercise test The exercise test has been described on page 480. Exercise will induce ischaemia with coronary stenoses of 70% or more (Fig. 12.57). If a population of asymptomatic indi-

12 FIG. 12.57 Exercise ECG showing typical and severe down-sloping ST segment depression

viduals is subjected to an exercise test there will be an incidence of false positive tests up to 20%. The predictive capacity of the exercise test depends on the prevalence of the disease in the population (Bayes' theorem). Therefore, exercise tests should be avoided in patients with a low pretest probability of having coronary heart disease, as there are likely to be more false than true positive results. The exercise ECG has several uses: • To detect CHD in patients with suspicious symptoms, especially those with an intermediate probability of disease; • To identify patients at high risk of further infarction or death early after myocardial infarction; • To assess functional capacity and adequacy of treatment. Although changes in the ST segment are the most characteristic features of a positive exercise test (Fig. 12.57), O the development of angina and the workload performed up to the development of symptoms or ECG change also aid interpretation. A normal blood pressure response consists of a rise in systolic pressure by at least 20mmHg. Diastolic pressure falls with exercise, but is so difficult to measure under these circumstances that it is not routinely taken. The gravest prognostic indicator in exercise testing is a failure of systolic pressure to rise, or even a fall. The appearance of multiple ventricular ectopic beats is also associated with severe disease. Radionuclide studies Radionuclide techniques (p. 485) are complementary to the other non-invasive methods used in assessing CHD. Information on ventricular wall motion abnormalities (MUGA scan) and regional perfusion (201T1 scan) can be obtained. Echocardiography and Doppler ultrasound Echocardiography (p. 483) may be useful to assess ventricular function and wall motion abnormalities, as well as excluding non-coronary causes of ischaemia, such as aortic stenosis or hypertrophic cardiomyopathy. Ambulatory ECG (Holter) monitoring Ambulatory ECG monitoring (p. 500) is generally used to detect arrhythmias, but may also detect episodes of ST segment change during normal activities. Some of these episodes may not be accompanied by symptoms of angina

545

(silent ischaemia). In the patient with known coronary disease, frequent episodes of silent ischaemia are associated with a high risk of infarction. In the patient with few or no symptoms the implications of this finding are not at all clear. The role of routine 24-hour ECG recording for ST segment analysis in CHD is not established, as it remains to be proven that the information gained is any more useful than that from the exercise test. Left ventricular and coronary angiography The aim of these investigations is to establish the functional state of the left ventricle, to exclude valve lesions that might account for the symptoms (such as occult aortic stenosis), and to obtain detailed pictures of the coronary arteries. O A global left ventricular ejection fraction (EF) of 50-70% is normal, and end-diastolic pressure is usually below 12mmHg. Subjective assessment is often sufficient to describe left ventricular wall motion abnormalities without recourse to the numerical estimation of EF. The coronary arteries are selectively intubated and cineangiograms taken in several projections. A 50% reduction in the diameter of an artery is generally considered to be a clinically significant stenosis. Cardiac catheterization provides unambiguous information on the location of coronary artery stenoses. However, the clinical significance of stenoses is sometimes difficult to judge; the exercise test and thallium scan may provide complementary information. Cardiac catheterization is indicated in all patients in whom coronary angioplasty or coronary bypass surgery is being contemplated. In some patients with troublesome symptoms and a low likelihood of CHD this examination is performed to finally exclude the possibility of coronary artery disease. This group may comprise between 10 and 20% of all patients undergoing cardiac catheterization. The invasive nature of catheterization is necessarily associated with a morbidity (<1.0%) and mortality (<0.1%). This means that angiography cannot assume the role of a screening test for coronary disease.

MANAGEMENT OF CHRONIC STABLE ANGINA Once the diagnosis is established, treatment should be aimed at restoring a normal exercise capacity. Investigations should establish the individual's risk, and contributory risk factors be reduced as far as possible. A flow diagram of the management of angina is shown in Figure 12.58.

General measures All patients should be advised to stop smoking, achieve O Fig. 12.10

546

FIG. 12.58 Flow diagram of management of angina PTCA = percutaneous transarterial coronary angioplasty.

their ideal weight and have raised blood pressure controlled. Lipid measurements should be made. Those with total cholesterol >5.2mmol/L, HDL/LDL <20%, triglyceride >2.0mmol/L should be given an appropriate diet (Ch. 5, p. 136) until repeat estimations are within the recommended ranges. First-degree relatives of young patients with CHD should have their lipid levels measured. The patients with angina of effort, a normal resting ECG and chest X-ray, and whose symptoms are well controlled on simple treatment can probably be managed adequately without further investigation. The prognosis in CHD is dependent on the severity of the stenoses, the number of coronary arteries involved and on left ventricular function. This information can only be obtained accurately at cardiac catheterization, which should be offered to patients below 40 years of age, whose risk is highest. The risk of death is approximately 4% per year in other patients, but increases to between 7 and 9% if there are associated abnormalities in the resting ECG. Thus, although a resting ECG may be sufficient investigation for many patients, further determination of the risk to the individual is obtained by non-invasive tests, especially the exercise test and myocardial perfusion scanning with Tl201 (p. 486).

12

TABLE 12.23 Drugs commonly used in the treatment of angina Drug

Route

Dose

Frequency

Nitrates Glyceryl trinitrate (GTN) Tablets Spray Buccal Patches

Sublinguai Oral Buccal Skin

0.3-0.6 mg 0.4mg/spray 1-1 Omg 5-10 mg

PRN and prophylaxis PRN and prophylaxis PRN and prophylaxis Daily (omit for 8-12 hours)

Isosorbide dinitrate (ISDN)

Oral (some sustained release)

10-80mg

4-8-hourly

Isosorbide mononitrate

Oral (some SR)

10-1 20 mg

6-24-hourly

Parenteral preparations GTN ISDN

i.v. i.v.

0.6-1 2 mg/h 1-12mg/h

Infusion Infusion

fi-Blockers. Many alternatives, including: Atenolol (fa selective) Bisoprolol Propranolol (non-selective)

Oral Oral Oral

50-200 mg 5-1 Omg 30-360 mg/day

Daily Daily 6-hourly

Calcium antagonists Nifedipine Long-acting Diltiazem Modified release Verapamil Slow release

Oral Oral Oral Oral Oral Oral

10-20mg 30-90 mg 60-120 mg 90-240mg 40-120 mg 90-240 mg

8-hourly (12-24-hourly for sustained release) Daily 8-hourly (12-24-hourly for sustained release) 12 hourly-daily 8-hourly (12-24-hourly for sustained release form) Daily

New agents Nicorandil (ATP-sensitive K+ channel opener) Trimetazidine

Oral Oral

10-20mg Not available in UK

12-hourly

Catheterization may then be required for those with a positive test at low workloads, those with large reversible Tl201 perfusion scan defects, and those in whom medical treatment is not being successful.

Drug treatment The drugs commonly used in the treatment of angina are listed in Table 12.23.

Aspirin Although not a symptomatic treatment, all patients should be given 75-150 mg of aspirin per day as an antithrombotic agent. Nitrates Therapy begins with sublingual tablets or aerosol spray nitrates. These are used to treat an anginal episode and can be successfully used in prophylaxis by patients with predictable symptoms. The aerosol spray has the advantage of a long storage life compared to glyceryl trinitrate (GTN) tablets, which deteriorate rapidly in their container once

opened. The onset of action is very rapid (within 1-2 min) and its duration can be up to 20 minutes. With all nitrates troublesome headache can prevent their use in certain susceptible individuals, although this side-effect usually stops after a few days of treatment. Longer-acting nitrate preparations include glyceryl trinitrate given via the skin, using sustained-release plasters or patches. Isosorbide dinitrate is absorbed from the gut and metabolized by the liver to the active mononitrate. The duration of action is longer than short-acting nitrates, but 4-8-hourly doses are needed to maintain blood levels. Mononitrates are absorbed in the gut and do not require hepatic conversion; thus, some variability in effectiveness due to liver metabolism is avoided. Their duration of action is longer, so one or two daily doses are adequate. An idiosyncratic reaction to GTN (and also other nitrates) is syncopal attacks. Clinically, this has to be distinguished from syncope secondary to angina, which is suggestive of an arrhythmia, whereas nitrate-induced syncope is associated with bradycardia and hypotension (a vasovagal type of reaction).

547

RECENT ADVANCES IN ISCHAEMIC PRECONDITIONING There is good experimental evidence that heart muscle is able to adapt to conditions of ischaemia, so that short bursts of ischaemia (in the order of 2-5 minutes, singly or repeated over a short period) will render the myocardium resistant to a much longer period of lack of blood supply. In laboratory conditions 30-60 minutes of coronary occlusion will produce infarction of 40-60% of the muscle supplied by that coronary artery. Merely occluding the artery temporarily, for 2-3 minutes once or twice before the long occlusion, can reduce infarct size to just 10-20% of the muscle at risk. This phenomenon is called preconditioning, and is associated not only with reduced infarct size, but also with a dramatic reduction in reperfusion arrhythmia. These observations are being extended to human heart muscle suffering global ischaemia, as in cardiac surgery, or regional ischaemia, as in angina and PTCA. The mechanisms underlying preconditioning are being sought, but there is good evidence to suggest adenosine receptors and opening of ATP-sensitive potassium channels may be involved. This opens up the prospects of using drugs to mimic ischaemia-induced preconditioning.

Tolerance Tolerance to nitrates develops as sulphydryl (SH) groups on the vessel walls become oxidized by constant exposure; this prevents the production of nitric oxide and stimulation of guanilate cyclase, which is believed to be fundamental to the smooth muscle relaxation produced by these drugs. A nitrate-free period of 6 hours in every 24 is sufficient to allow recovery of SH groups when longer-acting preparations are used. The tendency for patients to develop a tolerance to longacting nitrates (tachyphylaxis) limits their usefulness. No controlled trials have demonstrated an improvement in mortality with stable angina or after myocardial infarction. Their primary use therefore remains as an effective symptomatic treatment of coronary heart disease.

548

-Blocking drugs |3-Adrenoreceptor blockade has become established as a cornerstone in the treatment of angina. Although many -blocking drugs with varying pharmacological properties are available, there is no overwhelming evidence to suggest that one is any better than the others. Some patients find they are unpleasant owing to a 'sedative' action and an interference with exercise capacity. They are specifically contraindicated in patients with asthma, and relatively contraindicated in diabetes and in those with peripheral vascular disease, bradycardia or heart block. A number of patients may have been mislabelled as asthmatics in the past. A safe way of testing whether -blockade may be used in these patients is by infusing esmolol (50-

200 g/kg/min), because it has the shortest half-life and any detrimental effects may be reversed relatively rapidly. A cardioselective -blocker (selective for blockade of 1 receptors > p2) may then be used as maintenance treatment, e.g. bisoprolol (2.5-5mg o.d.). In angina, these drugs achieve their benefit by reducing myocardial oxygen demand consequent on a fall in heart rate, blood pressure and myocardial contractility. There is good evidence that -blockers improve prognosis after myocardial infarction. However, no prospective trial has extended this observation to the treatment of stable angina. Calcium antagonists Calcium antagonists are effective in angina, achieving their effect by smooth muscle relaxation in the coronary arteries and peripheral circulation, increasing myocardial supply and reducing myocardial work. The three types of calcium antagonist available for clinical use are the dihydropyridines (e.g. nifedipine, nicardipine, amlodipine, felodipine), diltiazem and verapamil. They vary in their pharmacological properties, and so their indications may be different. The mode of action varies with the type of drug, all of which inhibit cellular calcium entry but differ in their tissue specificity. The dihydropyridines act primarily by arteriolar vasodilatation with minimal effects on the SA node and the myocardium. They may cause a reflex tachycardia. The action of diltiazem is more marked on the SA and AV nodes, producing a mild bradycardia, but less negative inotropy than verapamil. It is advisable to avoid using a dihydropyridine as monotherapy in angina. These drugs are best used in combination with a |3-blocker. If a |3-blocker is contraindicated, or when monotherapy with a calcium antagonist is desirable, then the use of verapamil (40-120mg t.d.s., SR 240mg o.d. or b.d.) or diltiazem (SR 90-120 mg b.d.) is recommended. The negative inotropic and chronotropic effects of verapamil are useful in antianginal treatment in those without heart failure or heart block. In heart failure the least negatively inotropic calcium antagonists are to be used, such as nicardipine, felodipine or lacidipine. Calcium antagonists are a useful symptomatic treatment for angina, with added beneficial blood-pressure-lowering action. As with nitrates, there are no prospective trial data that suggest any effect of these drugs on prognosis. There are some data to suggest that patients with good LV function may do better on diltiazem or verapamil after transmural and non-Q-wave myocardial infarction.

Nicorandil Nicorandil is now accepted as another class of antianginal agent. It is the first potassium channel opener (KCO) that has been shown to be effective in the treatment of angina. It can be used in preference to some of the above drugs in coronary artery spasm, angina in asthmatics and diabetics, and in patients with heart failure or peripheral vascular disease. It is an arterial and venous vasodilator which combines the arteriolar vasodilation of ATP-sensitive

potassium channel opening with a nitrate-like venodilatory capacity. A major prospective trial on the drug in stable angina will report in 2001/2002. Other compounds Amiodarone was introduced as an antianginal preparation but its use is now limited to the treatment of serious arrhythmia. Perhexilene alters the metabolism of cardiac muscle from FFA to glucose, and can be effective in refractory angina. In certain identifiable genotypes, however, it can have serious side-effects, including hepatotoxicity and peripheral neuropathy. Trimetazidine is marketed in some parts of Europe. It has a poorly characterized mode of action. ACE inhibitors Although there has been a demonstrable mortality benefit when these drugs are used in patients following MI, particularly when LV function is impaired, there are no convincing data on their use in stable angina. Drug combinations in angina There is a progression in therapy before medical treatment can really be said to have failed. Most patients begin treatment with nitrates or KCO and a (3-blocker. A calcium antagonist is used if -blockers are contraindicated. Elderly patients may well be adequately controlled with nitrates and a reduction in activity, and many find that -blockers induce unpleasant lethargy. Younger patients may not wish to reduce their activity, and find the prophylactic action of -blockers or calcium antagonists a considerable advantage. Verapamil and diltiazem produce profound bradycardia or heart block when used with a (3blocker, so these combinations should be used cautiously. Patients with severe disease commonly require 'multiple therapy' to control their symptoms, but the benefits of these three- or four-drug combinations are dubious. Practical prescribing in angina Provided there is no contraindication to the use of the drugs in individual patients, the schema of drug therapy shown in Figure 12.59 may be adopted: • It is generally wise to start antianginal medication by selecting the shortest-acting drugs and small doses several times a day, and build up the doses gradually if clinically indicated and in the absence of adverse reactions. In this way any unexpected adverse reaction will not be as prolonged or as severe as if larger doses and longer-acting versions had been used. Drug compliance is generally not problematic (as it can be in the treatment of hypertension). • The exception to the above rule occurs when treating coronary arterial spasm and arrhythmia secondary to myocardial ischaemia. In these conditions it is imperative to institute full therapeutic cover for the entire 24 hours using known effective doses, which can subsequently be reduced if side-effects supervene. • In vasospastic or Prinzmetal's angina the aim of treat-

ment is to prevent the powerful vasoconstriction: combination therapy using long-acting isosorbide mononitrate, dihydropyridine calcium antagonists and nicorandil may be required. p-Blockers may have to be added to counteract the reflex tachycardia and to reduce the intensity of angina by reducing oxygen demand during an attack. • In patients with other conditions concomitant with angina, drugs that precipitate angina (e.g. theophylline, digoxin, thyroxine) may have to be reduced or stopped, and an alternative therapy sought. In angina patients with heart failure and low blood pressure, or with hypertension, it may be necessary to replace their prescribed vasodilators (e.g. ACE inhibitors, a-blockers, hydralazine) with those possessing antianginal properties.

12

Coronary artery bypass graft surgery (CABG) Effective surgical treatment has had a profound effect on the management of CHD in the developed world. Coronary artery bypass grafting can now be performed with an overall operative mortality of under 1 %. Symptoms can be eradicated in 80-90% of patients and improved considerably in a further 5-10%; a few are no better or, indeed, are worse following surgery. It might therefore be asked why this form of therapy, which aims to 'revascularize' the heart, is not offered to all patients with angina. Several factors are responsible, including: • Finite morbidity from surgery • Deterioration of grafts • Survival advantage for surgical treatment has been proven for only moderate- and high-risk subgroups of patients • Benefit in symptoms has to outweigh the cost in wellbeing for up to 12 weeks postoperatively. Deterioration of grafts Saphenous vein bypass grafts have an appreciable rate of deterioration: more than 90% are patent immediately after surgery, but then occlude at a rate of approximately 2% of grafts per year between 5 and 10 years, and at 4% per year thereafter. About 60% of grafts are patent after 10 years. This implies that surgery is not a permanent cure, merely another method of symptom control in these patients. Improvements in early vein-patency rates have been achieved by the use of antithrombotic agents, such as aspirin. Even more significant has been the use of the internal mammary arteries (IMA) as grafts. This is more demanding technically, but has been shown to provide greater than 90% patency rates at 10 years. Unfortunately, complete cardiac revascularization is not often possible even when both IMA are used. Survival advantage Survival is improved by surgery in those with left main stem disease and those with three-vessel coronary disease and impaired left ventricular function (Table 12.24). It is

549

FIG. 12.59 Schema for medical antianginal therapy

very likely to improve survival in patients with severe angina, three-vessel coronary disease and normal ventricular function. The situation is less clear for patients with less marked symptoms and those with little in the way of symptoms but with a positive exercise test, although evidence is mounting that this group of patients also benefit from surgery. Surgery appears to be a more effective treatment for patients with continued symptoms not controlled by drugs. There is a diminution in fatal heart attacks, but probably not in non-fatal events. Randomized studies have failed to show any differences between medical and surgical treat-

O MCQ12.19

550

0 Fig. 12.23

Case 12.5

TABLE 12.24 Indications for coronary artery surgery in angina Indication

Comment

Left main stem stenosis Symptoms uncontrolled despite maximal medical treatment Impaired LV function especially with three-vessel coronary artery disease Proximal stenosis in left anterior descending coronary artery with positive exercise test at low workload

Proven survival benefit of surgery Improved survival with surgery in some subgroups of patients Improved survival with surgery Likely improved survival with surgery

ment in the numbers of patients who return to work, or in the amount of recreational activity they take up. However, the sociological measures may not reflect the changes in perceived quality of life in the two groups of patients.

TABLE 12.25 Indications for coronary angioplasty Indication

Comment

All indications for surgery (except left main stem stenosis) where coronary lesions suitable* Symptomatic single-vessel coronary disease

Single- and double-vessel PTCA. Low risk, high efficacy

Coronary disease where surgery impossible ('salvage PTCA') In AMI, especially if thrombolysis contraindicated

PTCA is as good and safe as surgery but patients face a higher incidence of recurrence A suitable approach in the very elderly or infirm Survival advantage over thrombolysis

* Suitable coronary lesions: proximal, discrete, not involving major branches, recent onset, non-calcified, circumferential lesions.

Thus, in general, patients are considered for coronary artery surgery if medical therapy fails to control their symptoms adequately, or if they fall into the above higherrisk groups. These broad guidelines conceal a large number of patients in a rather grey area, in whom individual assessments must be made. These might include young patients with severe stenoses causing highly positive exercise tests, in whom the consequences of occlusion might be a disabling infarction and in whom an internal mammary graft might be expected to be a highly successful treatment. Patients with extensive left ventricular aneurysm may often improve, in terms of heart failure control and anginal symptoms, with plication or resection of the aneurysmal area.

Changes in surgical technique A number of new, less invasive methods of coronary revascularization are currently in use. They share the aim of avoiding the requirement for cardiopulmonary bypass and its attendant complexity and morbidity. Their application is more limited than that of conventional CABG and the long-term advantages over CABG, or indeed angioplasty, have not yet been assessed by trial. O

Percutaneous transarterial coronary angioplasty (PTCA) (Table 12.25) © The indications for percutaneous transarterial coronary angioplasty (PTCA), first introduced in 1977, are similar to those that select for surgery. In most centres one or two stenoses are tackled by angioplasty, and patients with severe three-vessel coronary disease are generally considered for CABG. Patients with left main stem disease are also not considered for PTCA, unless surgery is impossible for technical reasons or because of coincidental disease. Intracoronary stents are now routinely deployed following balloon angioplasty. Debate continues regarding their optimal use, but the great improvement in immediate angiographic appearances, the virtual eradication of the

need for urgent CABG in emergencies, and the rapid reduction in the cost of stents has fuelled their adoption in advance of randomized trials.

IB

Restenosis Initial experience with PTCA revealed a >30% angiographic restenosis rate, with a need for further intervention in up to 25% of patients. The routine use of stents has halved these incidences, and in selected cases symptomatic restenosis rates may be less than 10% at 1 year. Patients with small arteries (<2.5mm diameter) and those with very large arteries (>4.0mm diameter) are best managed without stents, but they form less than 10% of the PTCA population. The determinants' recurrence are unknown, but most patients are given antithrombotic treatment with aspirin and newer antithrombotic agents, clopidogrel or ticlopidine. Some patients receive calcium antagonists and/or nitrates as treatment for vasospasm. Immediate complications such as vessel occlusion, myocardial infarction or the requirement for urgent surgery have diminished dramatically with the development and increased use of coronary stents. Overall, complications for elective PTCA now should approximate 1 %. O

Newer techniques Interventional cardiology is a rapidly evolving field. In

RECENT ADVANCES - INHIBITORS OF PLATELET AGGREG/VTORY RECEPTORS Despite the availability of aspirin, heparin, warfarin and fibrinolytic agents (streptokinase, tPA, urokinase), the treatment of certain conditions such as unstable angina and the result of angioplasty of acute ischaemic syndromes are still unsatisfactory because of unpredictable tendencies to thrombus formation and acute occlusion. Indeed, it has been suggested that intracoronary infusion of fibrinolytic agents can, paradoxically, hasten thrombus formation. There is therefore a need to find more effective agents that will prevent intravascular thrombus formation without causing significant bleeding complications. Platelet IIB/IIIA receptors are necessary to form a meshwork of interlinked platelets, which is an essential process to induce new thrombus formation. Inhibitors of this receptor prevent this and have shown promising results in the treatment of unstable angina and in reducing the incidence of acute occlusions peri- and postangioplasty. Three types of such inhibitors are being tested: monoclonal antibody against the receptor, peptide inhibitors, and chemical non-antibody and nonpeptide inhibitors. The relative efficacies of these agents in the treatment of diverse manifestations of the acute ischaemic syndrome have yet to be elucidated.

551

farction. Symptomatic treatment alone is associated with a mortality of about 12-15%, with a further 12% of patients suffering full transmural myocardial infarction within 12 months. An enormous amount of effort has gone into establishing the most appropriate protocols to manage these patients, particularly as many of the options are technically sophisticated or demand the use of very expensive new drugs.

FIG. 12.60 Technique of coronary angioplasty A The guide catheter is introduced into the coronary ostium. B A steerable guide wire is advanced across the lesion. C The balloon catheter is advanced from the guiding catheter over the guide wire and positioned so that the balloon straddles the narrowing. D Balloon inflation splits and compresses the atheromatous plaque. E After deflation, the balloon is removed and the arterial wall shows irregularity at the site of the narrowing. F A balloon-mounted stent is deployed G The stent remains in situ to be re-endothelialized within weeks.

order to extend the principles of PTCA to more demanding lesions, various devices have been developed. These include laser-assisted PTCA (mostly for occlusions on calcified lesions), atherectomy devices, devices to remove thrombos, and catheters capable of delivering drugs or radiation therapy directly to the lesion. Their precise roles continue to be defined.

ACUTE CORONARY SYNDROMES(ACS)

552

These syndromes, which include the clinical diagnoses of unstable angina, non-Q-wave myocardial infarction and rest angina, share much of the pathophysiology of acute myocardial infarction. Rupture of the fibrous cap over an atheromatous plaque, intracoronary thrombus formation and myocardial ischaemia are common to all. Numerically, the acute coronary syndromes now account for more hospital admissions than acute myocardial in-

Management The diagnosis relies on the clinical history of crescendo angina (p. 544) or rest pain, which is not associated with ST segment elevation on the ECG or with significant cardiac enzyme elevation (CK or CKMB). The patient should be admitted into a coronary care unit where bed rest is imposed. Light sedation can be helpful. Medical therapy includes oxygen, aspirin (150-200mg orally) and heparin with i.v. nitrate. Close monitoring of blood pressure is required during nitrate infusion, and the systolic pressure should not be reduced by more than 20%. p-Blockers and calcium antagonists may be added when symptoms persist despite increased doses of i.v. nitrate. Use of thrombolytics in unstable angina does not improve the prognosis and may even worsen outcome. Opiates should be used if pain is persistent. Heparin and aspirin are well established in the treatment of ACS. Low molecular weight heparin has the advantage of ease of use (subcutaneous injection rather than i.v. infusion), more predictable effects on clotting parameters and probably better outcomes in ACS than with conventional i.v. unfractionated heparin. It is, however, more expensive, although increased drug cost may be mitigated by savings in overall patient care (no i.v. pump, fewer laboratory tests, etc.). Recently, more powerful antiplatelet agents, including the oral agents ticlodipine and clopidogrel and the parenteral platelet glycoprotein Ilb/IIIa receptor inhibitors, have been used in ACS. The former can be used when there is aspirin intolerance or allergy. Several trials of IIb/IIIa receptor antagonists have shown that these parenterally given drugs can reduce the combined complications of death and myocardial infarction. They are very expensive and will initially be used in the highest-risk subgroup of patients, as well as those undergoing urgent angioplasty. The timing of cardiac catheterization and angioplasty in ACS is hotly debated. In the USA and some countries in Europe early investigation and PTCA is the norm. In countries such as the UK, with less well-developed cardiac intervention facilities, fewer patients are investigated, and in those that are, the tests are performed later, sometimes electively in the post-discharge recovery period. Long-term outcomes appear comparable whichever strategy is used.

Risk stratification in ACS (Table 12.26) Straightforward clinical features can be used to identify patients at high risk of death or myocardial infarction from

12

CASE STUDY 12.3 INCREASING CHEST PAIN A previously well man of 68 years came to the emergency department with a history of increasingly severe chest pains. The pains were central and radiated to the angle of the jaw. Although he had initially noticed milder pains precipitated by activity and emotion, for the previous 2 nights he had been woken from sleep with increasingly severe episodes of the same discomfort. The last bout of pain had been about 8 hours previously and had lasted altogether for 40-45 minutes. It was not associated with sweating, nausea or other sensations. The pain had woken him from sleep. Initially he thought it was indigestion, but its persistence, and lack of response to antacids, led him to seek advice. He had no significant past medical history but was having treatment with a statin for a hyperlipidaemia detected by a routine screen by his GP. He was an ex-smoker, having given up 6 months prior to presentation, and he had a strong positive family history of ischaemic heart disease with three first-degree relatives suffering myocardial infarction before the age of 60. In the emergency department his initial ECG was normal, but half an hour later he began to suffer chest pains and a further ECG revealed flattening of the T wave in lead AVL and 1.5mm horizontal ST segment depression in leads V4, V5 and V6. He was given sublingual glyceryl trinitrate (GTN), the pain resolved and ECG returned to normal after 10 minutes.

Question How might the diagnosis of an acute coronary syndrome be confirmed and how should the patient be subsequently managed?

Discussion Acute coronary syndromes rely upon a classical history related to myocardial ischaemia in the absence of transmural infarction. Infarction should be excluded by the absence of evidence from the ECG (ST elevation and subsequent development of Q waves and T-wave changes). In addition, to diagnose acute myocardial infarction a 2-3 fold elevation in creatine kinase enzyme and its isoform CK-MB should be demonstrated. The characteristics of the chest pain in infarction also suggest the diagnosis in that the pains tend to be more prolonged than in an acute coronary syndrome, not responsive to GTN and often associated with other symptoms such as nausea, sweating and breathlessness. Acute coronary syndrome is often therefore a diagnosis of exclusion. In this case the association of ST segment depression with pain is a highly important prognostic factor. Additional important prognostic features include the presence of an elevated myocardial specific enzyme in the blood, either troponin T or troponin I. Elevations of these enzymes during acute coronary syndrome presentation is associated with a much higher incidence of

ACS. Continuation of chest pain despite adequate treatment, the presence of ST depression with chest pain, or the development of T-wave changes identify a high-risk group. An early indicator is the detection of elevations in the cardiac-specific troponins T or I. These patients will generally benefit from cardiac catheterization with a view to revascularization, usually

complications. Indeed patients with angina-type pains in whom the troponins are negative 6-8 hours after the symptom have a very low risk of subsequent events and can usually be discharged safely from hospital for further outpatient evaluation. This patient is in a high-risk subgroup which requires stabilization. Mainstays of treatment include intravenous nitrate therapy plus heparinization using low molecular weight heparin. Beta-blockers are used and help to control the pain. The smooth-muscle potassium channel activator nicorandil causes arterial dilatation and may also improve the immediate situation. There is discussion as to whether all patients with acute coronary syndromes in the higher risk groups should automatically be offered invasive investigation with cardiac catheterization and angiography. This is the most common approach in Europe, the USA and Australia. Cardiac catheterization and angiography are essential if patients have further episodes of pain with or without ECG change despite pharmacological therapy. A case can be made for delaying angiography in those patients whose symptoms disappear on the initial treatment. If patients require cardiac catheterization, additional treatment with drug inhibitors of platelet aggregation (by binding to the Ilb/IIIa cell surface receptor), such as clopidogrel or ticlopidine, further improves their prognosis and decreases the risks of acute . angiography and angioplasty treatment.

by PTCA. Angioplasty and stent treatment is further improved, in terms of both immediate and long-term outcome, if these patients are given bolus plus infusions of Ilb/IIIa receptor inhibitors immediately before or during the coronary intervention. Those patients - often the majority - without ECG changes on troponin elevation (Fig. 12.61) may safely be

553

RECENT ADVANCES IN LASER

TRANSMYOCARDIAL REVASCULARIZATION

FIG. 12.61 Cardiac specific troponin I levels to predict risk of mortality in patients with acute coronary syndromes

TABLE 12.26 Risk stratification in ACS, The presence of just one feature may be sufficient to classify a patient who is clinically at high risk of death or Ml from ACS. Feature

Low risk

High risk

ST depression 3=1 mm with pain 8 hr troponin T or troponin I

Absent

Present

Negative

Failure to control episodes of chest pain T-wave changes Haemodynamic disturbance with pain Conventional risk attributes in acute Ml/ischaemia, e.g. old age, poor LV Previous Ml

Absent

Positive - the higher the enzyme, the higher the risk Present

Absent Absent

Present Present

Absent

Present

discharged early for conventional outpatient investigation by exercise testing or Tl201 scanning, and management similar to that for those with stable angina. The diagnostic work-up should include identification of other conditions that may exacerbate angina, such as anaemia, hypertension, thyrotoxicosis, severe anxiety and cocaine abuse.

FURTHER READING ON CORONARY HEART DISEASE 554

Braunwald E 1998 Unstable angina: an etiologic approach to management. Circulation 98: 2219-2222.

One of the prerequisites of coronary artery bypass surgery or angioplasty is that the coronary arterial stenoses should occur proximally in epicardial coronary arteries. If the stenoses occur distally, especially if the distal intramyocardial vessels are affected, neither of these two procedures will produce much benefit. Certain medical conditions, such as diabetes mellitus, predispose to such distal disease. Hitherto, the only option has been cardiac transplantation, for either severe cardiac failure or intractable angina. Recently, laser transmyocardial revascularization has been developed as a potential alternative. During open chest surgery, thin rays of powerful laser beams are directed from the epicardial surface towards the left ventricular chamber in a myocardial region that has been shown to be viable, and either ischaemic or hibernating. Through a process which has not yet been elucidated, the epicardial portion of these artificial channels spontaneously seals up, leaving intramyocardial channels communicating with the ventricular chamber, and effectively forming new coronary vessels. Preliminary studies suggest that perfusion and oxygenation of these myocardial regions can be restored towards normal, and that symptoms and functional status can be significantly improved in selected patients. The effects on ventricular function need investigation. Clinical studies are currently in progress, and the applicability of this technique in practice will depend on the results.

Shepherd J, Cobbe S M, Ford I et al. 1995 Prevention of coronary heart disease with pravastatin in men with hypercholesterolaemia. West of Scotland Coronary Prevention Study Group. N Engl J Med 333:1301-1307. The Bypass Angioplasty Revascularisation Investigation (BARI) Investigators. 1996 Comparison of coronary bypass surgery with angioplasty in patients with multi-vessel disease. N Engl J Med 335: 217-225. Yusuf S, Zucker D, Peduzzi P et al. 1994 Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomised trials by the Coronary Artery Bypass Graft Surgery Trialists Collaboration. Lancet 344: 563-510.

MYOCARDIAL INFARCTION In the UK about 200000 people suffer a myocardial infarction each year. Half die from the infarction, and, of these, half do so before they reach hospital (Fig. 12.62). Most of this group suffer an arrhythmia, and death occurs instantaneously or within 1 hour of the first symptom. Clearly, these deaths will only be avoided by measures that prevent the development of coronary disease, although

TABLE 12.27 Mortality from acute myocardial infarction Class

Definition

1

Uncomplicated, without clinical signs of LV failure Generally asymptomatic, but manifesting a ventricular third sound, basal crackles or raised venous pressure Clinical evidence of pulmonary oedema Cardiogenic shock indicated by a systolic BP below 90 mmHg and at least one of: oliguria (<20mL/h), low skin temperature, or mental confusion

II

III IV FIG. 12.62 Mortality following onset of acute myocardial infarction A high proportion of patients die of ventricular fibrillation within 1 hour of onset.

prehospital and community resuscitation schemes can make an impact. Hospital admissions for infarction total 70-80000 per year in the UK, with an overall in-hospital mortality around 10%, plus a 10% mortality in the first year. The acute mortality varies with the type of infarct, from 3-6% for the patient with an uncomplicated infarction injuring only a limited amount of myocardium, to 50-80% with massive myocardial necrosis and shock (Table 12.27). The amount of muscle damaged determines both early and late outcomes from infarction. The advent of coronary care units (CCU) has led to a reduction in acute mortality, which has fallen from over 20% in 1966-67 to under 10% in the 1980s.

Admissions (%)

Mortality

33

6

38

17

10

38

19

81

12

(%)

Data from: Killip and Kimball 1969 American Journal of Cardiology 20: 457-467.

buccal nitrates. Associated symptoms are sweating, nausea, vomiting, breathlessness and collapse. Like angina, the pain may radiate to the arms, throat and jaw. Many patients believe they are suffering from indigestion and consume antacids. With the advent of thrombolytic therapy, it has become even more important to be aware of possible differential diagnoses in the patient with acute chest pain (p. 544).

Examination DIAGNOSIS OF MYOCARDIAL INFARCTION The clinical syndrome is less precise than the pathological definition, but a working diagnosis includes two of the following: • A typical history • The evolution of characteristic ECG change ending with Q-wave formation • A significant elevation and subsequent fall in creatine phosphokinase, preferably the MB isoenzyme fraction, aspartate transaminase, hydroxybutyrate dehydrogenase, or cardiac troponin-T.

CLINICAL FEATURES History A squeezing or tight, chest or epigastric pain, with its onset at rest, is reported by 80% of patients, a large proportion (40%) of whom may have an antecedent history of angina. The pain lasts longer than angina - typically more than 20 minutes - and is resistant to treatment with sublingual or

There are no specific features that accompany myocardial infarction. During the acute phase patients vary a great deal in their appearance, and the physical signs are largely those associated with pain and fear and are initially dominated by autonomic nervous activity. With acute anterior infarction there is a tendency for sympathetic activity to predominate, with a tachycardia, cool pale periphery and a normal, or even slightly raised, blood pressure in the early minutes. This contrasts with acute inferior infarction, which is commonly associated with massive vagal discharge, producing a cold sweaty periphery, bradycardia, hypotension, nausea and vomiting. The blood pressure tends to fall to lower than normal for the patient within a few hours of infarction, but some may have a hypertensive reaction. The venous pressure is generally normal in the early stages unless there has been extensive right ventricular involvement, which may occur with an inferior or posterior infarction. In left ventricular infarction, heart failure may develop and the pulmonary venous pressure will rise, signifying severe myocardial damage. Auscultation commonly reveals a fourth heart sound and a third sound over the apex in those patients in incipient left ventricular failure. The left ventricular impulse may be anterior and dyskinetic when a large

555

segment of left ventricular wall is moving paradoxically; this is particularly marked later in those individuals who develop an aneurysm. Orthopnoea and basal crackles are present in patients developing pulmonary oedema. A pyrexia develops within 12 hours in patients who suffer significant myocardial injury; it may persist for up to a week and is associated with a polymorphonuclear leukocytosis. Pericardial friction rubs are occasionally encountered soon after infarction, and signify a large loss of myocardium extending to the epicardium. This finding is distinct from the pericarditis and pneumonitis of Dressler's syndrome, which is a rare complication occurring during the first 6 weeks after a myocardial infarction and which has an immunological basis (p. 566). Other physical signs in myocardial infarction are due to the mechanical complications of the condition and are discussed individually below.

INVESTIGATION ECG changes in acute infarction The ECG changes of acute infarction may take some hours to develop, and occasionally are completely absent. Treatment of acute infarction must often begin before there is any objective evidence confirming the diagnosis (a rise in plasma enzyme levels also takes some hours to occur). 57 segment elevation The most characteristic feature of infarction on the ECG is the development of ST segment elevation, with a convex upward pattern in the leads facing the area of ventricle infarcted. The diastolic isoelectric line is depressed, so accentuating the ST elevation that follows it (Fig. 12.63). Reciprocal ST depression is often seen in leads opposite the infarction; this is usually indicative of severe multiplevessel coronary disease, but may on occasion represent a purely electrical phenomenon. ST elevation (the current of injury] is produced by ischaemia, and can occur in the absence of infarcted tissue. Thus, such ST changes can be associated with anginal pain due to coronary vasopasm (Prinzmetal's angina), and resolve rapidly and completely when spasm is relieved and ischaemia reversed. ST elevation during acute infarction lasts for some days but may persist longer. If it is still present some weeks after infarction, left ventricular aneurysm or extensive dyskinesia is likely.

556

T-wave inversion The terminal portion of the T wave begins to invert within hours of infarction and produces symmetrical inversion after a few days (Fig. 12.63). This appearance persists for weeks or months, and is occasionally a permanent feature of the ECG. In some patients at the very earliest stage of infarction the only ECG change may be in the T waves, which become very prominent, peaked and symmetrical.

FIG. 12.63 Evolution of ECG change in an acute inferior infarction Initially there is marked ST segment elevation in II, III, aVF, and reciprocal depression in V2-V4. At 12 hours a Q wave is developing in II, III and aVF, and the ST segment has returned to the baseline. At 48 hours the Q wave has deepened and broadened and there is T-wave inversion.

This appearance is sometimes called the hyper acute pattern of infarction. Symmetrical T inversion alone, without Qwave formation, may indicate a limited infarction in some patients (see non-Q-wave infarction, below) and this has important implications for their management. Q-wave formation Q waves are often the latest change to be seen on ECG. They are often permanent and are sometimes considered the definitive ECG change of infarction. They must have a depth of at least 25% of the succeeding R-wave height (or greater than 3mV) and last longer than 40ms (one small square at 25mm/s). They may appear within 24 hours of infarction and are associated with loss in R-wave height of the corresponding leads. Alternative patterns Ideally, ECG proof of infarction should include successive traces showing the presence and resolution of ST segment elevation. T-wave inversion and pathological Q-wave formation (Fig. 12.63). Often such a classic progression is missing, and the diagnosis may rest on a slightly deeper than normal Q wave in V4, loss of R-wave height or the absence of the small q waves of initial septal depolarization in I, aVL, V5 and V6; this may occur in septal infarc-

12

tion, but is also produced by incomplete left bundle branch block. In the presence of bundle branch block the ECG changes due to infarction may be difficult to detect or absent. With right bundle branch block septal depolarization occurs normally. Changes in ST segments and T waves may appear, but the ECG diagnosis of infarction often cannot be made with certainty. Bundle branch block patterns may precede infarction or be a consequence of it.

Localization of infarction from the ECG Although myocardial infarction can be endocardial, epicardial, or involve the whole thickness of the ventricular wall, these distinctions cannot reliably be made by ECG. Localization to an approximate region of the left ventricle is possible, however. The precordial leads give most information in this respect. If they have been correctly positioned, leads VI and V2 face the right ventricle, although an extra lead (V4R, at the level of V4 on the right of the sternum below VI) is better; V3 faces the interventricular septum; V4 the anterior and lateral borders; and V6 the posterolateral wall of the left ventricle (this is because the apex faces backwards).

FIG. 12.64 ECG of acute anterior infarction ST segment elevation in V1-V6 and Q waves in V2-V4.

Inferior infarction In inferior infarction, changes are seen in leads III, aVF and II, with reciprocal change in I, aVL and the anterior leads. Inferolateral infarction extends to involve lead V6. The right coronary and its posterior descending branch account for most of these cases (Fig. 12.63). Anterior infarction Anterior infarction is extensive when leads V2 to V6 are involved with changes in I, aVL and II. There may be reciprocal changes in the inferior leads II and aVF. Anterior infarction implies involvement of the LAD coronary artery. Infarction may be anteroseptal when the ECG changes are limited to leads V3 and V2. Anterolateral infarction may present with changes in I, aVL and V5, with some involvement in V4 and V6 (Fig. 12.64). Posterior infarction The true posterior infarction is usually caused by occlusion in the circumflex coronary artery, and its hallmark on the ECG is a tall R wave in VI (Fig. 12.65). There is often associated ST depression in leads V2 and V3, with subsequent peaking of T waves in Vl-3. Non-Q-wave and subendocardial infarction Non-Q-wave infarction is a rather imprecise term used to identify patients who have the hallmarks of MI without full ECG change or enzyme rise. Non-Q-wave MI is now more correctly considered with acute coronary syndromes.

Laboratory tests Infarction causes the release of intracellular enzymes into

FIG. 12.65 ECG of acute posterior infarction Inferior ST elevation (in II, III, aVF) and a tall R wave in V1 (mirror of the Q wave) are shown.

the circulation. These may be used to estimate the size of the infarction. Classically, creatine phosphokinase (CK) has been used to detect muscle damage and specificity for cardiac muscle injury improved by assays of a more cardiac-specific isoenzyme (CK-MB). CPK and CPK-MB rise within hours of infarction, and fall within 24-48 hours. The absolute peak of CPK or CPK-MB gives a crude estimate of the size of infarction. Lactate dehydrogenase (or its more cardiac-specific isoenzyme LDH1 is assayed as hydroxybutyrate dehydrogenase (HBD); it appears in blood within 12-24 hours after infarction and may remain elevated for some days (see Table 12.28). Elevations of the contractile protein components, troponin I (cTnl) and troponin T (cTnT) are now taking the place of LDH (HBD) enzymes. Bedside quantitative troponin testing is now widely available and is particularly

557

TABLE 12.28 Biochemical markers of myocardial infarction Marker

Range of time to initial elevation in h

Mean time to peak (h)

CK-MB

3-12

24

cTnl

3-12

24

cTnT

3-12

12h-»2days

LDH

10

24-48

Myoglobin

1-4

6-7

Adapted from Taegtmeyer H Biochemistry of acute myocardial infarction. In: Francis GS, Alpert JS eds Coronary Care. Boston: Little, Brown & Co 1995: 45

useful for stratification in ACS. In patients presenting late, or where CK enzyme analysis is equivocal, the tendency for cTnl and cTnT to persist for days after the event can be diagnostically helpful. Myoglobin is highly specific and appears early, giving promise as a tool for more rapid identification of patients likely to benefit from thrombolytic treatment. A raised polymorphonuclear leukocyte count and an elevated erythrocyte sedimentation rate (ESR) are non-specific accompaniments of acute infarction. A very high ESR in the early stages of infarction raises the possibility of infarction being secondary to a giant cell arteritis or immunological disease such as polyarteritis nodosa.

Imaging techniques Radionuclide studies may be used to measure ventricular function (MUGA scan, p. 487) or document areas of potentially reversible ischaemia (201T1 scan, p. 486). The echocardiogram can be used to assess ventricular function and localize areas of infarction and dyskinesia.

can be produced by the sudden increase in right ventricular work associated with an acute massive pulmonary embolism. Syncope is often the first manifestation of the embolism, which is associated with marked dyspnoea, tachypnoea and cyanosis. Acute pericarditis (p. 586) produces chest pain, often accompanied by a fever and a friction rub.

MANAGEMENT OF ACUTE INFARCTION The aims of management are pain relief, infarct limitation, and the treatment of arrhythmic and mechanical complications. Management is subsequently directed to defining the relatively small proportion of patients at a greater than normal risk in the postinfarction period, and to rehabilitation and secondary prevention.

Analgesia and oxygen Fear and pain exacerbate some of the ill effects of sympathetic hyperactivity. Analgesia and sedation is achieved with opiates, preferably diamorphine (2.5-5.Omg i.v.) or morphine sulphate (4-8mg i.m.). These drugs combine sedation with an advantageous vasodilatory action, which improves haemodynamics. Respiratory depression is rare in the setting of acute infarction. The relief of pain itself may help raise blood pressure and reduce the heart rate. Associated nausea is treated using metoclopramide (10 mg i.v.) or prochlorperazine (12.5-25mg i.m.). An inhaled mixture of nitrous oxide and oxygen (Entonox) provides pain relief prior to arrival in hospital. Oxygen by mask is frequently given to patients with acute infarction to counteract the hypoxaemic effects of pulmonary oedema. However, many patients are not hypoxaemic, and in such cases oxygen delivery to the tissues will not be improved by this treatment.

Acute (3-adrenoreceptor blockade DIFFERENTIAL DIAGNOSIS Acute chest pain syndromes can pose diagnostic difficulties in acute myocardial infarction (p. 544), particularly as there are no specific physical signs in acute infarction and the diagnostic tests may be normal in the initial hours of the attack. With the advent of thrombolysis and acute p-adrenoreceptor blockade used early in the management of acute infarction, diagnostic accuracy has become even more important. The symptoms of aortic dissection (p. 599) may be very similar. Pain indistinguishable from acute cardiac ischaemia

Acute intravenous (3-blockade can be given safely to many acute infarction patients. Large-scale trials have shown a small rate of reduction in early deaths caused by reinfarction and cardiac rupture. An additional effect is a reduction in pain. Acute intravenous -blockade would have to be given to 200 patients to prevent one death, one reinfarction and one cardiac arrest. Chronic therapy after infarction saves two lives for every 100 patients treated for 1 year.

Limitation of infarct size Six hours is the time limit within which it is possible that measures to restore blood supply will salvage ischaemic muscle. If collateral channels are poor, this time limit may be considerably less.

MCQ 12.20

558

Thrombolysis In approximately 80% of patients a myocardial infarction

is caused by thrombotic occlusion of a vessel, usually superimposed on an atherosclerotic plaque. Despite the first use of thrombolytic agents in myocardial infarction as early as 1958, their adoption as one of the mainstays of treatment followed the publication of two giant trials using streptokinase. A single injection of 1.5 x 106 units i.v. over 1 hour within 12 hours of the first symptom reduced mortality by 18%. More than 200000 patients have now been randomized in therapeutic trials with thrombolytic agents. The most widely used agents remain streptokinase and the recombinant tissue plasminogen activator rtPA. Haemorrhagic complications occur, though sufficiently rarely as not to interfere with the overall beneficial effect of therapy. Patients with clotting disorders, a previous history of stroke or peptic ulceration, and those with severe hypertension or any injuries are excluded from this treatment. It is important to exclude aortic dissection, as fibrinolytic treatment can prove catastrophic. Other relative contraindications are a known hypersensitivity to the drug, recent streptococcal infection, survivors from cardiac arrest who have had cardiac massage, and those with heart block who may need a temporary pacemaker. Streptokinase should not be administered within 2 years of a previous treatment. Benefit from thrombolysis is increased if it is given within 6 hours (and preferably 3 hours) of the first symptom. This has led to newer fibrinolytic agents that can be administered at the site of the presenting symptoms, by either the ambulance crew or the GP. Aspirin and heparin Aspirin alone reduces mortality by 25%, and with streptokinase gives a combined reduction of 45%. Continuous infusion of heparin after thrombolysis and after coronary angioplasty is routine in many centres. Patients with large infarctions suffer fewer systemic emboli from ventricular clot if treated with heparin. Bleeding complications and stroke are slightly more common. Heparin treatment is potentially more important after the shorter-acting rtPA, when it should be given for 48 hours. Warfarin treatment is limited to those in AF, or where a large LV thrombus is shown. Angioplasty and surgery Early revascularization during acute infarction may be the revascularization method of choice, but the cost of providing a 24-hour service for these complex, labourintensive techniques is considerable. Individual cases will continue to be so treated in centres where angioplasty or surgical suites are closely linked to the acute admission wards.

Nitrates and calcium antagonist drugs Nitrates by intravenous infusion (2-10 mg/h for isosorbide, 1-12 mg/h for GTN) are frequently used in the acute phase of infarction, where persistent pain is a problem; their use

is associated with a modest reduction in mortality. They have the advantage of causing reduction in preload and afterload, which may relieve left heart failure (see p. 493). Hypotension can be a problem, particularly in patients who are dependent on a high right heart filling pressure to maintain stroke volume, such as those with right ventricular infarction. Cutaneous therapy is best reserved for the less acute stages of infarction, when converting patients from intravenous infusion to maintenance therapy. Calcium antagonist drugs are effective in improving ischaemic cardiac pain, but their use has not been associated with any improvement in infarction mortality. Calcium antagonists, particularly the rate-slowing agents diltiazem and verapamil, can be used when there are contraindications to -blockade. O

12

ACE inhibitors Given within a few days of infarction there is an early benefit, which has now been proved to persist for many years (>4) post MI. These drugs should not be given while patients are hypotensive, particularly during the acute stages of myocardial infarction. Benefit from ACE inhibitors is probably not adequately explained purely by their vasodilatory and antihypertensive effects.

HMG-CoA reductase inhibitors - the statins More than 70% of myocardial infarction survivors in the UK have a total cholesterol >5 mmol/L. There is now convincing evidence that the majority will benefit from being treated with a statin.

Bed rest, mobilization and rehabilitation Bed rest during the acute phase of infarction is aimed at minimizing harmful cardiac remodelling processes. However, prolonged bed rest is unnecessary and can cause complications of its own (such as deep vein thrombosis, pulmonary embolism, deconditioning and bedsores). Prophylactic subcutaneous heparin (12 500 U b.d.) may be indicated if prolonged bed rest is necessary. Mobilization can begin within 24 hours in uncomplicated infarction. Patients can generally be discharged on the 10th day, having demonstrated their ability to walk and manage a flight of stairs before returning home. Mobilization should be slowed if there is a return of chest discomfort, a heart rate below 50bpm or above 120 bpm, the return of ischaemic ST elevation on the ECG, or an increase of more than 15mmHg in systolic pressure with gentle ambulation. Acute myocardial infarction is quite a distressing experience. Patients should be counselled individually and many of them appreciate a clear explanation of the condition and prognosis in simple layman's terms. Advice on risk factors (e.g. stopping smoking) can be initiated at this stage.

559

TABLE 12.29 Thrombolytic agents in myocardial infarction Feature

t-PA

Streptokinase

Anistreplase (APSAC)

Reteplase

Plasma clearance time (min) Fibrin specificity Piasminogen binding Antigen icity Risk of hypotension Method of administration

4-8

15-25

50-90

>20

Minimal Indirect Yes Yes i.v. infusion (1 h)

Minimal Indirect Yes Yes Bolus

High Direct No No Bolus

Licensing

Moderate Direct No No Bolus + i.v. infusion Licensed

Licensed

Licensed

Expense

High

Low

Moderate

Unlicensed in UK ??

TABLE 12130 Contraindications to thrombolytic therapy

TABLE 12.31 Benefits of treatment during and after acute myocardial infarction

Major contraindications Any previous history of haemorrhagic stroke History of stroke, dementia or CNS damage within 1 year Head trauma or brain surgery within 6 months Known intracranial neoplasm Suspected aortic dissection Internal bleeding within 6 weeks Active bleeding or known bleeding disorder Major surgery, trauma or bleeding within 6 weeks Traumatic cardiopulmonary resuscitation

[ACUTE Ml] Deaths prevented/1000 patients treated Treatment i.v. p-Blocker

Relative contraindications Oral anticoagulant therapy Acute pancreatitis Pregnancy or within 1 week postpartum Active peptic ulceration Transient ischaemic attack within 6 months Dementia Infective endocarditis Active cavitating pulmonary tuberculosis Advanced liver disease Intracardiac thrombi Uncontrolled hypertension (SBP >180mmHg, DBP >110mmHg) Puncture of non-compressible blood vessel within 2 weeks Previous Streptokinase therapy (for repeat treatment with SK)

FURTHER READING ON MYOCARDIAL INFARCTION

560

Ball S G (ed) 1995 Myocardial infarction: from trials to practice. Petersfield: Wrightson Biomedical, p. 194 Task Force on the Management of Acute Myocardial Infarction of the European Society of Cardiology 1996. Acute myocardial infarction: prehospital and in-hospital management. Eur Heart J. 17: 43-63

7

Oral -blocker Aspirin

[POST Ml] Major event prevented (including death and Ml) per 1000 patients treated

21 deaths 21 reinfarction 24

7 deaths 9 reinfarction 3 strokes

Streptokinase and aspirin

52

ACE inhibitor HMG-CoA reductase inhibitor

Smoking cessation

5-8 -

? 7 deaths 11 CABG/PTCA 12 Ml 3CVA 4 heart failures 15 deaths 46 reinfarction

(Adapted from British Cardiac Society ACCSAP Commentary)

•

COMPLICATIONS OF MYOCARDIAL INFARCTION ARRHYTHMIAS Arrhythmias are a common complication of myocardial infarction.

Tachycardias Sinus tachycardia Sinus tachycardia (heart rate greater than 100 bpm) is associated with a poor prognosis if it persists despite control of pain, anxiety and cardiac failure. -Blockers will reduce its incidence but cannot be given to patients with severe impairment of ventricular function or hypotension. Ventricular fibrillation VF is a lethal arrhythmia which can complicate acute myocardial infarction from the moment of coronary occlusion. It is the cause of most cases of sudden death due to CHD. Its peak incidence is in the first few hours of the attack, when it is termed primary VF and, if successfully dealt with in an otherwise uncomplicated infarction, has a good prognosis. Secondary VF occurs several days or weeks after infarction, usually in patients who have suffered extensive muscle damage and heart failure. Its prognosis is then very poor. Treatment is by DC defibrillation; the earlier it is used, the more likely is a successful outcome. The procedure itself is described fully on page 519. Lidocaine (lignocaine) is also given after successful defibrillation to prevent further episodes. Limitations on the use of lidocaine (lignocaine) include its narrow therapeutic margin, central nervous system side-effects being the most common. Doses have to be reduced in the elderly and in those with poor hepatic function. The use of lidocaine (lignocaine) to prevent VF is controversial. Many episodes of VF are preceded by multiple or frequent ventricular ectopic beats, or early ectopics falling on the vulnerable period of the ECG - the R on T phenomenon. Warning arrhythmias give frequent false positive and false negative associations and their treatment does not reduce the incidence of VF. Ventricular ectopic beats At an early stage (6-72 hours) in an acute myocardial infarction, frequent ventricular ectopic (VE) beats (>10VE/h) are seen in 70% of patients. Ectopy then diminishes, so that only 10% of patients have frequent VEs at the time of discharge. Frequent VEs at the time of discharge from hospital carries an increased (two- to threefold) risk of death in the first year. This association appears to be independent of ventricular function. Although drugs to effectively inhibit ectopic activity are available, controlled trials have not shown any benefit and, indeed, some worsening survival associated with treatment of frequent or complex VE. Treatment of frequent VEs is indicated if their occurrence causes symptoms due to an effective bradycardia or diminished output. Ventricular tachycardia Haemodynamic disturbance and degeneration to VF is common with sustained VT in acute infarction. DC shock therapy is indicated in almost all cases, to be followed by intravenous or oral therapy (p. 515). Lidocaine (ligno-

caine) is usually used in the acute situation, followed by other Vaughan-Williams class I drugs and amiodarone. Although drug therapy may suppress the arrhythmia the effect on survival is less certain. Patients with non-sustained VT should probably be monitored carefully, and receive a predischarge exercise test and oral antiarrhythmic therapy for the first few weeks after infarction. The risk of death is increased if VT is present at discharge.

12

Accelerated idioventricular rhythm Occasionally a ventricular rhythm at a rate of 60-90 bpm can arise. It generally has broad complexes, but they may be narrow if the origin is high in the bundle of His. This rhythm rarely requires specific treatment, but may be an indication of profound sinus and AV nodal suppression, which may respond to atropine (0.6-1.2mg). Investigation and adjunctive therapy Patients suffering from VF or early ventricular arrhythmia may be treated conventionally. Those with late or secondary arrhythmia deserve very careful inpatient assessment, including angiography. Revascularization by CABG or PTCA may be necessary. Increasingly, patients who remain at risk of sudden death post MI will be offered treatment with an implantable cardioverter defibrillator (AICD), especially if an arrhythmic tendency persists after adequate revascularization. Supraventricular tachycardia Atrial fibrillation (AF) is the commonest supraventricular tachycardia seen in acute infarction, atrial flutter being relatively uncommon. Tachycardias causing severe haemodynamic disturbance are usually best dealt with by DC cardioversion. In atrial flutter this is generally the treatment of choice, as low energies are often required to convert to sinus rhythm. In AF, digoxin is given by oral loading (p. 511) (0.5-1.0mg a.d.) followed by 0.125-0.25 mg o.d. Control of AF rate may require the concurrent administration of another drug (p. 511). -Blockers are often used, and verapamil or diltiazem are also effective. Paroxysmal atrial tachycardia may complicate acute infarction; if hypotension is severe, DC cardioversion is the therapy of first choice. Intravenous adenosine or verapamil is effective in terminating many of these episodes, and intravenous -blockade (atenolol 5mg or metoprolol 515 mg) may be an alternative. If the first-choice drug fails to produce a satisfactory slowing of the rhythm or conversion to sinus rhythm, it is best to use DC cardioversion to avoid additive negative inotropic effects of antiarrhythmic drugs. An alternative therapy for atrial flutter and atrial tachycardia is overdrive pacing; this has the advantage of avoiding negatively inotropic drugs but requires a temporary transvenous pacemaker placed in the right atrium.

Bradycardia Bradycardia is often transient and requires no treatment.

561

When it produces symptoms or impairs the haemodynamics, treatment with a pacemaker or drugs is required. Sinus fyradycardia Up to 80% of patients with acute inferior infarction develop a heart rate less than 60bpm in the very early stages of the condition. This usually responds to atropine (0.5-1.Omg i.v.), but if symptomatic bradycardia persists a temporary pacemaker is required.

Heart block First-degree heart block The lengthening of the PR interval may be seen during the evolution of an infarction, particularly inferior infarction. When pre-existing it requires no action. However, lengthening PR intervals deserve observation, as this implies interference with AV node function, and progression to higher grades of block may occur. Second-degree heart block Both Mobitz types I and II (p. 502) require very careful observation. Progression to complete heart block and sudden death is uncommon with inferior infarction, where Mobitz type I is a frequent, though transient, complication (about 15% of cases). In anterior infarction with extensive involvement of the interventricular septum, Mobitz type II block may be seen and carries a poor prognosis (80% mortality) which owes more to the associated large mass of muscle loss than to the conduction disturbance itself. Temporary pacemakers should be used in all cases where second-degree block produces symptoms (e.g. syncope) or haemodynamic disturbance, or when antiarrhythmic drugs are to be given. A permanent pacemaker should be considered for survivors of anterior infarction with persistent block. Third-degree or complete heart block Complete heart block is a common complication of inferior infarction (about 10% of cases), when treatment is along the same lines as for second-degree block. It needs no specific treatment if well tolerated, but temporary cardiac pacing should be employed if heart failure, or rates less than 50bpm, occurs. In anterior infarction, complete block (incidence approximately 5% of cases) may occur suddenly or be preceded by bifascicular block (see below). It generally occurs in large infarctions involving the septal tissue where the conducting tissue is concentrated, and always requires treatment with a permanent pacemaker. This may not always improve survival but management is considerably eased. Complete heart block rarely persists, but if it does persist beyond 7-14 days after infarction a permanent pacemaker is required.

562

Bifascicular and trifascicular heart block Interruption of conduction through the right bundle and

either fascicle of the left bundle constitutes bifascicular block. When the anterior fascicle is involved (left anterior hemiblock, LAH) the QRS axis is greater than -30° with small r waves in II, III and aVF and absent initial q waves in V5 and V6. Posterior fascicular block (left posterior hemiblock, LPH) produces a rightward axis shift (greater than +100°) with an S wave in I and Q in III. Trifascicular block is any combination of bifascicular block associated with a long PR interval. The association of right bundle branch block with LPH in anterior infarction indicates considerable ischaemic damage, and its appearance is an indication for prophylactic use of a temporary pacemaker, as progression to complete block may be sudden. Right bundle branch block with left anterior hemiblock does not carry such a poor prognosis, and is a slightly less powerful indication for pacing.

Temporary transvenous cardiac pacemakers The indications for implanting a pacemaker in acute infarction are summarized in Table 12.32. There are various methods by which pacing can be achieved, but in general transvenous routes are the most stable and can be maintained for the few days that are required. The methods require the insertion of a bipolar electrode into the right ventricular apex, usually under radiological control. Pacing thresholds of under 1 V at 2 ms pulse width duration should be established, as the threshold tends to rise after implantation. To achieve this, it may be necessary to try various positions in the right ventricular apex. The pacemaker is an external box which is always set in the demand mode; this ensures the patient's ventricular activity is sensed and pacing only begun when the intrinsic heart rate falls below a preset rate (usually between 70 and 90bpm). A stable position of the temporary electrode is confirmed by establishing pacing at an output of 1 V, and ensuring ventricular capture is not lost during maximal inspiration, coughing or sniffing. There should be sufficient slack in the electrode to accommodate these movements of the thorax without displacing the tip of the electrode. Once the electrode position is finalized it is secured to the skin under a sterile dressing. Except in an emergency, it is rarely justified to manipulate a temporary pacemaker after it has been inserted for some time, as sterility cannot be assured. If displacement has occurred, a replacement pacemaker will be required. Threshold measurements should be checked daily, or more frequently if the patient is pacemaker dependent. The pacemaker settings are adjusted to provide a voltage output of at least three to four times threshold. A threshold of over 2 V is an indication to replace the electrode. Atrial and dual-chamber pacing Although pacing via a ventricular electrode is sufficient in most instances, there are circumstances where atrial or dual-chamber pacing is required. The former can be used to treat atrial flutter or bradycardia-dependent rhythms

TABLE 12.32 Indications for temporary pacemaker in acute myocardial infarction (Ml) A. Indicated regardless of Ml site Symptomatic and drug-resistant bradycardia due to:

B. Indicated in anterior Ml All indications in A

C. Not indicated First-degree heart block alone

sinus bradycardia, sinus arrest, SA block

RBBB + LPHB

Asymptomatic Mobitz type I

second-degree heart block

RBBB + LAHB

Asymptomatic bradycardia or junctional rhythm

third-degree heart block

Trifascicular block:

Sick sinus syndrome with bradycardia-tachycardia Drug-resistant tachycardia DC shock-resistant tachycardia Ventricular standstill Alternating LBBB with RBBB

long PR + LBBB long PR, RBBB + LPHB or LAHB Second-degree heart block, especially if Mobitz type II

12

LAHB ' LBBB RBBB

in isolation

LPHB .

LAHB = left anterior hemiblock; RAHB = right anterior hemiblock; LPHB = left posterior hemiblock; LBBB = left bundle branch block; RBBB = right bundle branch block

when AV nodal conduction is intact. Electrode positioning is more difficult and stability can be a problem. In dual-chamber or physiological pacing, sequential pacing of first the atria, followed, after a suitable pause, by the ventricles, offers a near physiological system maintaining the atrial contribution to cardiac filling. This can be very important in severe ventricular impairment, right ventricular infarction, or when there is associated valvular disease, such as aortic stenosis. The technique requires suitable dual-chamber pacemakers and the expertise to implant atrial as well as ventricular electrodes; availability is thus likely to be limited to specialized cardiac centres.

CARDIAC FAILURE Dyspnoea, basal crepitations and the radiographic signs associated with pulmonary oedema (p. 492) are frequently encountered in the early stages of infarction.

Management Prompt treatment with intravenous loop diuretics (furosemide (frusemide) 20-40 mg or bumetanide 1-2 mg) has been the mainstay of treatment. Larger initial doses are often given, but are rarely necessary and may produce hypotension, worsening the patient's overall condition. Diamorphine 2.5-5.0 mg i.v. is also helpful, and oxygen, given by a close-fitting, low dead-space mask (e.g. MC mask at 4L/min) is indicated. The preferred treatment is now with vasodilatation using intravenous nitrates, reducing preload. Patients with low blood pressure and heart failure are best managed using arterial and indirect left atrial pressure measurement. The latter is obtained with a Swan-Ganz catheter placed in a pulmonary arterial branch (see below). The aim of treatment should be to decrease the indirect left atrial pressure (or pulmonary capillary wedge pressure) to less than 20mmHg, while maintaining systolic blood pressure above lOOmmHg and diastolic

pressure over 60mmHg. Nitrates with inotropes (dobutamine ± dopamine) may be better than diuretics in these difficult cases. A week after infarction, introduction of ACE inhibitors has been shown to improve prognosis.

CARDIOGENIC SHOCK Cardiogenic shock in myocardial infarction is defined as hypotension with a systolic pressure under lOOmmHg; it is invariably accompanied by ventricular failure, peripheral vasoconstriction and oliguria (<20mL/h of urine), and usually arterial hypoxaemia, mental confusion and metabolic acidosis (Table 12.33). Remediable causes, such as uncontrolled pain, arrhythmia or excesses of negatively inotropic or vasodilating drugs (e.g. p-adrenoreceptor blockers or calcium antagonists and diamorphine) should be excluded and treated specifically. Shock is caused by damage affecting 30-40% of ventricular muscle mass and, if untreated, carries an 80-100% acute mortality. Treatment is by early revascularization, especially by PTCA. It may reduce early mortality to 50-60% at 1 month.

RIGHT VENTRICULAR INFARCTION Right ventricular infarction may produce cardiogenic shock, as defined above, but carries a very much better prognosis. The findings are a lower than expected indirect left atrial pressure and a relatively high right atrial pressure. The ECG shows changes of inferior or posterior infarction, and lead V4R may reveal the infarction clearly. Treatment consists of maintenance of a high right-sided filling pressure, which may have been inadvertently lowered by diuretics given to treat the seemingly high venous pressure. The cardiac output and systemic pressure may respond to infusion with colloid. Inotropes and diuretics may thus be avoided altogether in these patients.

563

TABLE 12.33 Cardiogenic shock in myocardial infarction Features (in increasing order of severity) • hypotension (SBP<100 mmHg) • thready impalpable peripheral pulses • pale and cool hands/feet/nose ± peripheral cyanosis • oliguria/anuria • sweat! ness • mental obtundation Exclude shock due to: • absolute/relative hypovolaemia (especially in the presence of RV infarction) • drugs (especially diamorphine, morphine, nitrates, p-blockers) • vasovagal reaction (tends to be bradycardiac) • heart block/bradyarrhythmia • tachyarrhythmia • tamponade • other causes (septicaemia, anaphylaxis) Investigations • echocardiography (essential) • haemodynamic monitoring - thermodilution Swan-Ganz catheter ± arterial cannula

Management Pressure monitoring, including indirect left atrial pressure, systemic arterial pressure and a facility for measuring cardiac output, are the minimum requirements for managing patients with shock. Therapy consists of oxygen,

RECENT ADVANCES p-BLOCKADE FOR HEART FAILURE

564

Following studies showing that angiotensin-converting enzyme inhibitors (ACEIs) can prolong the survival of patients with heart failure, a recent study has shown that carvedilol, a vasodilating (3-adrenergic blocker, used in addition to ACEIs, can further improve survival. This result has produced a major shift in our ideas of managing heart failure, away from that of using positive inotropic and chronotropic agents to enhance cardiac performance to that of using a negative inotropic and heart slowing agent instead. The exact mechanisms of why ACEIs and -blockers produce beneficial effects whereas oral positive inotropic agents (such as milrinone, enoximone, flosequinan) worsen mortality is unknown, but one simple likely explanation is that activation of the sympathetic and renin-angiotensin-aldosterone systems commonly seen in heart failure leads to increased loss of cardiac myocytes, which is further accelerated by positive inotropic agents, and can be substantially diminished by using ACEIs and P-blockers.

inotropic agents, diuretics and, very rarely, mechanical support of respiration by ventilation and of the circulation with intra-aortic balloon counterpulsation (see below). Inotropic agents are required to maintain the systolic pressure to perfuse the tissues, but are arrhythmogenic and may increase oxygen requirements. Diuretics are used to relieve pulmonary oedema, but adequate ventricular filling pressures must be maintained or else cardiac output falls. The inotropic agents of choice are dopamine, dobutamine and phosphodiesterase inhibitors (milrinone, enoximone and amrinone). Dopamine Dopamine has three important actions, which appear at increasing dose ranges. Low-dose dopamine infusion (0.5-2.5 jig/kg/min) primarily produces renal and mesenteric vessel dilatation, and is frequently used at this dose

RECENT ADVANCES IN VENTRICULAR ASSIST DEVICES AND ARTIFICIAL HEARTS With improvements in the treatment of various cardiac diseases (notably myocardial infarction and lifethreatening arrhythmia), the incidence of heart failure rises steadily. The demand for cardiac transplantation cannot be met because of a severe shortage of donor hearts. Alternatives are being explored. Transplanting genetically engineered immunologically compatible animal hearts may become feasible, but there are ethical problems raised by the antivivisection lobby and the theoretical fear of trans-species transmission of unknown viral or prion infections. Technological advances have made the development of a reliable artificial heart a distinct possibility in the next few decades. The first problem was infection, especially at the points of insertion of tubes through the skin. The development of a new battery system, which can be charged across the intact skin, has made possible the implantation of the entire device inside the body. The second major problem is compatibility with blood, which advances in material science have largely overcome. Other problems, such as avoidance of haemolysis, size of device, fracture, bleeding and thrombus formation, have to be surmounted. The comparison of today's pacemaker technology with that of 50 years ago provides optimism that artificial hearts may become a medical reality in this century. In the meantime, those patients who suffer severe transient reversible myocardial hypofunction (for example due to myocardial stunning) may require mechanical support of the circulation to 'rest' the myocardium and allow it to recover. The use of ventricular (usually left ventricular, but occasionally biventricular) assist devices to support the circulation for a short period allows the patient to buy time until myocardial function is restored.

12

CASE STUDY 12.4 CHEST PAIN AND PULMONARY OEDEMA A 67-year-old diabetic man with no previously known cardiac history was transferred as an emergency to the cardiology unit. He had presented a few hours earlier to his local emergency department with a 3-day history of chest pain and was noted to have an ECG consistent with an extensive antero-septal infarction. There were widespread Q waves visible in leads V2-V6, and there were virtually no R waves in the anterior chest leads. 3 mm of ST segment elevation was visible from V2-V4. They were unable to give him thrombolysis due to hypotension; his blood pressure on arrival was 70/40. His chest X-ray showed gross pulmonary oedema, and his oxygen saturation on air was low and could only be barely maintained with a high volume oxygen mask. On arrival to the cardiac unit, he was noted to be tachycardic with a blood pressure of 60/40 and barely conscious. Oxygen saturation was 88% with signs of gross pulmonary oedema. He had an indwelling urinary catheter and a documented urine flow of 50 mL of urine over the previous 3 hours. The ECG was as described by the referring hospital. Arterial blood gases confirmed hypoxaemia with an acidosis giving a pH of 7.14 and a base excess of 13.

Question How should this patient with cardiogenic shock due to recent cardiac infarction be managed? Discussion This man presents the picture of cardiogenic shock due to myocardial infarction. This carries a greater than 80% acute mortality. Firstly he requires resuscitation and stabilization. His gas exchange is poor, he has pulmonary oedema and is exhausted by the effort of breathing. He was therefore intubated as an emergency and ventilated. Unfortunately, the act of intubation led to a further fall in systemic pressure. Although pharmacological support of pressure could have been instituted with dobutamine by intravenous infusion, it was decided to support the circulation by inserting an intraarterial aortic balloon pump. He was transferred to the cardiac catheterization laboratory, where the pump was inserted percutaneously from the right femoral artery. One-toone ventricular assistance was set up with an immediate improvement in mean arterial pressure from 40 to 60 mmHg. Signs of ischaemia on the ECG improved as did blood gases and acidosis.

as a renal protector during times of systemic hypotension. Medium-dose dopamine infusion (5.0-lOug/kg/min) has a direct myocardial inotropic action, and at high doses (>10[ig/kg/min) the drug produces peripheral vasoconstriction (a-adrenoreceptor mediated), thereby raising systemic pressure at the cost of vasoconstriction in tissue beds, including renal. The drug must be infused through a large central vein, and at medium to high infusion rates may precipitate cardiac arrhythmias. Dobutamine Dobutamine is a synthetic dopamine analogue which directly stimulates (3i receptors; unlike dopamine, it does not release noradrenaline (norepinephrine) and so is less affected by the reduction in cardiac stores of this hormone in chronic heart failure. Dobutamine can be used via a

The patient proceeded immediately to angiography which revealed a proximal occlusion on the left descending coronary artery but normal circumflex and right coronary. An infusion of a Ilb/IIIa antagonist was begun and the occlusion re-opened by balloon angioplasty with a subsequent stent insertion. The angioplasty was uncomplicated apart from a short burst of ventricular tachycardia soon after reperfusion. This did not require treatment. He was transferred back to intensive care and was able to be extubated the next day. Cardiogenic shock is an extremely severe complication of myocardial infarction with a very high mortality. It appears likely that mortality is reduced to 40-50% by revascularization using emergency angioplasty. In these circumstances intra-aortic balloon counter-pulsation is the treatment of choice to support the circulation without the complications of additional pharmacological therapy. The patient recovered and was able to be discharged from hospital. He had some residual impairment of ventricular function and subsequent treatment was along the lines of secondary management of a myocardial infarction.

peripheral vein and the dose is usually between 2.5 and 40|ig/kg/min. Its action is primarily inotropic, and hence improves cardiac output while at the same time reducing ventricular filling pressures. Combination therapy Dopamine in low dose to promote renal bed vasodilatation and dobutamine in doses to achieve an improved cardiac output are often used in combination in cardiogenic shock; lower doses of each can thus be used. Filling pressures (measured via a Swan-Ganz catheter) are titrated against systemic pressure and the drug dosages adjusted every 10 minutes to achieve an optimal cardiac output and sufficient perfusion pressure to maintain renal, cerebral and myocardial function while avoiding excessive heart rates and pulmonary oedema. Ideally, therefore, the

565

dose of dobutamine is adjusted every 10 minutes until the cardiac output reaches a plateau. Dopamine is adjusted to achieve a reasonable urinary output (>60mL/h); this may require support with intermittent bolus doses of loop diuretics. Left atrial pressure is maintained between 15 and 20mmHg (below the threshold to produce pulmonary oedema and above the level that begins to impair cardiac output). The right atrial pressure is maintained high enough to support the cardiac output and left atrial pressure. The heart rate is kept between 100 and 120 bpm and systolic blood pressure between 110 and 120mmHg, with a diastolic pressure near SOmmHg. The patient should be comfortable in this state, with near normal arterial oxygen saturation and no acidosis. Mechanical support Mechanical support of the circulation can be achieved using intra-aortic balloon pump counterpulsation. Mechanical ventilation of patients in cardiogenic shock is justified if they are to be offered more definitive therapy. Coronary angioplasty improves the outlook in these patients. The technique involves passing a balloon catheter into the descending aorta, by percutaneous techniques allied to cardiac catheterization. The balloon is then inflated with helium and deflated, triggered to the ECG. This produces an enhanced aortic diastolic pressure.

ACUTE VENTRICULAR SEPTAL RUPTURE AND RUPTURED PAPILLARY MUSCLE Acute ventricular septal rupture and ruptured papillary muscle are uncommon and very severe complications of acute myocardial infarction. They tend to occur 2-10 days after the acute event and produce sudden deterioration, with acute heart failure, shock, and the development of a new systolic murmur. It may be impossible to differentiate the two conditions clinically, although echocardiography and Doppler ultrasound can do so in most cases. Treatment is early surgery in both instances, although mortality remains high. There may be an advantage to allow tissues to 'heal' before closure of a VSD, but unfortunately most patients will succumb before sufficient time elapses. There are some conspicuous successes with early surgery and, when possible, circulatory support prior to operation may be life-saving.

cally by the sudden onset of severe pain followed rapidly by cardiovascular collapse. Resuscitation from this event is unusual; the clinical picture is of electromechanical dissociation (EMD), where an ECG rhythm is present but no output ensues.

VENTRICULAR ANEURYSM Ventricular aneurysm is a late complication of myocardial infarction in 10-15% of patients. True aneurysms cause thinning and paradoxical motion of portions of the ventricular wall. Recurrent arrhythmia, refractory heart failure, embolization and rupture are the complications of true aneurysms, which are manifest clinically by paradoxical precordial pulsation, persistent ST elevation on the ECG and an abnormal LV border contour on the chest Xray. False aneurysms are limited cardiac ruptures whose structure has become organized and stabilized; the main complications include those of true aneurysms as well as a greater tendency for late and fatal rupture. Surgery is the treatment of choice for all false aneurysms and symptomatic true aneurysms. Antiarrhythmic therapy may be necessary and anticoagulation for true aneurysms is wise, as 60-70% of them contain thrombus.

DRESSLER'S OR POSTMYOCARDIAL INFARCTION SYNDROME Dressler's syndrome is an immunological syndrome of pleuropericarditis which may occur in 1-4% of patients, 2-12 weeks after infarction (or open heart surgery). Antibodies directed against the sarcolemma and subsarcolemma of myocytes may be found. The clinical syndrome is similar to acute pericarditis (p. 586), with pain, fever, leukocytosis, a high ESR and a 'pericarditic' ECG in 50% of cases. It must be differentiated from extension of myocardial infarction, pulmonary embolism or pneumonia. Treatment usually consists of NSAIDs, although steroids may be required for severe and recurrent episodes. The prognosis is good.

PROGNOSIS AFTER MYOCARDIAL INFARCTION

CARDIAC RUPTURE

566

Cardiac rupture is a much commoner cause of sudden death in hospital than has previously been recognized (10-20% of in-hospital deaths). It typically occurs several days after the acute infarction and appears to be commoner in men, in those with higher blood pressures, and in those who suffer anterior infarction. Acute -blockade protects from this complication, which is manifested clini-

After a myocardial infarction the prognosis of the patient is dependent on the success of treatment of the complications described above, the likelihood of recurrence or extension of the infarct, and the extent of eventual impairment of cardiac function. Despite modern advances in medicine, some 40% of myocardial infarction victims die without reaching the hospital. In the prethrombolytic era the introduction of coronary care units resulted in an

approximate halving of hospital mortality rates from about 30% to about 15%, and the predominant cause of death changed from fatal arrhythmias to heart failure. After surviving an acute myocardial infarction, the prognosis is dependent on: 1. The extent of ventricular impairment 2. The propensity to life-threatening arrhythmia 3. The possibility of further infarction. The extent of left ventricular dysfunction is dependent on: • The state of left ventricular function prior to the infarction • The size and nature of the infarction • The subsequent repair and remodelling processes • The amount of viable myocardium still in jeopardy from ischaemia (inadequate blood supply and cardiac myocytes functioning in energy deficit), stunning (reasonable blood supply but myocyte contraction impaired) or hibernation (reduced blood supply but myocytes downregulated so as not to be in energy deficit). Thus, patients with diminished starting numbers of cardiac myocytes (due to either previous myocardial infarctions or old age - secondary to natural attrition of myocytes, which averages about 35% from the age of 18 to 90 years, or previous cardiomyopathy) or pre-existing ventricular dysfunction from any cause (e.g. valvular or hypertensive heart disease) would not tolerate the same myocardial infarction as well as those with normal or athletic hearts. Obviously, patients incurring a larger myocardial infarction are likely to have a worse prognosis than those with a smaller infarction. The repair processes post infarction may be impaired (e.g. by concomitant therapy with corticosteroids or NSAIDs), leading to a greater likelihood of infarct expansion and ventricular aneurysm formation, resulting in a ventricle that functions less well than one with normal scar tissue formation. The most important long-term prognostic indicator is left ventricular dysfunction. The clinical diagnosis of heart failure is a good indicator of dysfunction; patients with heart failure post infarction have 2-4 times greater mortality rates than those without. The simplest measurements of dysfunction are left ventricular ejection fraction or endsystolic volume, the depression of which has been found to correlate with mortality. These patients deserve a more aggressive approach to treatment, in terms of risk factor reduction, ACE inhibition and, if indicated, revascularization (see below). The presence of arrhythmias at rest or during exertion is an adverse prognostic indicator after myocardial infarction. Arrhythmias can be identified by Holter monitoring, exercise tests or direct electrophysiological studies. Frequent ventricular ectopics or inducible ventricular tachycardia are predictive of sudden death. Late potentials in averaged ECGs and reduced heart variability are also predictive of life-threatening arrhythmias. High-grade AV

block and bundle branch block resulting from myocardial infaction are associated with a poor outcome. Residual myocardial ischaemia, as manifested by postinfarction angina or positive exercise tests, is an important predictor of reinfarction and mortality. High-grade stenosis or occlusion of the infarcted related artery supplying a large portion of the left ventricle is an independent prognostic indicator. Much prognostic information can be obtained from clinical data and non-invasive investigations (e.g. ECG, chest X-ray, exercise tests, echocardiography) and, if indicated by these, more sophisticated or invasive investigations may then be performed. It is important to point out that, although various parameters may be predictive of prognosis, this does not imply that therapeutic attempts to improve these parameters necessarily lead to benefits to patients or improved outcome. A rational approach to management is required.

12

FURTHER READING ON PROGNOSIS AFTER MYOCARDIAL INFARCTION Emanuelsson H, Karlson B W, Herlitz J 1994 Characteristics and prognosis of patients with acute myocardial infarction in relation to occurrence of congestive heart failure. Eur Heart J 15: 761-768. Gaudron P, Eiles C, Kugler I, Ertl G 1993 Progressive left ventricular dysfunction and remodelling after myocardial infarction. Potential mechanisms and early predictors. Circulation 87: 755-763. Stevenson R, Ranjadayalan K, Wilkinson P, Roberts R, Timmis A D 1993 Short and long term prognosis of acute myocardial infarction since introduction of thrombolysis. Br Med J 307: 349-353.

MANAGEMENT OF MYOCARDIAL INFARCTION SURVIVORS The objectives of management post infarction are to initiate or continue the management of ongoing complications described above, and to minimize the chances of further myocardial infarction, either as infarct extension or as infarction in a new territory. The ideal treatment for the latter is to produce a regression of all coronary artery stenoses, but this remains unattainable. Current techniques of revascularization (coronary bypass surgery or angioplasty) are not without risk and should therefore be reserved for those with most to gain in terms of symptomatic relief and/or prognosis. Assiduous attention to risk factor management is required as part of the rehabilitation programme.

Investigations after myocardial infarction The objective of investigation is: • To select high-risk patients for therapy which has been shown to reduce mortality, reinfarction or heart failure

567

• To establish the cause (s) of symptoms • To exclude patients who will derive little or no benefit from therapy, or are likely to be harmed by therapy • To assess functional capacity and prognosis. To this end many patients would benefit from a formal predischarge or follow-up stress test (e.g. exercise ECGs, exercise/stress radionuclide myocardial perfusion scan) and an assessment of left ventricular function (using either echocardiography or radionuclide ventriculography). (A cautionary note: interpretation of exercise ECGs after infarction requires skill, as false positive results are not infrequently found.) The incidence of subsequent coronary events is low in those who have a good exercise capacity without positive test results. Coronary angiography may be considered in those with stress-induced angina or evidence of myocardial ischaemia, especially in patients with impaired left ventricles. Patients with postinfarction angina, especially after non-Q-wave myocardial infarction, even without positive exercise ECGs, should be considered for coronary angiography.

Drug therapy A simple principle recommended when introducing drug therapy in the coronary care unit is that any cardioactive agent is preferably introduced using small frequent doses of the shortest-acting drugs to avoid major or prolonged adverse reactions, and the doses increased subsequently or changed to longer-acting versions. The exceptions to this rule are the treatment of life-threatening arrhythmias or prevention of coronary artery spasm or acute occlusion, when known full effective doses that will cover the entire 24 hours of the day should be introduced, and the doses reduced or changed to alternatives if side-effects become intolerable. p-Blockers The most successful agents in secondary prevention of further coronary events following myocardial infarction are the (3-blocking drugs. Despite differences in the pharmacology of different agents, it is likely that this effect is common to the whole class of drugs, the only exceptions being those with high intrinsic sympathomimetic activity, such as oxprenolol. Sudden death and reinfarction appear to be successfully prevented. Data from all the randomized trials suggest a 10-25% reduction in mortality. Most events occur within the first 6-12 weeks of infarction. Treatment with (3-blockers is therefore recommended for at least 3 months, and in high-risk individuals for a year. Patients at very low risk of reinfarction or death may not require this form of treatment.

Case 12.6

568

ACE inhibitors In clinical trials involving some 105 postinfarction survivors, ACE inhibitors have been shown to reduce mortality significantly by 5-27%, the greater reduction being in patients with clinical heart failure or demonstrable left ventricular dysfunction after infarction. Although mortality reduction is demonstrable when ACE inhibitors are given to unselected patients post infarct, the overall statistical extension of life achieved thereby amounts to only a few days. Hence this approach may not be considered cost-effective. Initiating ACE inhibitors too early following an acute infarction was associated with no benefit, and in some patients it may be harmful. In patients who benefit most from ACE inhibition (i.e. those with heart failure) it is advisable to delay introduction until the haemodynamic status becomes stable. ACE inhibitors are contraindicated in cardiogenic shock or preshock, and in unstable angina vasodilators with antianginal properties should preferentially be used. The initiating dose should be small to avoid protracted hypotension in unexpected cases, and the dose increased stepwise to a maintenance dose of, for example, ramipril 5mg b.d., lisinopril 10 mg o.d. or trandolapril 4mg o.d. Treatment may need to be continued indefinitely in patients with continuing heart failure requiring chronic diuretic therapy, and in whom there is demonstrable left ventricular systolic impairment, provided the ACE inhibitor is tolerated. Those who have intolerable sideeffects may be considered for angiotensin II receptor inhibition (e.g. with losartan), although whether this drug possesses the same beneficial effects as ACE inhibitors requires further studies. Those with transient failure after infarction may need ACE inhibition for about 6 months when the cardiac remodelling processes have stabilized. Antithrombotics Aspirin is a safe antithrombotic. It is economical and well tolerated in the absence of peptic ulcer disease. Its role after infarction is not well established and the precise dosage is contentious. A very small dose of 75 mg dissolved in lOOmL of water every day, or even alternate days, may have advantages over larger doses, as the desired antiplatelet activity is then achieved without the unwanted inhibition of endothelial wall prostacyclin activity. Other antiplatelet drugs, e.g. sulfinpyrazone and dipyridamole, have no demonstrated advantage over aspirin alone. Ticlopidine and clopidogrel may be used in patients who are intolerant of or allergic to aspirin. Anticoagulants Anticoagulation with warfarin following myocardial infarction has been a topic for controversy for many years. Although a small benefit in survival is apparent, particularly for a small subgroup of compliant elderly patients, there are considerable problems associated with its safe usage over many years. Where warfarin has been com-

pared with aspirin, the benefits are equivalent and sideeffects are equally common, although different. However, aspirin therapy requires much less supervision and hence is preferable for patients and physicians alike. Patients with extensive anterior infarction, with demonstrable intraventricular clot (shown by echocardiography), can have the risk of further emboli reduced by warfarin anticoagulation. The risks of emboli are highest in the first few weeks, hence treatment may only be required for 3-4 months. Lipid-lowering agents As in treatment of any condition, there needs to be a weighing of benefits versus risks. Lipid-lowering agents are no exception in possessing unwanted effects, such that, although they are capable of reducing cardiovascular mortality and morbidity, there was a fear that they might increase non-cardiac deaths. Low-cholesterol and low-fat dieting is much safer, and all infarct survivors should be strongly encouraged to adhere to this recommendation. If the serum cholesterol levels are still elevated (above 5.5mmol/L) despite strict dieting, an HMG-CoA reductase such as simvastatin or pravastatin (10-40 mg o.d.) may be introduced. These drugs have been shown to improve longterm (about 6 years) survival as primary or secondary prevention in subjects with elevated cholesterol, although their effect beyond 10 years is unknown.

Cardiac rehabilitation Rehabilitation is a process whereby patients are helped realistically to regain their physical, mental, social, vocational and employment potential in their community. It is an intrinsic component of the clinical management of cardiac disease. In a climate of limited resources, the patients to be targeted are those with the highest potential for health gain. After myocardial infarction, patients with the highest risks (such as those with heart failure, arrhythmia, angina and hypertension) and those in employment prior to infarction are the ones who should be preferentially entered into rehabilitation programmes. In many centres there has been a tendency to select for rehabilitation those who are already in excellent prognostic categories, and the potential for gain is small. Results of rehabilitation studies suggest a benefit in survival with the additional decrease in other coronary events. The phases of rehabilitation are: 1. Week 1 post infarction, either in hospital or at home (if the patient is not admitted) 2. For a few weeks during convalescence 3. The active outpatient hospital-supervised period - up to 6 months 4. Community and long-term follow-up. Hospital-based rehabilitation programmes usually encompass phases 1, 2 and 3, whereas phase 4 takes place in

the primary care setting, health clubs or cardiac patient support groups. The components of rehabilitation include counselling, psychosocial care, risk factor management, education, exercise training, relaxation therapy and vocational assessment. One purpose of counselling is to explain the medical diagnosis, the results of investigations, the pros and cons of intervention, and the prognosis and implications in sympathetic, positive but realistic terms which patients can understand. Another main objective is to encourage behavioural change rather than simply to provide emotional support. Persistent anxiety and depression occur in about 25% of patients and relatives, and these usually respond to correction of misconceptions and addressing their main concerns. Patients who undergo exercise rehabilitation attain their optimal functional state more rapidly, have fewer visits to their doctors and are more likely to return to work. Formal exercise testing is essential to determine the correct individualized exercise prescription. It can also be used to monitor progress. Patients with a negative exercise test at a good workload have an excellent prognosis, and this information is of great psychological benefit. The type of exercise most beneficial for cardiovascular fitness is aerobic isotonic exercise (e.g. brisk walking, cycling, swimming or jogging). The amount and duration of exercise training that is required to produce benefit is 20 minutes of exercise at 75-80% of the maximal heart rate achieved on exercise testing, 2-3 days per week. Training effects may be seen after 6-12 weeks of this degree of effort. Ventricular function and, possibly, myocardial perfusion improve with prolonged exercise training. Most centres encourage the continued use of -blockers throughout exercise programmes. O

12

FURTHER READING ON MANAGEMENT OF MYQCARDIAL INFARCTION SURVIVORS Betteridge D J, Dodson P M, Durrington P N et al 1993 Management of hyperlipidaemia: guidelines of the British Hyperlipidaemia Association. Postgrad Med J 69: 359-369. Cleland J G F 1995 Clinical perspective. ACE inhibitors for myocardial infarction: how should they be used? Eur Heart J 16: 153-159. Scandinavian Simvastatin Survival Study Group 1994 Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 344:1383-1389. Shepherd J, Cobbe S M, Ford I et al 1995 West of Scotland Coronary Prevention Study Group. Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. N Engl J Med 333:1301-1307. Smith G D, Song F, Sheldon T A 1993 Cholesterol lowering and mortality: importance of considering initial level of risk. Br Med J 306:1367-1373. Tan L B, Ball S G 1994 ACE inhibition after myocardial infarction. London: Science Press, pp. 1-78. Thompson B R, Bowman G S, Kitson A L, de Bono D P, Hopkins A 1996 Cardiac rehabilitation in the United Kingdom: guidelines and audit standards. Heart 75: 89-93.

569

TABLE 12,34 Causes of myocarditis Cause

Example

Viral

Coxsackie B, but a wide range of enteroviruses, echoviruses, adenoviruses, polio, influenza Chlamydia and coxiella (Q fever) Infants and immune-compromised adults T. cmzi (Chagas1 disease of S. America) Often toxin-mediated, e.g. diphtheria Radiation, severe hypothermia Emetine, carbon monoxide

Rickettsiae Toxoplasma Trypanosomiasis Bacteria Physical Drugs and poisons

DISEASES OF HEART MUSCLE

MYOCARDITIS The causes of myocarditis are shown in Table 12.34. In the Western world the commonest causes are viral, of which the most important are Coxsackie B and echovirus infections. Toxoplasmosis affects the myocardium in the neonate. Trypanosoma cruzi is the major cause of infective heart muscle disease in South America and is discussed on pages 354-355. Acute viral myocarditis may complicate many severe viral infections and has been described with almost all common viral diseases. Clinically important myocarditis is uncommon, but ECG changes, possibly due to myocarditis, are more frequent.

Clinical features A febrile illness followed by acute unexplained heart failure, cardiac arrhythmias and chest pain are the main presenting features of myocarditis. Chest pain usually results from a myopericarditis. An arrhythmia, such as frequent ventricular ectopic beats, is probably the most frequent manifestation of a mild attack of myocarditis, and is often seen after any viral infection. The physical signs may include gallop rhythm and heart failure with cardiac enlargement.

Diagnosis and investigation Diagnosis is made on the basis of the history and an ECG, which may show a variety of arrhythmias, conduction defects, and myocardial injury with ST and T-wave abnormalities. An echocardiogram is helpful in diagnosis and

Fig.

570

12.24

shows reduced ventricular contractility. Endomyocardial biopsy is occasionally necessary to make the diagnosis. This technique is used to obtain a small piece of myocardium for histological examination. In acute myocarditis there is a marked lymphocytic and inflammatory cell infiltrate. Necrosis of myocytes is seen and, often, interstitial fibrosis. Later in the disease the histological appearances become less marked and merge into those of congestive (dilated) cardiomyopathy (see below).

Management Strict rest is advised until symptoms and signs and ECG abnormalities have resolved. Strenuous exercise is prohibited during this period. Cardiac arrhythmias are treated. Anticoagulants should be considered, as intraventricular clots are common. Although most patients make an apparently full recovery, about 15% have a poor prognosis with continuing evidence of cardiac enlargement and, in some cases, of progressive heart failure. These patients may only be saved by cardiac transplantation. It seems likely that some patients will progress to a dilated cardiomyopathy. Immunosuppressive therapy may be helpful in the more severe cases when they present in the early phases of myocarditis, preferably with histological confirmation of the diagnosis.

CARDIOMYOPATHIES Cardiomyopathy is the term used to describe heart muscle disease, and is usually only used when hypertension and coronary artery disease have been excluded. Cardiomyopathy is more frequent in people of African origin - not only in Africa and the Caribbean, but also in the UK. The reasons for this are still unknown. Cardiomyopathies are now classified into three main groups, shown in Table 12.35.

Dilated cardiomyopathy Dilated cardiomyopathies (DCM, also called congestive cardiomyopathy - COCM) comprise half or more of patients diagnosed with cardiomyopathy. The aetiology is usually not identifiable, because an identical clinical histological and pathological picture can arise after damage to the heart from many causes, including toxins (such as cobalt), drugs (such as adriamycin), or following an attack of acute myocarditis. Chronic alcoholism is associated with DCM, and, as alcohol has a direct suppressant effect on the heart, it is thought to be an aetiological agent in many cases. Other causes are given in Table 12.35. Clinical features Presymptomatic DCM may be present for several years without coming to the notice of a doctor unless there is a routine chest X-ray. At this stage there may be some

12

TABLE 1245 Types of cardiomyopathy' Type

Features

Causes

Dilated cardiomyopathy (DCM)

Large cavity, poor contraction. AV valve regurgitation. Low ejection fraction

Hypertrophic (obstructive) cardiomyopathy (HCM or HOCM) Restrictive cardiomyopathy

Idiopathic ventricular wall thickening, often asymmetric. Small cavity. High ejection fraction Thick-walled ventricle or fibrosed endocardium restricts filling, but systolic function (ejection fraction) maintained. Much less common

High percentage unknown (idiopathic) Following myocarditis of any cause Alcohol, cobalt, adriamycin, daunorubicin Systemic illness: amyloid; sarcoid; haemochromatosis Usually familial, may be sporadic, Similar picture may arise in the elderly, secondary to hypertension/aortic stenosis Endomyocardial fibrosis and Lb'ffler's endocarditis Amyloid; sarcoid; haemochromatosis

*Some patients may have features of more than one type of cardiomyopathy, e.g. dilated and restrictive, and individual diseases, such as haemochromatosis, may commonly have features of both.

diminution in maximum effort tolerance, but unless the subject concerned is an athlete, this usually goes unnoticed or is ascribed to normal ageing. Later on, symptoms are dominated by heart failure in severe cases, with both left- and right-sided failure being eventually intractable. AF and thromboembolism are frequent complications and 10% of patients have recurrent chest pain or angina on effort, despite normal coronary arteries. The clinical picture is of a dilated heart that contracts very poorly, with a markedly reduced ejection fraction and a high filling pressure. Physical examination usually reveals cardiomegaly, with a gallop rhythm. Both mitral and tricuspid regurgitation are common when the patient presents in heart failure, and are secondary to the ventricular dilatation. Investigation Chest X-ray reveals the cardiomegaly, which often involves all the chambers of the heart. O The ECG is non-specific and may show features of ventricular hypertrophy, Q waves and widespread T-wave inversion. Echocardiography can be used to exclude a pericardial effusion and haemodynamically significant valve disease. The distinction between cardiomyopathy with mitral regurgitation and heart failure due to valvular disease is important and can be made by echocardiography, which shows vigorous LV contraction when the valvular disease is primary. Management and prognosis When these patients present with heart failure the prognosis is poor, with a 50% mortality in the following 2 years. Treatable causes of heart failure, such as valve disease, must be excluded, and endomyocardial biopsy may be necessary in the young patient to exclude an active myocarditis. Heart failure is treated conventionally (p. 493). Anticoagulation is advised if a high risk of thromboembolism exists.

Hypertrophic cardiomyopathy Hypertrophic cardiomyopathy (HCM) is characterized by marked left ventricular wall thickening and a small left ventricular cavity. Usually, the wall thickening is asymmetric and predominantly affects the interventricular septum; this is known as asymmetrical septal hypertrophy (ASH). The hypertrophy can lead to left ventricular outflow obstruction with abnormal systolic anterior motion (SAM) of the mitral valve; when this occurs, there is almost always a variable degree of mitral regurgitation. The condition is then known as hypertrophic obstructive cardiomyopathy (HOCM). Similar processes can occur, but less frequently, on the right side of the heart, and sometimes both sides of the heart are involved. Aetiology and pathology There is a familial incidence of HCM. It is inherited as an

RECENT ADVANCES IN GENETICS OF HYPERTROPHIC CARDIOMYOPATHY CARDIOMYOPATHY

OF

Hypertrophic cardiomyopathy is a phenotypically and genetically heterogeneous disorder linked to at least four different loci on chromosomes 1, 11, 14 and 15. Three of the disease genes have been identified, encoding for proteins of cardiac |3-myosin heavy chain ([3MHC), cardiac troponin-T and a-tropomyosin. Other loci and genes remain to be identified. In each of these genes, different mutations have been identified which are statistically associated with different disease manifestations and severity. In clinical practice, the identification of these genes and mutations may aid earlier diagnosis and provide prognostic information, complementing information from clinical assessments. However, the implications of these on the treatment and counselling of patients have yet to be evaluated.

571

autosomal dominant, but sporadic examples account for most cases. The condition may be mild and show only on echocardiographic examination. Some cases have associated 'triggers' for left ventricular hypertrophy, such as hypertension or aortic valve disease, but there is usually no obvious cause. On microscopy the muscle fibres are thickened and appear to be arranged haphazardly. The appearances are not specific. One theory suggests that the geometrical arrangement of the fibres is such that they pull antagonistically during systole, and that this results in myofibrillar hypertrophy and disruption. A number of genetic mutations have been identified in families with HCM. Statistically, some mutations are associated with a very high risk of premature arrhythmic deaths, but others are associated with a more benign course. When confirmed, these findings will have fundamental implications for managing the disease and for the screening of affected families. Clinical features Syncope and palpitations from a variety of arrhythmias are common and sudden death may occur. Angina can develop without coronary artery disease. The thickened myocardium can produce an inflow obstruction and lead to heart failure from a raised filling pressure. In later stages there is an enlarged heart and a risk of systemic embolism. The physical signs depend on whether there is outflow obstruction. In its absence there may only be third or fourth heart sounds and a rather muscular feel to the left ventricular impulse. In the presence of outflow obstruction, systolic ejection is shortened, there is a jerky carotid pulse (compared with a sustained and weak pulse in aortic stenosis) and the apex beat has a characteristic presystolic lift. Outflow obstruction produces a harsh late systolic murmur, which may be masked by a pansystolic murmur of coincidental mitral regurgitation. The ejection murmur becomes louder on forced expiration or with the Valsalva manoeuvre (creating a positive intrathoracic pressure), or an inhalation of amylnitrite or administration of an arterial vasodilator or positive inotropic agent (e.g. isoprenaline). It becomes diminished on deep inspiration or with Mueller's manoeuvre (creating a negative intrathoracic pressure), or the administration of an arterial vasoconstrictor, owing to the diminution of the ventriculoarterial pressure gradient. The presence of exertional angina, exertional syncope and a systolic murmur with left ventricular hypertrophy can easily lead to a mistaken diagnosis of aortic stenosis. Although the physical signs differ from those of true aortic stenosis, echocardiography has proved invaluable in establishing the correct diagnosis.

Q MCQ 12.21

572

A significant number of sporadic cases are discovered at autopsy in young people, not infrequently when they have collapsed during physical activity or sport. Investigation Chest X-ray is often unremarkable in the absence of heart failure. The ECG is non-specific, showing left ventricular hypertrophy and, occasionally, Q waves. Echocardiography demonstrates the left ventricular wall thickening, particularly ASH, and, with obstruction, shows SAM of the mitral valve (see above). Two-dimensional echocardiography shows that the septal hypertrophy can be a localized bulge, and also the high ejection fraction. Careful echocardiographic examination is required not to miss isolated apical hypertrophic cardiomyopathy. Doppler ultrasound will demonstrate the mitral regurgitation, as well as the highvelocity jet in late systole that develops in the outflow tract of the left ventricle in the presence of obstruction. Cardiac catheterization and angiography are necessary to confirm normal coronary arteries, to judge the degree of mitral regurgitation and to measure the apex-to-base pressure gradient in the left ventricle. Manoc/emenf Sudden death is a continuing hazard even with antiarrhythmic therapy, and various agents are used to control symptoms of palpitations without apparently preventing sudden death. (3-Blockers can abolish outflow obstruction, and may help in the management of heart failure. Calcium antagonists, such as verapamil, have been used to reduce ventricular hypertrophy, with variable results. Both types of drug slow the heart rate and aid diastolic function. Surgical intervention has been used quite extensively, with mitral valve replacement and outflow tract myomectomy being used to abolish outflow tract obstruction. There is a high surgical mortality and, unless the myomectomy is radical, outflow tract obstruction can recur. The role of surgery therefore remains controversial. More recently, therapeutic embolization of septal branches of the left anterior descending artery has been shown to reduce outflow obstruction, leading to clinical improvement. Screening of siblings of those affected reveals a high incidence of mild degrees of asymptomatic asymmetrical hypertrophy. Investigation and monitoring of these subjects may eventually enable the mortality and morbidity of this condition to be reduced, but may also cause great anxiety in an asymptomatic individual.

Restrictive cardiomyopathies O There is a small subgroup of heart muscle disorders characterized by a very high filling pressure to the ventricles resulting from rigidity of the walls owing to infiltration with abnormal tissue. Reasonable systolic function is usually maintained. Clinically, such restrictive cardiomyopathies present with very similar features to constrictive pericarditis, with evidence of severe heart failure and a relatively small heart.

Cardiac amyloid Cardiac amyloid is the commonest cause of death in primary amyloidosis (Ch. 23, p. 1263). The gastrointestinal tract and tongue are often also involved, as are the nerves and skin. Presentation is either with heart failure secondary to poor systolic function (like a DCM), or with severe right heart failure secondary to a restrictive pattern. There may be evidence of tricuspid and mitral regurgitation with a loud gallop rhythm, and echocardiography shows thickening of the ventricular walls with a small left ventricular cavity. Characteristically, the ventricular walls, infiltrated by amyloid, have a very bright echo appearance, and there may be conduction abnormalities which may require a pacemaker. The diagnosis may be confirmed by rectal or gingival biopsy or, sometimes, by cardiac biopsy. There is no specific treatment for these patients, who fortunately present late in life. Haemochromatosis and transfusion haemosiderosis Overload of the body with iron can damage the heart. Haemochromatosis is a genetic disease in which iron overload occurs (see Ch. 16, p. 868). Transfusion haemosiderosis is caused by repeated blood transfusion (usually more than 100 units) for chronic anaemias, such as thalassaemia, and results in similar tissue loading with iron. The iron damages the liver, heart, gonads and pancreas, causing hepatomegaly, heart failure, gonadal dysfunction and diabetes. There is a characteristic dusky skin pigmentation. Cardiac manifestations are found in the majority of patients. Cardiac involvement does not necessarily parallel other organ involvement, although cardiac failure is always associated with high levels of iron in the heart. There is usually myocardial thickening and dilatation, and so the patient presents with features of both a dilated and a restrictive cardiomyopathy. The diagnosis is based on demonstration of excessive iron in the body, with the serum ferritin being particularly high. Cardiac findings are of heart failure and arrhythmias both supraventricular and ventricular - with disorders of conduction. There may be a low-voltage ECG. The main aim of management should be to reduce body iron levels by chelation or, in the case of haemochromatosis, by repeated venesection.

Cardiac carcinoid In the carcinoid syndrome, where a carcinoid tumour (usually ileal) has metastasized widely, over 50% of patients have cardiac involvement, consisting of endocardial plaques of fibrous tissue. This is almost always on the right side of the heart and involves the tricuspid and pulmonary valves, the vena cava and the coronary sinus, producing distortion and valvular stenosis. Serotonin and kinin peptides produced by the tumour are normally inactivated in the liver and lung; thus the left side of the heart is protected. Clinically, the patient has the systemic manifestations of

carcinoid syndrome: flushing, diarrhoea and, sometimes, bronchoconstriction. In addition, there are symptoms and signs of right heart failure, often with pulmonary stenosis and tricuspid regurgitation. The circulating vasodilators may also produce a hyperkinetic circulation with highoutput heart failure.

12

Sarcoidosis Between 20% and 30% of autopsied cases of generalized sarcoid show cardiac involvement, although clinical manifestations are present in only about 5% of patients. Presentation may be with a cardiomyopathic picture of either the dilated or the restrictive type, although it can also present with conduction abnormalities, heart block and sudden death. Heart block in a young patient may be the first manifestation of sarcoid. Diagnosis and management In the presence of the classic pulmonary findings diagnosis does not require myocardial biopsy, but where there is doubt a biopsy can be extremely helpful. Isotope and magnetic resonance imaging may show infiltration. Treatment is on the whole unsatisfactory. Pacing may reduce the risk of sudden death in patients with conduction system damage, and large doses of corticosteroids are usually advocated in an attempt to prevent progressive myocardial infiltration. Endomyocardial fibrosis Endomyocardial fibrosis is relatively common in regions of Africa, particularly Nigeria and Uganda, where it accounts for more than 20% of cardiac deaths. It is also found in other tropical areas and, occasionally, in Europeans who have been resident in endemic areas. Pathologically, there is gross fibrous thickening of the endocardium of one or both ventricles, usually associated with severe AV valve regurgitation. It commonly presents in young people of both sexes, often from low-income groups. Clinical features The clinical presentation of endomyocardial fibrosis depends on whether one or both ventricles are involved. The disease may be heralded by fever, but usually presents with severe heart failure and sometimes arrhythmias. The heart is not usually markedly enlarged, although pericardial effusion is not uncommon. Clinical signs of valvular regurgitation and heart failure predominate. Systemic and pulmonary emboli may complicate the clinical course. Echocardiography, particularly 2D scanning, may demonstrate the thickened endocardium, which can obliterate parts of the ventricular cavity. Intraventricular clot may also be shown. Surgical excision of the thickened endocardium, with tricuspid and mitral valve replacement, has given some encouraging results in a disease that appears to have, for most, a relentless progression. Endocardial fibroelastosis Endocardial fibroelastosis is a condition of infancy, in

573

which there is gross thickening of the endocardium, with subsequent heart failure. It may present as a primary condition or follow aortic stenosis, other left-sided heart conditions or cardiac surgery in which endocardial perfusion may have been at risk. Loffler's endocarditis Loffler's endocarditis - a syndrome of prolonged eosinophilia with associated localized or widespread eosinophilic infiltrates - involves the heart in over 90% of cases. The hypereosinophilic syndrome is most commonly idiopathic, but may occur with eosinophilic leukaemia or secondarily with Hodgkin's disease, polyarteritis nodosa, tumours and parasitic infestation. The clinical picture is of a severe progressive cardiomyopathy with a restrictive picture, often with systemic embolism, in a patient with a wasting disease and marked eosinophilia. The disease process starts as an eosinophilic myocarditis, and progresses with the formation of multiple endocardial thrombi. In the third stage there is fibrosis and a hyaline membrane forms over the endocardium. The disease thus has some similarities to endomyocardial fibrosis. The clinical course of the disease can be from a few months to 12 years.

This usually occurs in someone with a pre-existing severe disease. Over the last two decades the clinical spectrum of infective endocarditis has changed greatly. This is owing to the gradual decline in the numbers of patients with rheumatic heart disease, an increase in the number of patients who have had cardiac surgery, invasive investigation being used in elderly patients, and the epidemic of intravenous drug abuse. Nowadays, infective endocarditis occurs quite frequently in elderly patients with degenerative disease of the heart valves and in whom there is no evidence of either rheumatic or congenital heart disease. Furthermore, infective endocarditis is not easily divided into the subacute and acute forms. There are a wide range of infective agents of different virulence, and the same organism, such as Staphylococcus aureus, can behave quite differently on different occasions, depending on the patient's host defences and the particular strain of organism. The intravenous drug abuser may develop infective endocarditis with a variety of unusual organisms, sometimes on apparently normal valves and quite commonly on the tricuspid valve. Bacteria are not the only causes of infective endocarditis. Fungal and rickettsial infections, as well as viral infective endocarditis, can occur. For these reasons, the more general term infective endocarditis is preferred to the earlier classification.

INFECTIVE ENDOCARDITIS

Pathogenesis Infective endocarditis is a multisystem disease caused by infection of the heart valves or adjacent endocardium. It is characterized by prolonged fever, weight loss, embolic phenomena and renal failure.

Classification Until recently, it was usual to divide the disease into two groups: • Subacute bacterial endocarditis • Acute bacterial endocarditis.

574

Subacute bacterial endocarditis (SEE) affects relatively young patients with an identifiable pre-existing cardiac abnormality, usually rheumatic valvular heart disease or a congenital abnormality of the heart associated with a jet lesion. The heart valves become infected with an organism of normally low virulence which slowly forms vegetations on the damaged valve and produces progressive valvular regurgitation. This is accompanied by intracardiac abscesses, embolic episodes and a slow, wasting fever, often associated with evidence of formation of immune complexes, leading to glomerulonephritis. Acute bacterial endocarditis occurs in immunocompromised patients or those suffering from a staphylococcal septicaemia. Apparently normal mitral or aortic valves are rapidly destroyed in the course of 24-48 hours, resulting in torrential regurgitation and death from heart failure.

Pre-existing cardiac lesions Endocarditis usually develops in congenitally abnormal or rheumatic hearts in which there is a 'jet' lesion. The altered flow of blood allows bacteria to impact directly on the endocardial surface. Previous damage to the endothelium (e.g. as a result of rheumatic heart disease) facilitates colonization by bacteria present in the bloodstream, which stick to the endocardial lesion and multiply. A chronic lesion, such as one produced by a jet of blood impinging on the endocardium, remains susceptible. However, transient lesions, such as those produced by passing a cardiac catheter or cannula through the heart valves for as little as 15 minutes, heal rapidly and, as a result, rarely become infected. Bacteraemia Transient bacteraemia is common with trauma to the mouth and teeth; the organism most commonly present, Streptococcus viridans, is also very likely to stick to an endocardial lesion and colonize it. Gram-negative bacilli, which can be grown from the blood in Gram-negative infections and at the time of urogenital manipulation and surgery, are less likely to cause endocarditis and stick less readily to endocardial lesions. In practice, any surgical procedure involving manipulation of any organ or part of the body not normally bacteria-free should be regarded as a potential cause of bacteraemia.

Infective endocarditis may occur following a chest or urinary infection or septicaemia. It can also occur following a variety of operative and investigative procedures, but the risk of these procedures producing a bacteraemia is variable. Minor medical procedures that may provoke bacteraemia include the following: • Dental extraction and manipulation. Transient bacteraemia with streptococci and some enterococci lasts from 15 minutes to several hours. Chewing can cause bacteraemia even in edentulous individuals. • Oropharynx. Tonsillectomy, rigid bronchoscopy, orotracheal intubation and nasopharyngeal suction. • Gastrointestinal tract. Barium enema, sigmoidoscopy and colonoscopy. • Urogenital tract. Urinary catheterization, prostatectomy and prostatic massage have all been found to produce a bacteraemia, but normal vaginal parturition does so only occasionally. • Intravenous cannulae. These are a potential source of bacteraemia and should not be used unnecessarily in patients at risk. Growth of the vegetation The vegetation grows as a concretion of bacteria and the reaction of the blood to their presence. The bacteria become covered in platelets and layers of fibrin. The exuberance of the vegetation depends on the organism. Fungal infections with Candida produce large vegetations which can even occlude the orifice of prosthetic valves. Other organisms may provoke so little reaction that the vegetations are too small to be demonstrated by echocardiography (which has a maximum resolution of a few millimetres). Infection can perforate the valve cusps and track into the adjacent myocardium and valve ring, causing abscesses and a variety of complications, depending on the site. These include fistula formation, bundle branch block, complete heart block, infective pericarditis and papillary muscle and chordal rupture. Embolism and immune complex manifestations Embolism of fragments of infective vegetation can produce embolic abscesses in the spleen, kidneys, brain and elsewhere, with 'mycotic' aneurysm of the arterial wall which may result in rupture and sudden haemorrhage. Embolic episodes result in cerebral and peripheral ischaemic attacks, transient haematuria, splenic infarcts and coronary occlusion. Long-standing infection leads to the formation of immune complexes, which may themselves complicate the presentation and course of the disease. These complexes are probably responsible for the glomerulonephritis that may accompany infective endocarditis, and for the skin lesions. Valve involvement The frequency of involvement of the four cardiac valves

is in direct relationship to the mechanical loading on the valves, i.e. mitral, aortic, tricuspid and pulmonary, in descending order of frequency. This is related to the much higher jet velocities usually found on the left side of the heart, although the very low incidence of involvement of the pulmonary valve in pulmonary stenosis remains unexplained. Untreated infective endocarditis of the aortic valve may lead rapidly to infection of the mitral valve, as the regurgitant jet of blood through the aortic valve most commonly strikes the anterior cusp of the mitral valve. Tricuspid valve endocarditis leads to a picture of recurrent lung 'pneumonitis' of a patchy nature caused by embolism of vegetation and organisms into the lung.

12

Infecting organisms Streptococci About 50% of cases of infective endocarditis are caused by cx-haemolytic streptococci from the mouth and oropharynx, formerly classified as Strep, viridans but now subdivided into many organisms, e.g. Strep, mitior and Strep, sanguis. These organisms are most likely to reach the bloodstream after dental extraction or manipulation, bronchoscopy, and tonsillectomy and adenoidectomy. The enterococci of group D streptococci have become more frequent causes and can be found in the gastrointestinal tract, genitourinary tract and periodontal tissue. Accurate identification is sometimes difficult, and the prognosis is worse than for Strep, viridans infection. Strep, bovis, a non-enterococcal organism from group D, is an increasingly common finding in infective endocarditis, and the disease in these cases is very similar to that caused by Strep, viridans. Strep, pneumoniae is a less common agent but still causes some cases of acute infective endocarditis; the infection may be complicated by meningitis. Anaerobic streptococci, organisms that are initially difficult to grow, are found increasingly frequently but necessitate anaerobic cultures. The discovery of Strep, faecalis should lead to investigation of the gut, as colonic carcinoma may be an underlying cause. Staphylococci A staphylococcal septicaemia with Staph. aureus often involves the heart valves; if several positive cultures are obtained, a prolonged course of antistaphylococcal therapy is indicated even without definite evidence of cardiac pathology. Acute endocarditis can damage apparently normal valves very quickly. The same organism can produce a subacute picture. It is a frequent finding in intravenous drug abusers and diabetics with endocarditis. Staph. epidermidis (albus} may be grown as a contaminant of blood cultures, but repeated positive cultures are significant as it is a common cause of infective endocarditis, both in patients with prosthetic valves and in intravenous drug abusers. It is occasionally found in patients with indwelling subcutaneous intravenous lines. Many

575

months may pass between the introduction of the infection and the presentation with endocarditis. Other organisms Neisseria gonorrhoeae and Gram-negative bacilli are relatively uncommon causes of infective endocarditis, but Pseudomonas is becoming more frequent. Q fever endocarditis is uncommon and is diagnosed by finding a rising titre of antibodies in the blood. Fungal endocarditis occurs with prosthetic valves and in patients who have had prolonged periods of antibiotics or corticosteroids, or who are intravenous drug abusers.

Clinical features The clinical features of infective endocarditis can be divided into several categories. Features of chronic infection The features of chronic infection include a remittent pyrexia, myalgia, arthralgia and general fatigue. These symptoms are accompanied by anaemia, anorexia and weight loss. Splenomegaly may be present but is usually not marked. Splenic infarction can lead to pain and a friction rub over the spleen. A splenic abscess can delay resolution of the fever. In long-standing cases cachexia and mild finger-clubbing may occur. Cardiac manifestations Heart murmurs are present in over 85% of cases but may easily be missed, especially in the presence of heart failure. Early diastolic murmurs are notoriously difficult to hear, and a wide pulse pressure is often the first clue that aortic regurgitation is present. Changing heart murmurs are uncommon but, when present, are very suggestive of the diagnosis. Cardiac arrhythmias and conduction abnormalities may follow the spread of infection from the valve ring to form myocardial abscesses. Heart failure can be a presenting symptom. Its presence signifies a poor prognosis and is an indication for early surgery. Sudden heart failure can occur with rupture of a valve cusp or chordae tendinae, or sinus of Valsalva aneurysm. Systemic embolism Embolism is a prominent feature of infective endocarditis and the presentation may be with a cerebrovascular event (e.g. a dense hemiplegia), an ischaemic limb, mesenteric embolus or haematuria from renal infarction. Over 50% of cases exhibit embolic phenomena: fungal endocarditis with large friable vegetations is notorious for producing large emboli. Right-sided endocarditis usually presents with recurrent patchy consolidation (pneumonitis).

Fig.

576

12.25

Fig.

12.26

SUMMARY 10 Clinical features of infective endocarditis Infection

Fever, anorexia, weight loss, splenomegaly, cachexia, clubbing. Also arthralgia, bone pain

Cardiac

(History of cardiac murmur, bacteraemic procedure) New heart murmur, increasing valvular regurgitation, conduction disorders

Systemic embolism (about 50%) Arterial Pulmonary

Immunological Renal Skin

CVA, mesenteric, renal Infective emboli cause abscesses and arterial rupture. Recurrent pneumonitis Glomerulonephritis leading to renal failure Nails - splinter haemorrhages Conjunctival and buccal mucous membranes - petechiae Finger and toe tips - Osier nodes Thenar and hypothenar eminences - Osier nodes and Janeway lesions Retina - Roth spots and boatshaped haemorrhages

Immunological phenomena Renal A diffuse glomerulonephritis is present in many patients and is due to the deposition of immune complexes. The renal disease is shown by proteinuria and microscopic haematuria. In untreated endocarditis the blood urea begins to rise, but renal failure develops in less than 10% of cases. Even when advanced, renal failure recovers with adequate treatment of the infective endocarditis. Skin and mucous membranes Typical skin manifestations of infective endocarditis were much more common in the pre-antibiotic era, but are now seen in only 20^0% of cases. Thorough examination of the patient's skin, the mucosae of the oropharynx, conjunctivae and retinae should be repeated during the course of the disease, to check for new lesions. Petechial haemorrhages. Examination of the soft palate, pharynx, conjunctivae, hands, feet and anterior body wall may reveal petechial haemorrhages, often with a pale centre. Subungual or 'splinter haemorrhages' are usually seen as a faint dark line away from the nail margin in the fingers and, occasionally, the toes. O They may also be seen in patients who are debilitated for other reasons, or after trauma. When present, they occur in 'bursts' and indicate disease activity.

Osier nodes are painful, nodular, red to purple lesions that usually form on the fingertips 0 but can also occur on the backs of the toes, soles of the feet, the thenar and hypothenar eminences and, occasionally, elsewhere. Formerly common, they are now seen in less than a fifth of cases. The lesions can last a few hours to a few days. Janeway lesions are painless, 1-4 mm erythematous macules which are most common on the thenar and hypothenar eminences and the soles of the feet, and which blanch on pressure. Retinal lesions. The Roth spot is a 'cottonwool exudate' consisting of a perivascular collection of lymphocytes. It is not specific to infective endocarditis. Boat-shaped haemorrhages with a pale centre and white spots have also been described. All of these retinal and cutaneous lesions have been ascribed to 'septic emboli', but are now thought to be immunological manifestations of the disease.

Diagnosis Infective endocarditis is a possible diagnosis in any patient with a fever and a heart murmur, and is established by taking blood cultures. A patient with a heart murmur may have an intercurrent viral infection or some other cause for fever. Where the source of the fever is not clear, it is essential to take blood cultures before starting antibiotics, should they be considered necessary. This is especially important when a patient at risk of endocarditis develops a fever following a procedure that has not been covered with antibiotic prophylaxis. Any febrile patient who has had open heart surgery (other than coronary bypass grafting) should be considered at risk from infective endocarditis, even if no murmur is evident. In the elderly patient, endocarditis can exist without significant fever. Other signs of chronic infection, such as unexplained anaemia and high ESR in a patient with a heart murmur, should arouse suspicion and lead to blood cultures being taken. Echocardiography is an essental investigation in any patient with suspected endocarditis. It will demonstrate vegetations greater than 1 mm and document valvular and ventricular function. Blood cultures Blood cultures should be taken on six to eight occasions over the first 48 hours. It is not necessary to wait for peaks of fever. Anaerobic cultures should be taken and, although a higher percentage of positive cultures has been claimed for arterial blood culture, venous sampling is usually sufficient. Close liaison with the laboratory is essential so that any organism that grows is kept in subcultures and not discarded. Appropriate culture techniques can then be instituted for poorly growing organisms. Where the clinical evidence of the disease is overwhelming, especially if there is evidence of deteriorating valve function, antibiotic therapy may be started as soon as four to six cultures have been taken and before positive

cultures are reported by the laboratory. There is, however, an inevitable risk that this may interfere with identification of the organism if the initial cultures are negative. Where infective endocarditis is a high probability, and particularly if Strep, viridans is grown from blood culture, a dental opinion on possible septic foci in the mouth should be obtained. It is then possible to time any dental extraction or other dental work to follow by 1 hour or so the initiation of antibiotic therapy. This then removes the possibility of septic foci in the mouth harbouring antibiotic-resistant organisms which can then superinfect the heart after the initial organisms have been eradicated.

12

Culture-negative endocarditis In 20% of cases no organisms are grown, despite many other features of the disease being present. This may occur because the patient has been treated with an antibiotic (which may render cultures negative for days or weeks after stopping treatment), because the organism involved is present in very small numbers, or because it is difficult to culture. Even in average cases of endocarditis there may be only 100 organisms per mL of blood and, particularly in long-standing cases with renal failure, the patient's antibodies may render the blood almost sterile, despite the presence of organisms in large vegetations. Repeated blood culture samples of l0mL should be inoculated into 100 mL of culture broth and incubated aerobically and anaerobically. The organisms may require special nutrients and may not grow for several days, or, in a few cases, for several weeks. If repeated cultures are negative, or the patient's condition is deteriorating, it may be necessary to embark on treatment without positive cultures. This course of action must only be followed after close consultation with a microbiologist, who may suggest unusual sampling techniques (occasionally bone marrow or urine is positive when blood is not), and after blood has been taken for fungal and rickettsial antibodies.

Management Successful treatment of infective endocarditis should involve close cooperation between physician and microbiologist. In uncomplicated cases 4-6 weeks of bactericidal antibiotic therapy is required, with serum levels of antibiotic sufficient to penetrate the vegetations and abscesses and eradicate the organism. Bacteriostatic antibiotics and bacteriostatic levels of bactericidal antibiotics are not suf-

SUMMARY 11 Features suggestive of infective endocarditis • • • • •

Fever Evidence of chronic infection A heart murmur suggestive of a jet lesion Positive blood culture on more than one occasion Presence of vegetations on echocardiography

577

ficient, even with the host's natural defences, to eradicate the organism. Antibiotic therapy Although oral antibiotic therapy can be used for sensitive Strep, viridans infection, the intravenous route is preferred for initiating therapy and while awaiting 'back titrations'. For intravenous therapy a central venous catheter may overcome the problems of superficial venous cannulae. All cannulae must be inserted under strict aseptic conditions and kept scrupulously clean to prevent the risk of superinfection with fungi. They should be changed at least every third day. After culture and identification of the organism, its antibiotic sensitivity must be urgently assessed. The bacteriology laboratory assists by monitoring the patient's blood levels of antibiotic and titrating to determine what serum dilutions are bactericidal to the organism in culture. A level of 1 in 8 is considered satisfactory. It is essential that the antibiotic regimen for an acute septicaemic endocarditis include antistaphylococcal antibiotics. Where a subacute infection is being treated it is usual to use wide-spectrum antibiotics, such as gentamicin and ampicillin. The response to treatment is then monitored and the treatment changed or continued, depending on whether it appears to be controlling the infection. Culturenegative endocarditis, particularly with large vegetations or other indication of severe valvular involvement, may be an additional criterion for early surgical intervention, particularly if potentially hazardous antibiotics are being used. Monitoring treatment Successful antibiotic treatment ideally results in disappearance of the fever within a few days, a fall in the ESR, a rise in the haemoglobin and an increase in the patient's weight and wellbeing. Measurement of C-reactive protein may give a more reliable guide to response than the ESR. Evidence of microemboli, such as fresh splinter haemorrhages and microscopic haematuria, take longer to go, and large vegetations take several weeks to regress. Occasionally, the fever and high ESR do not settle, although repeated blood cultures remain negative. If other causes for persistent fever (e.g. splenic abscess or superinfection) are eliminated, then the fever may be antibiotic induced and disappear with cessation of antibiotic treatment. Surgery Severe valve failure and uncontrolled infection may require emergency valve replacement. If the patient presents with heart failure early surgery is essential, as even with rapid successful antibiotic treatment of endocarditis, damaged valves shrink and fibrose with progressive increase in the degree of valve regurgitation, and the heart

MCQ 12.22

578

Case 12.7

failure may become irreversible. Prosthetic valve endocarditis almost invariably requires valve replacement for the infection to be eliminated. A few days of antibiotic therapy prior to surgery reduces the risk of valve dehiscence from sewing the new valve into an infected valve ring, and reduces the risk of prosthetic valve endocarditis. A heavily infected valve with large vegetations may be electively replaced where there is risk of major embolism, or where the patient cannot be given effective bactericidal therapy. Tricuspid valve endocarditis with large vegetations can be treated by excision of the tricuspid valve without replacement, while the course of antibiotics is given. The prosthetic valve can then be sewn into a sterile ring. Apart from valve replacement, surgical closure of intracardiac fistulas, such as aortic-to-right-atrial shunts, and limited thoracotomy for epicardial pacing for heart block are occasionally required. Prophylaxis The recommended regimens for endocarditis prophylaxis vary slightly, but all are based on the use of a bactericidal antibiotic to cover a procedure likely to produce bacteraemia. If a general anaesthetic is involved, the antibiotic is given intravenously or intramuscularly. If no general anaesthetic is involved, oral preparations are given if possible. Regimens are shown in Table 12.36. A safe general rule is that prophylaxis should be given for all dental work where the teeth are manipulated, all pelvic, colonic and bladder instrumentation, and surgery. Normal parturition does not need routine cover unless the bladder is catheterized or forceps used. OO

SYPHILITIC ENDOCARDITIS Pathogenesis Syphilis is now much less common as a cause of heart disease. Cardiac involvement is a manifestation of the tertiary stage of the disease, which often follows a prolonged quiescent period. An endarteritis obliterans of the vasa vasorum of the aorta weakens the media, which undergoes a slow necrosis and causes thickening and scar formation in the intima. The changes are irregular, and complicated by fibrosis and calcification. Dilatation of the aorta occurs and can be localized, giving a fusiform or saccular aneurysm; it often also involves the valve ring, causing progressive, and often severe, aortic regurgitation. There may be damage to the aortic valve cusps leading to retraction, but never stenosis. Intimal scarring of the aorta can lead to coronary ostial stenosis. The changes do not always lead to overt heart disease. There is a higher percentage of cases of aortic regurgitation and aortic aneurysms in manual labourers with aortitis, than in sedentary men and women with necropsy evidence of aortitis. Thus, it may be that physical activity produces additional loads on the aorta, which induce these changes.

TABLE 12.36 Endocarditis prophylaxis Procedure

Recommendation

1. Dental procedure under local anaesthetic 2. Dental procedure under general anaesthetic 3. High risk - Refer to hospital • Patients with prosthetic valves • Patients allegic to penicillin or who have had penicillin more than once in the previous month • Patients who have had a previous attack of endocarditis 4. Surgery - upper respiratory tract

A

5. Surgery - genitourinary 6. Obstetrics and gynaecological patients 7. Gastrointestinal procedures

Penicillin allergy-B or C D or E or F

G

Pencillinallegy-H H

G

Penicillin allergy-H As 1 and 2 above Any postoperative antibiotic may have to be given intramuscularly or intravenously if swallowing painful Sterile urine, as 1 and 2 Infected urine - prophylaxis should also cover pathogens involved Cover only for patients with prosthetic valves - D or H Cover only for patients with prosthetic valves - D or H

A: Oral amoxicillin 3 g single dose taken under supervision 1 hour before procedure 1 Children under 10- /2dose Children under 5 - V2 dose

B: Oral erythromycin 1.5g under supervision 1-2 hours before procedure Children under 10- 1/2 dose Children under 5 - 1/4 dose

C: Oral clindamycin 600 mg single dose taken under supervision 1 hour before procedure Children under 10 - 6mg/kg body weight single dose taken under supervision 1 hour before procedure

D: Intramuscular amoxicillin 1g in 2.5 ml 1% lidocaine hydrochloride just before induction plus 0.5 by mouth 6 hours later Children under 10 -1/2 dose

E: Oral amoxicillin 3g dose 4 hours before anaesthesia followed by a further 3g by mouth as soon as possible after operation Children under 10- 1/2 dose Children under 5 - Y4 dose

F: Oral amoxicillin and probenecid Amoxicillin 3 g together with probenecid 1 g 4 hours before operation

G: Intravenous amoxicillin and gentamicin 1 g amoxicillin plus 120mg gentamicin intravenously just before induction, then 0.5 g amoxicillin orally 5 hours later Children under 10 - amoxicillin Ys dose, gentamicin 2mg/kg body weight

H: Intravenous vancomycin 1 g slow infusion over 1 hour before procedure followed by gentamicin 120mg intravenously just before induction Children under 10 - vancomycin 20mg/kg, gentamicin 2mg/kg

Clinical features The patient is usually a man (there is a 3:1 sex ratio for uncomplicated aortitis, but this rises to 10:1 for aortic aneurysm) in late middle age.

12

Aortic regurgitation The presentation may be with pure aortic regurgitation with evidence (best shown echocardiographically) of a dilated aortic valve ring, and serological proof of previous Treponema pallidum infection. If untreated, the regurgitation tends to be progressive and severe; indeed, the classic 19th century physical signs described for aortic regurgitation are almost all based on patients with severe symphilitic aortic regurgitation (waterhammer pulse, head rocking etc.). With a dilated aortic root, the early diastolic murmur is louder in the third right interspace than the third left interspace. If there is coronary ostial stenosis severe angina is usually present, often leading to status anginosus (continuous pain), heart failure and death, unless there is early surgical intervention. Aortic aneurysm Aortic aneurysm is much less common than regurgitation. It usually occurs in the ascending aorta or arch rather than the descending aorta, where aneurysms are much more likely to be arteriosclerotic. The clinical features depend on the position of the aneurysm. Classically, the ascending aortic aneurysm produces signs of aortic regurgitation, the arch aneurysm symptoms such as dysphagia, cough, pain and dyspnoea, and the descending aneurysm pain from erosion of the spine. Diagnosis and investigation Serological proof of previous spirochaetal infection is usually followed by CSF examination because of the close association with neurosyphilis. Positive serology cannot be taken as proof of syphilitic infection in patients with yaws scars, who often have persistent positive serology even after treatment. However, the combination of aortic dilatation and positive serology would be taken as confirmatory. Arteriosclerotic aneurysms, cystic medial necrosis of the aorta and aortic dissection are now common differential diagnoses. Aortic regurgitation is investigated with particular emphasis on excluding coronary ostial stenosis. Aortic aneurysm is now often investigated by CT scanning. There is often laminated clot in the aneurysm, and classically, in aortitis there is calcium seen in the ascending aorta. The treatment of syphilis is described on page 461. The Herxheimer reaction is particularly dangerous in aortic regurgitation with coronary ostial stenosis. The regurgitation may progress after effective penicillin treatment, owing to further fibrosis. Surgical treatment is as for other causes of aortic regurgitation or aortic aneurysm, after effective chemotherapy.

579

FURTHER READING ON INFECTIVE ENDOCARDITIS Dajani A S, Bisno A L, Chung K J et al 1991 Prevention of bacterial endocarditis: A statement for health professionals from the committee on rheumatic fever, endocarditis, and kawasaki disease of the council on cardiovascular disease in the young, the American Heart Association. Circulation 83:1174-1178. Endocarditis Working Party of the British Society for Antimicrobial Chemotherapy 1990 Antibiotic prophylaxis of infective endocarditis. Lancet 335: 88-89.

NON-INFECTIVE ENDOCARDITIS RHEUMATIC FEVER AND RHEUMATIC HEART DISEASE Rheumatic heart disease is now uncommon in western Europe and North America, but is an important historical cause of what was once the bulk of valvular heart disease. It is still important in the developing world as a disease of childhood and a major cause of heart disease. Its decline in western countries started before the widespread use of antibiotics, and probably reflects increased standards of nutrition and hygiene and, possibly, a decline in the virulence and prevalence of Lancefield group A haemolytic streptococci. Rheumatic fever and chorea are the exclusive causes of chronic rheumatic heart disease, in which there is interstitial muscle fibrosis as well as the heart valve disease. However, some 50% of patients with chronic rheumatic heart disease give no history of either condition. Rheumatic fever develops in about 3% of young people exposed to an epidemic strain of (3-haemolytic Lancefield group A streptococcus. These patients usually have pharyngitis, but may have scarlet fever. After an apparent recovery they develop fever, sweats, skin rashes, arthritis and, sometimes, cardiac involvement from what appears to be an autoimmune disease starting 2-4 weeks after the initial sore throat and lasting 6-8 weeks, but occasionally as long as 26 weeks. Some patients give no definite history of an antecedent sore throat. Following the attack of rheumatic fever, some 70% of patients will eventually have evidence of rheumatic valvular heart disease, although the time interval can be as long as 50 years. On the other hand, the valve disease may become manifest during the attack of rheumatic fever, especially where regurgitation develops. Rheumatic fever may be recurrent, and the chances of valve disease increase with the number of attacks.

Pathology

580

Although rheumatic fever follows a streptococcal infection, no bacteria are found in the lesions; however, they may still be present in the pharynx. There is an acute inflammatory reaction, with oedema in the affected tissues,

and then granulomatous lesions develop in the myocardium (Aschoff nodules). Oedema of the valve cusps is followed by verrucous fibrin deposits developing along the lines of apposition of the valve cusps, which eventually adhere at the commissures. In the acute illness there is an acute non-bacterial synovitis in the joints, and granulomata may develop subcutaneously on extensor surfaces of the limbs and over the Achilles tendons (rheumatic nodules). There is usually a high antistreptolysin O titre, and other antistreptococcal titres may also be high. It is thought that cross-reactivity between the streptococcal antigens and host tissue antigens leads to an autoimmune inflammatory process.

Clinical features and diagnosis The diagnosis of rheumatic fever depends on criteria proposed by Duckett-Jones (Table 12.37); two major criteria, or one major and two minor criteria, establish a high probability for the diagnosis. The patient is almost always between the ages of 3 and 30 years (usually 5-15); drenching sweats and a prolonged feverish illness occur, with extremely painful arthritis that moves from joint to joint (usually involving large joints such as the knee and elbow), which become hot, red, swollen and extremely painful for a few days and then recover fully. The flitting arthritis is characteristic, as is the classic skin rash, erythema marginatum. Subcutaneous nodules may develop after several weeks, usually over the elbows. Carditis may be pericarditis, myocarditis or, more importantly for the eventual appearance of valve disease, endocarditis. All may occur at once as pancarditis. The pericarditis is similar to any acute pericarditis (p. 586), with precordial pain, ECG changes (raised ST segments) and often an increase in cardiac silhouette on chest X-ray. Myocarditis is manifest by a tachycardia disproportionate for the degree of pyrexia, gallop rhythm and heart failure; there are also ECG changes, such as prolonged PR interval or greater degrees of AV block and T-wave flattening or inversion, and raised serum creatine phosphokinase AV. Death occasionally occurs from myocarditis. TABLE 12.37 Criteria for the diagnosis of rheumatic fever Major criteria

Minor criteria

• • • • •

• • • • •

Polyarthritis Carditis Subcutaneous nodules Chorea Erythema marginatum

Fever Arthralgia Prolonged PR interval on ECG Raised ESR Raised white blood count or C-reactive protein • Preceding p-haemolytic streptococcus infection • Previous rheumatic fever or inactive heart disease

Endocarditis is indicated by transient or changing murmurs. The classic murmur, which is specific for the condition, is the Carey Coombs murmur; this is a short rumbling diastolic murmur indicating mitral valvulitis. Sydenham's chorea Sydenham's chorea (St Vitus' dance) is usually a separate disease entity but, like rheumatic fever, appears to be related to recent p-haemolytic streptococcal infection and an autoimmune process; occasionally it accompanies the rheumatic fever, although it may follow some weeks later. It is characterized by sudden, jerky, purposeless movements of the face and limbs, such that the child is thought to be clumsy and grimacing. Pure chorea, even when unaccompanied by other evidence of rheumatic fever, is followed by a high incidence of rheumatic valve disease, indistinguishable from that following conventional rheumatic fever.

Investigation Laboratory investigations show a raised ESR, C-reactive protein and antistreptolysin O titre. Blood cultures are negative. Throat swab may be positive for p-haemolytic streptococcus, if the preceding pharyngitis has not been treated with antibiotic.

Differential diagnosis As there is no absolute diagnostic test it is extremely common to find patients who have been diagnosed as having rheumatic fever on weak evidence, such as a childhood fever with a few aches and pains and the presence of a systolic murmur. In one study some 50% of patients with mild congenital heart disease claimed to have had rheumatic fever, suggesting that a child with a heart murmur alone is very likely to be diagnosed as having rheumatic fever during the course of a febrile illness.

Management Aspirin is traditionally used to control the arthritis and fever; it is used in maximum doses. Where the patient is extremely ill, corticosteroids are commonly used, but there is no evidence that these reduce the incidence of endocarditis or subsequent development of rheumatic valvular disease. Prophylaxis Anyone who has had acute rheumatic fever is at risk of recurrence and is usually maintained on phenoxymethylpenicillin (250 mg orally daily, or an i.m. depot preparation of benzathine penicillin) or, if allergic to penicillin, oral sulphonamide, either until the age of 21 or for 5 years after the last attack of rheumatic fever, whichever comes later.

LIBMAN-SACKS ENDOCARDITIS

12

SLE (p. 1172) may lead to heart failure from hypertension and a myopathic process affecting the left ventricle. Involvement of the heart valves by a verrucous endocarditis is described as involving a significant proportion of postmortem specimens; however, valve disease is a clinical problem in only a few long-standing cases of SLE. The warty outgrowths occur at, and away from, the free edges of the valve cusps and over the mural endocardium. They involve the mitral valve in particular, and may lead to thickening of the chordae and subvalvular apparatus. Mitral regurgitation and aortic regurgitation are said to be the commonest lesions, but mixed valve lesions occur. Aortic stenosis has also been described in SLE. LibmanSacks endocarditis may become the site of infective endocarditis and, if haemodynamically important, is treated by prosthetic valve replacement.

RHEUMATOID ARTHRITIS A small percentage of patients with very active rheumatoid arthritis develop mitral and aortic valve disease, with exuberant nodular granulomatous lesions on the valve cusps and at other sites throughout the body. These lesions only occasionally cause significant valve disease, but very occasionally may cause significant stenosis or regurgitation in a patient who has very exuberant and severe rheumatoid arthritis.

REITER'S DISEASE AND ANKYLOSING SPONDYLITIS Reiter's disease and ankylosing spondylitis (p. 1148) can produce a degree of aortitis and aortic root dilatation and cusp retraction similar to that seen in syphilitic aortitis. There may be associated conduction abnormalities as well as aortic regurgitation, and occasionally these patients require aortic valve replacement and cardiac pacemakers.

CARDIAC TUMOURS Cardiac tumours (Table 12.38) are rare. They can present a variety of clinical features and mimic many other cardiac and systemic diseases. Benign tumours comprise 75% of primary cardiac tumours.

MYXOMAS The origin of myxomas is controversial. They appear as a gelatinous pedunculated mass, usually arising from the

581

TABLE 12.38 Most frequent cardiac tumours Tumour

Frequency (%)

Prognosis

Myxoma Lipoma Papillary fibroelastoma Rhabdomyoma Angiosarcoma Rhabdomyosarcoma Other (including secondary)

30 10 10 8.5 9 6 26.5

Benign Benign Benign Benign Malignant Malignant -

Syncope and sudden death are also recorded, and recurrent emboli can lead to severe pulmonary hypertension and a paradoxical embolization to the left side through a patent foramen ovale. Constitutional symptoms are as common as on the left. Physical signs are more frequent than on the left side, with systolic and diastolic murmurs being audible in over 80%, and endocardial friction rub audible in 20%. Myxomas may arise in the ventricles, but these are less common. As with atrial myxomas, they are likely to embolize and cause obstruction.

Investigation fossa ovalis in the atria; 75% occur on the left, but occasionally the tumour invades both atria. About 75% are found in women. There is a slight familial tendency, and current theory suggests that myxomas arise from embryonic cells in the septum. Clinical presentation is in late middle age, when they reach sufficient size. The clinical features are haemodynamic obstruction, embolization and constitutional symptoms.

Left atrial myxoma In left atrial myxoma the large gelatinous mass obstructs the mitral valve and simulates mitral stenosis, producing left heart failure. A ball-valve mechanism can produce recurrent pulmonary oedema or syncope, which can lead to sudden death. Pulmonary hypertension with subsequent right heart failure mimics mitral valve disease. In one-third of patients there are systemic emboli, the brain being the most frequent site. Systemic symptoms are found in 90% of patients and mimic those of a collagen disease or infective endocarditis, with fever, anorexia, weight loss, myalgia, arthritis, Raynaud's phenomenon, petechiae and fingerclubbing. These features are thought to result from immunological reactions to necrosing tumour. Physical signs can mimic those of mitral stenosis, with a diastolic murmur in about 20% of patients. More common is a systolic murmur of mitral regurgitation and, classically, there is a 'tumour plop' instead of the opening snap of mitral stenosis. However, the tumour can be silent or give only intermittent ausculatory findings.

Right atrial myxoma Right atrial myxomas are less common than those on the left side; they can cause severe right heart failure, with oedema, hepatomegaly, ascites and pleural effusion.

Q Fig. 12.27

582

Echocardiography has revolutionized the diagnosis of myxoma. Q The tumour can often be visualized popping backwards and forwards through the valve orifice. However, it can be missed, and a study from several windows may be needed to show small tumours and, particularly, right atrial myxomas. If satisfactory echo pictures are available there is now nothing to be gained from cardiac catheterization, which runs the risk of dislodging tumour and causing systemic emboli. Laboratory tests are frequently abnormal, especially in patients with systemic symptoms. Haemoglobin is low (with normochromic normocytic film), but may occasionally be high with right atrial myxoma, severe pulmonary hypertension and right-to-left shunting through a patent foramen ovale. A neutrophil leukocytosis occurs. The ESR is often raised, with an increase in plasma gammaglobulins.

Management Surgical excision of the tumour is essential but carries some risk of tumour embolization and recurrence, both from the site of excision, if incomplete, and from other sites in the heart. The tumour is therefore excised with its atrial attachment, and the atrial septum then patched. Many surgeons now inspect all four cardiac chambers, because of the high incidence of local recurrence and the occasional multiple tumour. Atrial myxomas are relatively easily treated and should be suspected wherever there is unexplained systemic embolism, particularly if there are features of a 'collagen disease', if a case of 'infective endocarditis' is persistently culture negative, and if features of mitral valve disease are accompanied by syncopal episodes. Right atrial myxoma should be suspected in unexplained right heart failure and recurrent pulmonary emboli.

CARDIAC DISEASE AND PREGNANCY Pregnancy places a sustained load on the heart which increases throughout the pregnancy and then peaks with labour. There is a considerable increase in cardiac output (30-50%), blood volume (40-50%) and extracellular

volume (about 7L), with an increase in heart rate of about lObpm and a slight fall in mean blood pressure. The circulation of the pregnant woman is as if she were vasodilated, with warm peripheries and an increase of up to 30% in resting oxygen uptake. The increase in cardiac output is accomplished by an increase in heart rate, stroke volume and ejection fraction. Venous pressure in the upper body increases marginally, if at all, whereas there can be an increase in lower body venous pressure when supine or standing, due in part to pressure of the gravid uterus on the inferior vena cava. This pressure increase, and the increase of up to 8.5 L in total body water, often leads to oedema of the ankles. The increase in plasma volume occurs earlier than a compensatory increase in total red blood cell volume, so that an apparent anaemia ensues.

Diagnosis Pregnancy makes the diagnosis of heart disease more difficult. The pregnancy may produce symptoms of fatigue, shortness of breath, ankle oedema and syncope. Physical examination may reveal flow murmurs, a mammary souffle, venous hum, a third heart sound, a raised a wave in the jugular venous pulse, displacement of the cardiac apex (by the elevated diaphragm) and basal rales (which clear on coughing) for the same reason, all of which may be taken to indicate heart disease. The above confounding symptoms and signs are unfortunate because of the increased risk posed by heart disease to both the mother and the fetus. Another reason for establishing whether heart disease is present, even if it appears to be asymptomatic, is the small risk of puerperal endocarditis. The murmur of pregnancy (Still's murmur} is a low-pitched, often vibratory murmur heard at the left sternal edge but usually loudest in the pulmonary area. It is extremely common, midsystolic and unaccompanied by any other signs of heart disease. High-pitched systolic murmurs, and all diastolic murmurs, indicate heart disease. The use of echocardiography and Doppler ultrasound will often be of great value in these cases. Chest X-ray is relatively contraindicated. If essential, the fetus can be shielded. The risk to the fetus late in pregnancy is much smaller than first thought. The ECG is usually normal but the axis may shift to the left, with T-wave inversion in lead 3, a small Q wave, and sometimes even ST and T-wave inversion in the limb and precordial leads; these ECG changes revert to normal at the end of pregnancy and recur in later pregnancies.

Management If the pregnant mother is known to have heart disease and is in functional class 3 or worse of the New York Heart Association classification (p. 468, Table 12.1), it is hazardous for the pregnancy to continue. Certain types of heart disease, e.g. coarctation of the aorta, severe hyper-

tension, severe aortic or mitral stenosis or cardiomyopathy, present particular hazards. Cyanotic congenital heart disease poses great risk to the mother if pulmonary hypertension is severe. Spontaneous abortion is frequent. Surgery during pregnancy may offer a solution to the mother who is anxious to continue with the pregnancy; closed mitral valvotomy, or open valvotomy and valve replacement under cardiac bypass, have been performed with a low maternal mortality and acceptable fetal risk. If valvular disease or a jet lesion (VSD etc.) is diagnosed, appropriate antibiotic cover is given at the time of labour, particularly if instrumentation (e.g. urinary catheterization) is used at the time of delivery.

12

Anticoagulants A particular problem may be the need for anticoagulants. Discontinuation of anticoagulants is dangerous for a mother with a prosthetic heart valve, and possibly also for one who has had a pulmonary embolus. Warfarin is avoided during the first trimester, because of an increased risk of fetal abnormality. For a planned pregnancy, heparin can be given during the first trimester. Warfarin is then given until heparin is used to cover delivery by caesarean section.

Peripartum cardiomyopathy Pregnancy and labour may make occult heart disease manifest. Occasionally, catastrophic heart failure supervenes and appears to be associated with a progressive cardiomyopathy.

Hypertension Hypertension in pregnancy and pre-eclamptic toxaemia and eclampsia are discussed on page 592.

SECONDARY HEART DISEASE A wide variety of pathological conditions may also increase cardiac output and produce heart failure. These include thyrotoxicosis, anaemia, cor pulmonale, beriberi, arteriovenous fistulas, generalized Paget's disease of bone and exfoliative dermatitis.

Thyrotoxicosis Thyrotoxicosis dramatically increases metabolic rate, but the thyroid hormones also have a stimulant effect on the autonomic nervous system and the heart itself. There is usually a sinus tachycardia which is indicative of the severity of the disease, but in 15% of cases thyrotoxicosis triggers a supraventricular arrhythmia, most frequently AF. Pulse pressure and blood pressure rise slightly and the heart contracts more vigorously in thyrotoxicosis.

583

Anaemia Severe anaemia makes a twofold demand on the heart. In order to supply the oxygen requirements of the body, the heart has to increase cardiac output, despite itself being perfused with anaemic blood. Mild anaemia is of marginal importance, but life-threatening anaemia may lead to cardiac dilatation and severe heart failure, with pathological changes in the myocardium of 'fatty heart'. As the myocardium extracts virtually all the oxygen from the blood in the coronary circulation, autoregulation of flow demands a dramatic increase in coronary flow to compensate for the extra myocardial work of severe anaemia. This cannot be accommodated if there is coronary disease of any severity, and continuous angina and heart failure may result.

TABLE 12.39 Inherited connective tissue diseases and the heart Condition

Cardiac condition

General features

Cutis laxa

Peripheral pulmonary stenosis Mitral regurgitation Arterial rupture Aortic dilatation, dissection, aortic and mitral regurgitation

General skin laxity, disrupted elastic fibres Hypermobile joints Hyperelastic skin Gracile habitus, arachnodactyly, lens subluxation, sternal abnormalities Fragile bones, blue sclera

Ehlers-Danlos Marfan's

Osteogenesis imperfecta Pseudoxanthoma elasticum

Aortic regurgitation Peripheral and coronary artery disease

Beriberi Beriberi is uncommon outside Asia, but minor forms may occur in some alcoholics and food faddists. Severe thiamine deficiency (classically due to subsistence on polished rice) results in autonomic and other nervous system damage, with profound vasodilatation and a high cardiac output. In its fulminant form the patient may die from pulmonary oedema within 48 hours of the onset of symptoms. Response to conventional heart failure therapy without thiamine is poor. Thiamine is given as soon as the diagnosis is suspected, and should be given to alcoholics in heart failure and in all cases of resistant heart failure where the nutritional background of the patient is doubtful.

Arteriovenous fistulas, Paget's disease of bone, exfoliative dermatitis If they are severe enough, arteriovenous fistulas, Paget's disease of bone and exfoliative dermatitis may all cause a considerable reduction in peripheral resistance leading to a large increase in resting cardiac output. The patient has a bounding pulse and hyperkinetic circulation, often with a third heart sound and systolic murmur. Heart failure may supervene if the heart is already damaged or if the stress is severe enough. Treatment of the condition will alleviate the strain on the heart.

Inborn errors of metabolism and connective tissue disorders Heart disease may complicate or be part of many other medical conditions (Tables 12.39-12.42). Marfan's syndrome is discussed in Chapter 3, page 71. (see also p. 601). O Aortic dissection is a frequent and lifethreatening complication.

Degeneration of elastin fibres leading to yellow pseudoxanthomas of skin, retinal angioid streaks, hypertension

TABLE 12.40 Inborn errors of metabolism and the heart Condition

Cardiac condition

Pompe's disease

Glycogen storage disease involving the heart muscle Aortic and pulmonary artery dilatation Multivalvular disease, coronary artery disease and cardiomyopathy

Homocystinuria Mucopolysaccharidoses: Hurler's and Hunter's syndromes (mental deficiency) Morquio's disease

Aortic regurgitation

TABLE 12.41 Chromosomal abnormalities and the heart Chromosomal defect

Cardiac abnormality

Trisomy 21 (Down's syndrome)

Endocardial cushion defects, ASD, VSD, pulmonary stenosis VSD, double-outlet right ventricle VSD, persistent ductus arteriosus, pulmonary stenosis VSD Coarctation Persistent ductus arteriosus

Trisomy 13 Trisomy 18 Cri du chat (short arm deletion 5) XO (Turner's syndrome) XXXY and XXXXX

ASD = atrial septal defect; VSD = ventricular septal defect

Athlete's heart O Fig. 12.28 584

Exercise training induces changes in the heart muscle similar to those in skeletal muscle. Power athletes who perform mostly isometric exercises (weight lifting, wrestling, tug-of-war) develop ventricular hypertrophy. The left ven-

TABLE 12.42 Skeletal syndromes and the heart Syndrome

Cardiac abnormality

Skeletal syndrome

Ellis van Creveld

Single atrium or ASD

TAR

ASD or tetralogy of Fallot ASD (and others)

Chondrodystrophic dwarfism Absent radius, thrombocytopenia Upper limb deformity, hypoplasia of clavicles Polydactyly, obesity

Holt-Dram Laurence-MoonBiedl-Bardet Noonan Tuberous sclerosis Leopard Cockayne Progeria Apert

Variable Pulmonary valve dysplasia Cardiomyopathy Pulmonary stenosis Accelerated atherosclerosis Accelerated atherosclerosis VSD

Pectus excavatum Bone lesions Broad fades, ribs abnormal Cachectic dwarfism Skeletal hypoplasia Craniosynostosis, syndactyly

ASD = atrial septal defect; VSD = ventricular septal defect; TAR = thrombocytopenia-absent radius syndrome

tricle has to generate high pressures to overcome increased impedance to flow during the isometric exercises. Continuous measurement of pressure indicates that the mean arterial pressure rises above 160mmHg during maximal isometric deadlifts. Remodelling of the heart occurs so that high intraventricular pressures can be generated. This objective can be achieved through augmented tension development in each myofibril, but as (without altering its constitution) each myofibril has a performance maximum, this limits the contractile reserve available during times of haemodynamic stress. An alternative mechanism is to increase the number of in-parallel myofibrils, which would increase wall thickness. Greater wall tension and intraventricular pressure can then be generated with each myofibril developing the same tension as before. The appearance of the LV that is induced by pressure overload is essentially that of increased wall thickness, commonly referred to as concentric hypertrophy, similar to the appearance in patients with chronic hypertension or aortic stenosis. Endurance athletes who perform mostly isotonic exercises (long-distance running, swimming, cross-country skiing, aerobics) develop dilatation of the ventricles. Remodelling of the heart occurs to ensure that large stroke volumes can be delivered. As each sarcomere has a maximum shortening distance, the stroke volume is increased by increasing the number of in-series sarcomeres. However, according to LaPlace's law, the resulting ventricular distension would result in greater wall tension being required to generate the same intraventricular pressure. Some increase in in-parallel myofibrils and hence wall thickening is therefore also necessary. The appearance of

the left ventricle induced by volume overload is chamber dilatation and eccentric hypertrophy, similar to that found in patients with 'high-output states' such as Paget's disease, beriberi and anaemia, or those with valvular regurgitation. In elite athletes, because of the marked ventricular dilatation, the cardiothoracic ratio on chest X-ray may exceed 0.5. Owing to the large stroke volumes, their heart rates at rest and during equivalent exercise workloads are significantly lower than those of normal sedentary subjects. Their peak exercise heart rates are not diminished and, combined with the enormous stroke volumes, their peak cardiac outputs are markedly augmented (up to 35L/min). They therefore have superior oxygen delivery, exercise capacity and performance. In certain types of endurance training, such as cycling, rowing and canoeing, the exercise contains not only isotonic but also resistance components. In these situations the observed hypertrophy is characteristically a combination of concentric and eccentric. It is important to note that the above remodelling processes are adaptive mechanisms to maintain or enhance the pumping capability of the heart in response to the physiological demands imposed upon it. Although the ventricular appearances of hypertrophy and dilatation are similar to those in pathological states of aortic stenosis or regurgitation, these changes observed in athletes should not be regarded as abnormal but as supranormal, as they are physiological and beneficial. If they need treatment, say for hypertension, drugs (e.g. (3-blockers, verapamil) that would impede these adaptive changes should be avoided. Diagnosis and management Clinically, apart from bradycardia, other commonly found features of athletes are sinus arrhythmia, benign systolic murmur usually due to pulmonary flow murmur, and occasional third or fourth heart sounds. The ECG may indicate left and/or right ventricular hypertrophy. A common abnormality is atrial or ventricular arrhythmia, but most of these do not cause symptoms. Confirmation that these are benign may be achieved by demonstrating the complete disappearance of the arrhythmia during exercise. Because of the high vagal tone some athletes are prone to develop heart block, which may require pacemaker insertion. Echocardiography indicates the types and extent of hypertrophy or dilatation. The left ventricular ejection fraction is invariably normal. To determine whether an athlete has supranormal cardiac function a cardiopulmonary exercise test can be performed. Achieving a supranormal maximal oxygen consumption is an indirect indication of superior cardiac reserve. If in doubt, direct measurement of cardiac reserve can be made noninvasively by measuring the cardiac output and blood pressure during maximal exercise. Athletes are also susceptible to cardiac disease. Detection may not be straightforward, because the same extent of damage (e.g. myocarditis) occurring in a normal subject

12

585

would have resulted in obvious cardiac dysfunction or failure, but its occurrence in an elite athlete may be masked because the baseline cardiac performance may be so enhanced that the eventual cardiac function falls within the normal range. Symptomatically, however, the athletic subject would complain that exercise capacity is significantly reduced. Certain heart diseases are lethal when combined with vigorous or competitive sport. Subjects with hypertrophic cardiomyopathy, aortic stenosis, Marfan's syndrome involving the aorta and heart, and life-threatening arrhythmias should be dissuaded from taking part in such sport. Patients with suspected or confirmed viral myocarditis should be persuaded to stop training, because viral replication rates are enhanced with continued exercise, and the resultant damage to the myocardium may be worse than detraining. Apart from this, there is the potential danger of triggering serious arrhythmia during myocarditis, which may result in sudden death.

Sternal depression and straight back syndrome Abnormalities of the shape of the thorax are a common cause of systolic murmurs; they may also give rise to musculoskeletal chest pains, an abnormal X-ray appearance of the heart and, in severe cases, ECG changes. Sternal depression and an abnormally vertical dorsal spine (straight back syndrome) both allow very little room for the heart in the anterior-posterior axis; this is well shown on the lateral chest X-ray. There is often a systolic murmur arising from the distorted great vessels; respiratory function may be impaired because of restricted chest wall expansion on inspiration, giving rise to a poor effort tolerance; and there may be associated mitral valve prolapse as a congenital anomaly. The ECG commonly shows right bundle branch block, and there may be anterior T-wave inversion in the V leads in the most severe cases. Assessment includes careful clinical examination to exclude primary cardiac pathology, and echocardiography and Doppler ultrasound where doubt exists as to whether there is more than a 'flow murmur'. Mild forms of this condition are commonly seen in pregnancy, when the patients are referred because of the murmur. Reassurance often helps the chest pains, which usually have a musculoskeletal basis and become worse when the possibility of heart disease is raised. Simple back exercises may also help.

PERICARDIAL DISEASE

586

The pericardium is a fibrous bag that supports and invests the heart. Like the pleura, it may be the site of inflammation, neoplastic infiltration and calcification, and may also be the site of fluid collection with or without inflammation.

TABLE 12,43 Causes of pericardial disease Infections Viral, e.g. Coxsackie, influenza Bacterial: pyogenic and tuberculosis Fungal Metabolic Uraemia Hypothyroidism Neoplastic Carcinoma of the bronchus Lymphomas Myocardial infarction

Autoimmune Collagen disease, rheumatoid arthritis, SLE Postpericardiotomy and postinfarction (Dressler's) syndromes Serum sickness and drug reactions Idiopathic (benign recurrent viral) relapsing pericarditis Rheumatic fever Haemopericardium Cardiac rupture Aortic dissection Trauma and surgery

Radiotherapy

Aetiology The causes of pericardial disease are shown in Table 12.43. All of these conditions can produce the clinical features of pericarditis with effusion. Heart failure can also cause a pericardial effusion, but these are small asymptomatic effusions, often discovered at routine echocardiography. Acute viral pericarditis The commonest viral causes are Coxsackie virus A or B, varicella, herpes simplex, influenza virus and echovirus (type 8). Frequently, a definite viral cause cannot be established. Typically, it occurs in young adults 1 or 2 weeks after an upper respiratory infection. The patient then develops fever, pericardial pain and malaise which may persist for a few weeks. Unfortunately, recurrences occur in about 25% of cases, possibly from an autoimmune mechanism. Tamponade is unusual. Acute pyogenic pericarditis Acute pyogenic pericarditis can follow pneumonia or a septicaemia and is treated with the appropriate antibiotics and repeated drainage, which may require surgical intervention. Staphylococcus and Haemophilus are the most common organisms involved. It may also occur in immunosuppressed patients and following cardiac surgery. Tuberculous pericarditis Tuberculous pericarditis may be asymptomatic until it presents as pericardial constriction (see below). It may present as a large pericardial effusion with enlargement of the cardiac silhouette on chest X-ray, or as a low-grade fever with weight loss, fatigue and features of pericarditis. Pericardial aspiration may reveal the organisms, but occasionally (as with pleural effusions) culture of the fluid or pericardial biopsy may be required before the diagnosis is established.

Malignant pericarditis Pericardial effusion fluid should always be sent for cytology, as malignancy may present with a pericardial effusion. The clinical features are similar to those of other types of pericarditis, but may be complicated by direct tumour extension into the pericardium, with a degree of constriction. Lung cancer is the commonest cause. Uraemic pericarditis Uraemic pericarditis develops with severe untreated uraemia and disappears with dialysis. It appears to be directly related to the blood urea level. Other types In most other types of pericarditis, such as in autoimmune disease, pericardial involvement is an incidental finding but may occasionally lead to constrictive pericarditis.

Clinical features The principal symptoms of pericarditis are pericardial pain, breathlessness (if the effusion becomes large enough) and, eventually, tamponade. There may be systemic symptoms related to infection, such as fever, rigors, weight loss and symptoms of the underlying cause (lung cancer, tuberculosis etc.). Pericardial pain Pericardial pain is a mediastinal pain that can be very similar to that of myocardial ischaemia, but differs in that it is more commonly referred to the shoulder and back and is often influenced by posture and mediastinal movements, such as coughing, sneezing and, particularly, swallowing. It is often relieved by leaning forward. The diagnosis may be difficult because of the occurrence of pericardial pain following myocardial infarction. There can also be a pleural element to the pain. Pericardial rub Pericardial rub - the main physical sign of pericarditis - is very variable and often absent. It is a rough scratching sound which seems nearer to the observer than the stethoscope headpiece. Typically, it has three components - systolic, diastolic and atrial-systolic. However, the atrialsystolic and diastolic components may not be present, and a purely systolic scratchy murmur may have other causes. The noise can sometimes be accentuated by leaning the patient forward or, if conditions allow, by putting them in the knee-elbow position. The sound conies and goes and does not necessarily disappear with the formation of an effusion; indeed, it may be found in up to 70% of patients with pericarditis and effusion. Pericardial effusion Pericardial effusion does not usually cause symptoms unless it is large, when there may be some breathlessness and, eventually, features of tamponade (see below). The

clinical signs of an effusion without tamponade (soft heart sounds, increased cardiac dullness etc.) are not as helpful as chest X-ray and ultrasound.

12

Cardiac tamponade If the pericardial cavity fills with blood or pericardial fluid, the pericardial pressure may rise sufficiently to embarrass filling of the heart and lead to a form of cardiogenic shock (cardiac or pericardial tamponade), with very low cardiac output, cerebral confusion, and severe hypotension proceeding to death if the pericardium continues to fill. The pericardium is viscoelastic and can therefore stretch slowly to accept as much as 2 L of fluid over a period of time without the development of temponade, but much smaller amounts of fluid can be fatal if they accumulate quickly.

Aetiology Tamponade is usually caused by bleeding or effusion into the pericardial space. It can complicate aortic dissection and myocardial infarction with rupture into the pericardium. Bleeding can also follow surgery or trauma to the heart and mediastinum, and may follow perforation of the heart at cardiac catheterization, or injury after pericardial aspiration. Malignant disease is a common cause, but tamponade is an uncommon complication of other causes of pericardial effusion. Clinical features Severe hypotension with impalpable or very weak peripheral pulses may be present. Classically, careful palpation of the arterial pulses reveals pulsus paradoxus, in which the pulse pressure drops considerably on inspiration and the pulse may even become impalpable. The paradox is not that the pulse volume diminishes with inspiration (which happens normally to a small extent), but that the stroke volume of the left ventricle declines markedly while - paradoxically - the heart seems to be beating normally. The mechanism of pulsus paradoxus is related to the pericardial temponade dictating a small fixed volume to the heart. Normally, right ventricular filling is aided by inspiration but with tamponade (or pericardial constriction), this rightsided mechanism inhibits filling of the constricted left ventricle, whose stroke volume falls. The jugular venous pressure is usually markedly raised, may show a marked systolic drop (prominent x descent) and, sometimes, rises with inspiration (Kussmaul's sign). Chronic tamponade can mimic constriction, with hepatic enlargement and oedema. The heart sounds may be faint and there can be a pericardial rub.

Investigation of pericarditis The ECG changes of acute pericarditis are shown in Figure 12.66. The ST elevation remains concave upwards and is

587

FIG. 12.66 EGG of acute pericarditis illustrating widespread concave ST segment elevation

followed later by widespread T-wave inversion, probably indicating a degree of epicardial myocarditis. A large effusion can produce a diminution of ECG voltages; in extremely large effusions there can be 'electrical alternans', with the voltage being influenced by the heart swinging backwards and forwards in the effusion. Chest X-ray (if there is an effusion) shows increased cardiac silhouette, with a globular appearance to the heart but no evidence of raised pulmonary venous pressure. This appearance is not specific. Echocardiogram is the technique of choice in demonstrating a pericardial effusion, preferably with a 2D scan. A small effusion can be localized, and may be missed unless the operator is experienced. With a large effusion echocardiography will show the heart surrounded by a large bag of fluid, which is seen as an 'echo-free' space. Pericardial aspiration is described below. The fluid should be sent for bacteriological and cytological examination.

Pericardial pain often responds to anti-inflammatory agents. Infective pericarditis demands prompt diagnostic aspiration and identification of the infective agent, with appropriate chemotherapy. In viral pericarditis antiinflammatories alone are used, but where there is an autoimmune element, such as in postinfarction syndrome and recurrent 'viral' pericarditis, corticosteroids may help. A large effusion is drained as described below. If the patient suffers repeated attacks, a pericardial window can be made into the pleural space at operation.

588

CONSTRICTIVE PERICARDITIS Constrictive pericarditis results from thickening and fibrosis of the pericardium with, in some cases, extensive calcification.

Management

O Case 12.8

Pericardial aspiration Pericardial aspiration can be both diagnostic and therapeutic. In an emergency, e.g. after a road traffic accident with trauma to the chest, if a patient is moribund with features suggesting tamponade (i.e. very high venous pressure), then aspiration from the apex of the heart with a needle of adequate length should not be delayed. Removal of even 50-100 mL of blood from the pericardium may reverse the state of shock. Elective aspiration is best performed in a coronary care unit with ECG, blood pressure and echo monitoring and a cardiac defibrillator available. After local analgesia of the chest wall, the pericardium is punctured either at the cardiac apex (which is free of major coronary vessels) or via a subxiphoid approach. The ECG monitor shows ventricular ectopics if the ventricles are touched. Using a metal needle connected to an ECG V lead, a current of injury (raised ST segment) is produced as soon as the myocardium is touched and, hopefully, before the ventricle is pierced. Ideally, a cannula, rather than a needle, is left in the pericardium, and aspiration or drainage follows. If the fluid looks like blood, pericardial blood from a haemopericardium can be distinguished from ventricular blood by the fact that it is defibrinated and will not clot. It also has a lower haemoglobin and haematocrit than venous blood, as it is almost always diluted with serous pericardial fluid. Diagnostic aspiration of a large effusion without subsequent drainage can lead to pericardial fluid draining into the pleural cavity. Drainage of a large effusion can take several hours, and recurrent effusion (e.g. in malignant disease) may require fenestration of the pericardium or insertion of a pericardioperitoneal drain by a cardiothoracic surgeon. O

Fig. 12.29

MCQ 12.23

Aetiology There may be an antecedent history of acute pericarditis or radiotherapy. Tuberculous pericarditis is now rare in western countries, but is the classic cause of calcific constrictive pericarditis.

Pathogenesis The heart is effectively encased in a rigid box, which eventually prevents it from filling. Haemodynamically, the diastolic pressures or filling pressures of the two sides of the heart become the same as the intrapericardial pressure; as this rises, the patient goes into bilateral heart failure, in which features of right-sided failure predominate.

Sometimes the constriction is selective, so that the left or the right side, inflow or even outflow, can be compromised most.

Clinical features Symptoms The major symptom is severe oedema, often with ascites, liver engorgement and progressive weight loss secondary to hepatic engorgement and low cardiac output. Patients with long-standing disease look cachectic, and with the liver enlargement and a degree of jaundice are easily thought to have carcinomatosis or cirrhosis of the liver. The absence of significant left-sided heart failure symptoms is characteristic. The patient can often lie flat without problems and does not have a history of paroxysmal nocturnal dyspnoea, but will always have a limited exercise tolerance and dyspnoea on exertion. Constriction may arise with no previous history of pericarditis. A previous history of pericarditis is a valuable clue to the diagnosis. Signs A very high venous pressure is invariable, as the enddiastolic pressure in both ventricles is about 25mmHg. The high venous pressure is interrupted by a rapid x and y descent or systolic collapse of the venous pressure. The pulse exhibits pulsus paradoxus, but this is not invariable and the presence of AF in about 30% of cases makes the sign difficult to elicit. The pulse will, in any case, be of small volume. The heart is not clinically enlarged and the rigid pericardium may impede palpation of the cardiac impulse. The heart sounds may be faint with a loud additional sound, the pericardial knock. This is a loud early third heart sound, resulting from the rapid filling of the ventricles being suddenly impeded by the rigid pericardium. Murmurs are uncommon.

Investigation The ECG is non-specific, showing low voltages and, usually, T-wave inversion; other changes may reflect epicardial myocardial damage resulting from pericardial impingement on the epicardial coronary arteries. Chest X-ray shows a normal-sized or slightly enlarged heart, often with an abnormal contour due to flattening of the atria. Calcification may be seen in the A-V groove on a lateral film. 0 Depending on the sites of constriction, there may be signs of left or right atrial hypertension. Pleural effusions are common. Screening of the heart reveals a typical 'diastolic shock', with lack of pulsatility of the atria. Echocardiography may demonstrate thickening of the pericardium. The ventricles fill rapidly in early diastole, but the differential diagnosis from restrictive cardiomyopathy is not easy. Cardiac catheterization would show elevation and equalization of diastolic pressures in all four chambers, with characteristic 'square root sign' of the early diastolic ventricular pressure waveform.

Management The only effective treatment is surgical decompression of the heart by removing enough of the pericardium to allow the heart to fill effectively. This means removing the adherent pericardium from around the great veins and atria, which are extremely thin-walled. Where the pericardium is adherent to the ventricles the coronary arteries are easily damaged, so considerable surgical skill and patience are required. ©

12

SYSTEMIC BLOOD PRESSURE AND HYPERTENSION Cardiac performance and vascular resistance form part of an integrated system which controls arterial pressure and blood flow to the tissues. Despite our knowledge of many of the component mechanisms of pressure and flow control, many aspects remain unclear in both health and disease.

CONTROL OF BLOOD PRESSURE Neurosympathetic blood pressure control Activation of preganglionic sympathetic neurons to the heart produces an increased heart rate and increased strength of contraction. Vasomotor sympathetic outflow constricts vascular smooth muscle, producing arterial vasoconstriction. The supply to the veins is richer and more sensitive than to the arterial system. Increased vagal activity reduces the heart rate but, with the exception of the coronary and salivary gland vascular beds, there is no parasympathetic control of the blood vessels. Moment-to-moment reflex control of the circulation is via the arterial baroreceptors in the wall of the aorta and in the carotid arteries. The speed of this pressurecontrolling mechanism is illustrated by its important role in maintaining systemic pressure during abrupt changes in posture. Neurosympathetic influences on blood pressure and the circulation are rapid but tend to be transient; this is partly due to a tendency for the reflexes themselves to undergo accommodation, and partly because peripheral autoregulatory mechanisms may overcome neural constrictor effects.

Renin and angiotensin The renin-angiotensin-aldosterone system (Fig. 12.25) plays a major part in the control of blood pressure, regional blood flow and blood volume, via its control of sodium balance. Renin secretion from granules in the juxta-

589

glomerular apparatus is stimulated by a fall in plasma volume, a drop in blood pressure and sodium depletion. The consequent rise in angiotensin II production and aldosterone secretion leads to sodium and water retention, and hence restoration of sodium balance and plasma volume. Under conditions of plasma volume reduction and sodium depletion (<50mmol Na+/day), angiotensin II contributes to the maintenance of blood pressure by both vasoconstriction and interaction with the sympathetic nervous system. In the kidney, angiotensin II has a preferential constrictor effect on efferent (postglomerular) arterioles, and thus maintains glomerular filtration pressure when the overall arterial pressure is diminished. The renin-angiotensin system plays a crucial role in the increased blood pressure seen with renal arterial constriction; the latter causes a rapid release of renin and angiotensin II and hence a rise in blood pressure. Angiotensin II causes changes in intrarenal flow patterns and stimulates aldosterone release. The combination of haemodynamic and direct effects on the renal tubules increases sodium reabsorption. An extrarenal local tissue renin-angiotensin system has been identified, and there is increasing evidence that abnormal function of this system may play a role in the pathogenesis of hypertension.

first sound (Kj) and diastolic pressure is usually within a few mmHg of sound muffling (K4). In many major trials of blood pressure treatment sound disappearance (K5) is used to define diastolic pressure. In some individuals K5 cannot be defined when peripheral vascular resistance is low, as sounds carry on to zero cuff pressure. The anauscultatory gap (K2-K3) is particularly prominent in the elderly; this is important, as mistaken identification of K3 as K] might lead to a serious underestimation of systolic pressure - by up to 15-20 mmHg in some cases. Its presence is accentuated by venous distension distal to the cuff. Accurate blood pressure measurement requires a cuff of adequate size: a small cuff will seriously overestimate pressures. The subject should be seated, with the manometer reservoir at the level of the sternal angle. The cuff bladder is arranged medially to overlap comfortably the medial aspect of the upper arm. Pressure in the cuff is inflated rapidly to 20-30 mmHg above that required to occlude the brachial or radial pulse. Pressure is then reduced slowly from the cuff while listening with a stethoscope over the previously located brachial pulse. The rate of pressure drop should not exceed 5 mmHg per second, and the line of sight should be such as to avoid parallax errors when reading the falling column of mercury.

Nitric oxide and endothelin

SYSTEMIC HYPERTENSION

More recent evidence suggests that nitric oxide (NO, previously known as endothelium-derived relaxant factor), endothelin, prostacyclin and other vasoactive factors that are locally produced by endothelial cells and vessel wall, play a role in the control of vascular tone and hence blood pressure. Suppression of NO production can result in hypertension, thus introducing the concept that blood pressure is controlled not only by vasoconstrictive neurohumoral factors, but also by paracrine vasodilators. Overproduction of NO, such as is seen in endotoxaemia, results in overvasodilatation leading to circulatory shock. Endothelin (first discovered in 1988), a peptide produced by endothelial cells and released abluminally towards the smooth muscle cells, is one of the most powerful vasoconstrictors. Its exact physiological role in the control of blood pressure is still unclear. Increased vascular synthesis and plasma concentrations of endothelin have been reported in hypertension, chronic renal failure and toxaemia of pregnancy. Hypertension has been reported to be due to a tumour overproducing endothelin. Endothelin receptor antagonists may have a role in the treatment of hypertension.

Definitions Blood pressure varies considerably, rising during exertion, and is subject to a circadian variation which may not be completely activity related. Ambient temperature and the degree of emotional or physical arousal will also affect systemic pressure levels. Systolic blood pressure rises with age

BLOOD PRESSURE MEASUREMENT

590

Korotkoff defined the auscultatory features (Korotkoff sounds) over the brachial pulse as pressure was released from an inflatable cuff. Systolic pressure correlates with the

FIG. 12.67 Changes in blood pressure with age in three different communities: Londoners, urban Zulu and Kalahari Bushmen

therapy), it is very important to pay attention to other contributing risk factors as well.

Hypertension in special groups in the community

FIG. 12.68 Distribution of diastolic blood pressure in a western population (black line) Risk of death is shown by the blue line. The diastolic blood pressure levels are indicated at which treatment reduced mortality in two large trials (see text). VA = Veterans Administration; MRC = Medical Research Council.

in developed countries (Fig. 12.67). Within a community, blood pressure (systolic and diastolic) has a normal or Gaussian distribution, a factor that becomes very important when definitions of abnormalities of blood pressure are considered (Fig. 12.68). Insurance companies were among the first to note that a reduced lifespan was associated with recordings of elevated blood pressure. It was clear from their data that no cut-off pressure level could be identified at which no increased risk of premature death was present. Hypertension could therefore be defined as a blood pressure above a particular value, e.g. 160/90 mmHg. A more flexible definition - able to take into account new information derived from the results of therapeutic trials - might be preferable, for example 'the level of blood pressure above which investigation and treatment do more good than harm.' Figure 12.68 illustrates how successive treatment trials (the Veterans Administration in 1965, and the Medical Research Council in 1985) have lowered the blood pressure at which treatment has been shown to have benefits.

Risk stratification A small number of individuals have very high blood pressure and are at very high risk of complications of the disease and death. At lower levels the risk attributable to blood pressure is relatively less for the individual, but that low risk is shared by so many people that the associated complications and premature mortality pose a significant burden on the community. Blood pressure is not the only risk factor in these calculations. Because drug treatment of great numbers of the population tends to carry unwanted financial and medical consequences (owing to the requirement for lifelong

12

There is a tendency to a high systolic blood pressure in old age and this carries with it the risks associated with hypertension at any age. Therapy is possible in older patients, and has been shown to reduce the risk of stroke and myocardial infarction. Urban black populations have a higher incidence of hypertension than their white neighbours. Both environmental and genetic factors are likely to be responsible. Hypertension is rare in some rural African communities but a very major problem in others. Stroke is the most common complication among black hypertensives, and myocardial infarction is relatively rare; the opposite is the case for white hypertensives. The effectiveness of treatment is similar among blacks and whites; however, there are minor differences in the response to certain drugs, with a slight resistance to the effects of B-adrenoreceptor blocking drugs being shown by black patients.

PRIMARY HYPERTENSION Borderline and labile hypertension About one-third of patients found to have an abnormal blood pressure at their first visit to the doctor will have lower blood pressure on subsequent visits. This borderline group may subsequently develop established hypertension. Some individuals appear to have wide swings in blood pressure, and have been called labile hypertensives. It seems likely that they do not form a pathologically distinct group, but manifest an accentuation of normal variability. As a group, they have a higher mortality than consistently normotensive individuals.

'White coat hypertension' Some patients have such a marked stress reaction when seeing a doctor (whether medical and whether wearing a white coat or not) that they raise their blood pressure to hypertensive levels. They appear calm externally and deny anxiety. This category of patients was first identified as a distinct subgroup on the introduction of ambulatory BP monitoring (see below). Whether patients with 'white coat hypertension' have a similar adverse prognosis as those with established chronic hypertension is unknown. There is a suggestion that some of them will eventually develop chronic hypertension later in life. These patients should be encouraged to undertake non-pharmacological methods of lowering blood pressure, such as avoiding obesity and increasing exercise. Some patients with essential hypertension may also have a component of 'white coat hypertension' superimposed.

591

They are often given increasing numbers and doses of antihypertensive drugs. Repeated ambulatory blood pressure monitoring may be required to gauge their true response to treatment.

Because of arteriosclerotic changes in the major vessels of the elderly, the pulse pressure widens, with a greater rise in systolic than in diastolic pressure. Thus, isolated systolic hypertension may be observed. Even in isolation, systolic hypertension is associated with an excess risk of morbidity through stroke, myocardial infarction and congestive cardiac failure. Benefit of treatment in such cases is likely, and it is reasonable to aim for a systolic pressure reduction to <150mmHg.

not. This differentiation has no clinical or prognostic value as the conditions are identical, with an untreated mortality of about 80%. Patients may have a past history of hypertension, but in most this is the first manifestation of their disease. It is one of the few occasions when hypertension is associated with symptoms, which include visual disturbance, headache and breathlessness due to pulmonary oedema; with severe degrees of cerebral oedema blunting of consciousness, coma and epileptic seizures also occur. It is usually easy to treat and therapy is undoubtedly life-saving. Many factors have been implicated in the genesis of malignant hypertension. The rate of rise of pressure may be more important than absolute levels of hypertension and there may be associations with other factors that damage blood vessels, such as cigarette smoking.

Benign or essential hypertension

Hypertension in pregnancy

'Essential hypertension' is commonly used to mean idiopathic or primary hypertension. The majority of patients with hypertension (over 90%) have no known cause, and thus fall into the category of essential hypertension. This condition affects a vast proportion of the population - possibly as many as 20% of the middle-aged subjects in the UK, depending on the blood pressure levels adopted as the definition of hypertension. The diagnosis depends on an established elevated systemic pressure in the absence of changes in the ocular fundi and of proteinuria, and where no identifiable cause of pressure elevation has been demonstrated. Essential hypertension can affect any age group and can, when untreated, lead to many complications, including malignant and accelerated-phase hypertension.

During pregnancy cardiac output rises, but blood pressure normally falls owing to a decrease in vascular resistance. Obstetricians regard as abnormal a blood pressure greater than 140/90 (measured using Korotkoff 4 for diastolic pressure) in the first half of pregnancy, and by this criterion nearly one-quarter of the pregnant population will be hypertensive. About 2% will have a blood pressure greater than 160/105. Pre-eclampsia and eclampsia are placental syndromes which are limited to the duration of pregnancy, and of which elevated blood pressure is just one manifestation. Pre-eclampsia was previously defined as hypertension with proteinuria (>0.5 g/L per 24 hours) and oedema. However, the latter is a frequent finding in normal pregnancies and often absent in pre-eclampsia, so that it is no longer included as a diagnostic criterion. Proteinuria is a late manifestation of pre-eclamptic renal impairment. Impaired uric acid clearance, which occurs much earlier in the disease process, is more useful diagnostically. Evidence of disseminated intravascular coagulation occurs late in the disease. Eclampsia is defined as the development of convulsions in the above setting. The diagnosis of pre-eclampsia thus depends on the observations of an abnormal rise in blood pressure between the first and second halves of pregnancy of >30/20mmHg, associated with abnormal urate levels (>0.35 mmol/L at 32 weeks or >0.4mmol/L thereafter); proteinuria, impaired renal function and clotting disorders confirm the diagnosis, but occur very late in the disease. The consequences of pre-eclampsia are placental ischaemia and infarction and, in severe cases, maternal cerebral haemorrhage. Treatment is elective delivery, which confirms the diagnosis by effecting a rapid and complete cure. Medical management may reduce blood pressure but has not been shown to ameliorate other features of preeclampsia, which continue until delivery is effected. Pre-existing hypertension, low social class and the first pregnancy with a particular partner are risk factors for preeclampsia; interestingly, it appears to be less common in smokers.

Isolated systolic hypertension

Malignant and accelerated-phase hypertension Malignant or accelerated-phase hypertension is associated with vascular fibrinoid necrosis and loss of precapillary arteriolar autoregulation. The consequence is capillary vessel rupture and tissue necrosis through haemorrhage and ischaemia. The brunt of the damage is borne by renal, adrenal, cerebral, retinal, pancreatic and mesenteric vessels, for reasons that remain unclear. The clinical findings are elevated levels of blood pressure (which may be severe), associated with haemorrhages and exudates visible on the retina. Proteinuria due to similar vascular changes in the kidney is usual. Papilloedema (when present) O reflects cerebral oedema and was previously used to differentiate malignant hypertension, where it was present, from accelerated-phase hypertension, where it was

1

592

Fig. 12.30

TABLE 12.44 Causes of secondary hypertension Type of cause Renal Parenchymal disease

Arterial disease

Endocrine disease Adrenal cortical overactivity Adrenal medullary overactivity Drug-induced

Congenital defects Miscellaneous

Examples Chronic renal failure of any cause Acute and chronic glomerulonephritis Polycystic disease Pyelonephritis Tumours, e.g. haemangiopericytoma; Wilms' tumour Diabetes Arteriosclerotic disease Fibromuscular dysplasia Embolism Polyarteritis nodosa Cushing's syndrome Conn's syndrome, adrenal hyperplasia Enzyme defects, e.g. congenital adrenal hyperplasia Phaeochromocytoma Contraceptive pill Liquorice-containing compounds Steroids Withdrawal of antihypertensives, esp. clonidine Cheese ingestion in patients taking monoamine oxidase inhibitors Sympathornimetic decongestant nasal sprays/appetite suppressants Coarctation of the aorta Porphyria (during acute attacks) Lead poisoning (during acute phase) Raised intracranial pressure

SECONDARY HYPERTENSION The causes of secondary hypertension are given in Table 12.44. As a general rule, the younger the patient with hypertension the greater is the likelihood that a cause will be found. In an adult hypertensive population less than 5% of patients will have a discernible cause. Although this proportion is small compared to those with essential hypertension, because many of the secondary causes are treatable by corrective surgery or specific therapy a high index of suspicion is required to avoid overlooking a treatable cause.

Renal disease The precise cause of high blood pressure in chronic renal failure is not clear. Adequate management of hypertension is important to prevent vascular complications due to hypertension and prevent further deterioration of renal function. There is no justification for surgical removal of

kidneys in unilateral renal disease, as medical treatment of hypertension is now likely to be more effective. Renal arterial stenosis activates the renin-angiotensin system, and restoration of the normal vascular supply may potentially cure the associated hypertension. Atheromatous disease is the commonest form of renovascular disease; it mainly affects middle-aged male smokers and is often part of a generalized degenerative disease of the arteries. Angioplasty may be useful, but the disease in smaller arteries limits the overall prognosis. In fibromuscular dysplasia (a disease of younger age groups), surgical reconstruction of renal arteries may be very effective. The importance of diagnosing renovascular hypertension has increased in recent years with the introduction of ACE inhibitors. These are to be avoided in renovascular hypertensives, as they can precipitate renal failure owing to the importance of angiotensin II in maintaining intrarenal haemodynamics in the underperfused kidney.

12

Endocrine disease Hypertension, which may be severe, usually accompanies Cushing's syndrome (p. 932). Hypemldosteronism Overproduction of aldosterone may be due to single or multiple autonomous adrenal adenomas (Conn's syndrome) or bilateral micronodular hyperplasia of the zona glomerulosa. Clinical suspicion of primary hyperaldosteronism is raised when untreated hypertension is associated with hypokalaemia (K+ < 3.5mmol/L), alkalosis (HCCV > 30mmol/L), and a serum sodium which is high or at the upper limit of normality (Na+ > 145mmol/L). Investigation of primary hyperaldosteronism is aimed at differentiating adenomas from adrenal hyperplasia, as the former may be removed surgically. Radionucleotide adrenal scanning with labelled cholesterol, CT scanning and adrenal vein catheterization to measure the production of aldosterone offer the best methods of diagnosis. Explorative surgery may be necessary in some instances. Hyperplasia is treated successfully with antialdosterone drugs, such as spironolactone or amiloride.

Phaeochromocytomas Phaeochromocytomas are tumours, usually found in the adrenal, that secrete a mixture of catecholamines. They are responsible for clinical syndromes where hypertension may be episodic and alternate with episodes of postural hypotension. Paroxysms of hypertension are often associated with a constellation of symptoms, including headache, sweating, palpitations, feelings of apprehension and tremor. These symptoms are more sensitive in making the diagnosis than are most biochemical tests. The majority (90%) of the tumours are benign, and 10% arise outside the adrenals, most commonly at the bifurcation of the abdominal aorta. There are associated phaeochromocytomas in Sipple's syndrome (often a familial disorder with the combination of medullary carcinoma of thyroid and

593

parathyroid adenoma) and in neurofibromatosis or von Recklinghausen's disease (p. 421). Investigation and management The diagnosis is confirmed by the detection of increased amounts of catecholamines and their breakdown products - normetadrenaline, metadrenaline and vanillylmandelic acid (VMA) - in urine or plasma. Localization of the tumour is by CT scanning and radionucleotide scans with MIBG (131I meta-iodobenzylguanidine). Treatment is surgical when possible, but can be hazardous owing to surges of catecholamine secretion during tumour manipulation. Preliminary treatment with a- and {3-blockade is essential, as are agents to raise blood pressure and prevent vascular collapse when the levels of these hormones fall rapidly as the tumour is removed.

Drug-induced hypertension Mineralocorticoid-induced hypertension with hypokalaemia and alkalosis accompanies the use of steroids, and may occur with carbenoxolone and excessive liquorice consumption. Hypertension induced by the contraceptive pill is generally mild and reversible on cessation of therapy, although this may take several months.

Coarctation of the aorta In coarctation of the aorta, children or young adults present with asymptomatic upper limb hypertension and are noted on physical examination to have characteristically late and low-volume femoral pulses compared with the radial pulses (radiofemoral delay). Hypertension may be difficult to treat medically and, if left until the patient is a teenager, is often only partially helped by surgical repair of the narrowed aortic segment. Those operated on late (after 20 years of age) frequently suffer the complications of uncontrolled hypertension, and about 12% may die in their 30s of cerebral haemorrhage or ruptured aortic aneurysm. Surgery is now recommended before the age of 6, but balloon angioplasty has recently been successfully used in this condition and may be used more frequently in the future if the initial good results are maintained. The hypertension of coarctation is not fully explained, but renal hypoperfusion, aortic baroreceptor activity and mechanical effects of aortic obstruction may all play a part. CLINICAL FEATURES OF HYPERTENSION

Until complications appear, essential hypertension has no symptoms and, by definition, no associated physical signs save the elevated blood pressure. Headaches (once widely

regarded as indicators of hypertension) are no more common in hypertension than in the general population, although very severe hypertension, associated with cerebral oedema, does produce headache. Breathlessness may be present, owing to elevated left ventricular and enddiastolic pressure and pulmonary venous congestion produced by left ventricular hypertrophy. Once the complications of essential hypertension are present, they are reflected in the symptoms and physical signs. Ventricular hypertrophy may lead to breathlessness, orthopnoea and frank cardiac failure; coronary disease to angina pectoris or myocardial infarction; cerebrovascular disease to stroke and dementia; and renal disease to all its associated symptoms. The presence of some symptoms (e.g. those associated with phaeochromocytoma) will suggest a secondary cause. Symptoms of renal impairment (e.g. nocturia and polyuria) may reflect the consequence of long-standing hypertension rather than its cause. Certain points in the examination of the hypertensive deserve emphasis, and require careful observation at diagnosis and during follow-up. Pulse Rate and rhythm. Atrial fibrillation is a late consequence of hypertension. A resting tachycardia might imply cardiac decompensation or a secondary cause, such as phaeochromocytoma or hyperthyroidism. Arterial wall. If thickened arteries are palpated in an under-50-year-old, aggressive treatment of blood pressure and other risk factors would be justifiable. Peripheral pulses. Radiofemoral delay is a sign of aortic coarctation, which may be a cause of hypertension in young patients. In older patients atherosclerotic disease may cause peripheral vascular bruits and absent distal pulses. Apex beat A displaced and thrusting apex beat indicates left ventricular hypertrophy (LVH) and therefore hypertension worthy of aggressive treatment. The absence of this sign does not exclude LVH. Abdomen Palpable kidneys might be found with polycystic renal disease. A systolic and diastolic abdominal or flank bruit raises the possibility of renal arterial stenosis. Optic fundus Haemorrhage, exudates and papilloedema (grades 3 and 4) are all signs of accelerated or malignant hypertension. Early signs of arterial wall thickening are narrowing of the arterial lumen (grade 1, which is difficult to identify) and arteriovenous nipping (grade 2). General aspects Other important general signs include:

1

594

Fig. 12.31

• Signs of risk factors for vascular disease (xanthomata), which might tip the balance in favour of treatment because of the increase in vascular risk;

Signs suggesting Cushing's syndrome or neurofibromatosis (associated with phaeochromocytoma); Indications of congestive cardiac failure.

INVESTIGATION The aims of investigation are: • To establish the diagnosis • To identify a cause of the hypertension • To determine the presence and extent of complications from hypertension • To assess efficacy of treatment. The history and clinical examination provide clues as to which investigations are appropriate. Numerically, attempting to find an aetiology is far less rewarding, as only about 5% of all cases of hypertension have an identifiable cause. This proportion is higher in young adults, all of whom should be investigated accordingly.

Persistent proteinuria requires detailed testing of renal function (p. 1037). The discovery of haematuria should be followed by tests for inflammatory and neoplastic disease of the kidney. A 24-hour urine collection is required to quantify protein excretion; this may be helpful in hyperaldosteronism, Cushing's disease and phaeochromocytoma.

12

Blood tests Routine haematological and biochemical tests are seldom rewarding. However, it is usual to obtain a full blood count and biochemical profile, which may occasionally provide useful information, for example: • An indication of renal function (urea and creatinine); • Electrolyte disturbances (low K+, raised HCO-3 and Na+) that might indicate hyperaldosteronism; • Other associated risk factors, such as hyperuricaemia or hyperlipidaemia; • Anaemia due to chronic renal disease, or polycythaemia secondary to renal disease.

ECG Ambulatory BP monitoring The difference between ambulatory BP (ABP) monitoring and clinic BP readings is similar to that between cinematography and snapshot photography. There are two ways of monitoring ambulatory BP: • With intra-arterial cannulation and direct pressure measurement • With indirect cuff manometry both of which record readings in a portable recorder which can subsequently be fed into a computer for display and analysis. The former method is now considered too invasive and is reserved for research purposes. Some patients with elevated blood pressure have 'white coat hypertension' (see above). The characteristic pattern is that they have significantly elevated blood pressure when the ambulatory BP monitor is first applied, and subsequent blood pressure readings at home or elsewhere show normal levels. A proportion of them show a further elevation of blood pressure as they make their way to hospital to return the equipment. Patients with essential hypertension possessing a 'white coat' component show elevated pressure at home, and further elevation during visits to hospital or clinic. The correlation between the effects of hypertension, such as left ventricular hypertrophy with isolated clinic BP reading, is much poorer than that with ABP. ABP provides information on diurnal variation in blood pressure. Blood pressure falls by about 20mmHg during sleep. The absence of nocturnal sleeptime falls in blood pressure is also associated with adverse prognosis. Urine testing Even though the yield of abnormal results by dipstick testing is not high, this simple test is essential in all cases.

An initial ECG may be helpful as a baseline for future changes, as well as in deciding whether to offer treatment. Features to be documented are signs of left atrial enlargement ('P mitrale', p. 477) and, with more severe levels of hypertension, evidence of LVH (Fig. 12.69) possibly associated with abnormalities in the ST segments and T waves that denote the 'strain' pattern. With effective blood pressure reduction these changes usually resolve. Chest X-ray The chest X-ray is helpful in establishing cardiac size and may reveal cardiac failure. 1 In most Caucasian patients it is normal, but cardiac enlargement is common in hypertensives of African descent. In young hypertensives, signs of a coarctation (p. 540) may be found.

FIG. 12.69 Typical ECG in severe hypertension showing left ventricular hypertrophy

595

Ultrasound techniques The echocardiogram can determine the presence and extent of left ventricular hypertrophy. The presence of LV hypertrophy in a new patient without other causes for hypertrophy (e.g. athletism, aortic stenosis) is supportive of the diagnosis that the hypertension has been substantial and chronic. In epidemiological studies, left ventricular hypertrophy has been found to be an independent indicator of adverse prognosis. These patients benefit from more attentive management to achieve as normal a blood pressure level as possible. Abdominal ultrasound examination shows kidney size and ureteric anatomy and gives some information on the renal substance. Doppler study of the renal arteries can detect stenosis, the suspicion of which should lead to radionuclide renal scintigraphy to assess residual renal function, or to renal angiography. Renal artery stenosis is a contraindication for ACE inhibitor, renin inhibitor and angiotensin receptor antagonist therapy. Nuclear medicine techniques Nuclear medicine techniques are used to measure dynamic aspects of kidney and cardiac function. These specialized methods should be reserved for those in whom a specific problem has been identified. Renal angiography A renal arteriogram is the definitive test to determine renal artery stenosis. Angioplasty may be performed if the lesion is deemed suitable for intervention. Renal venous sampling may be performed to measure plasma renin activity, which provides complementary information about the extent and effects of arterial stenosis. Venous sampling may also be indicated to identify the locations of tumours producing vasoactive agents.

COMPLICATIONS OF HYPERTENSION

596

The development of LVH in hypertension is an adaptive response, but the effects on diastolic function may give rise to pulmonary venous congestion and, with increasing hypertrophy, inner layers of myocardium become ischaemic. The associated acceleration of atheromatous coronary disease contributes to the progression to left ventricular dysfunction and cardiac failure. Atheroma deposition is accelerated by hypertension, and causes the most numerous of the complications of coronary artery disease; it also is responsible for most of the cases of stroke. Aortic dissection is a consequence of vessel wall degeneration and cystic medial necrosis. A pathologically distinct process affects smaller vessels, particularly of the renal and cerebral circulations; this consists of medial and intimal proliferation. Renal impairment may follow, although renal failure is usually a complication of the fibrinoid necrosis of the resistance

vessels, which characterizes malignant or acceleratedphase hypertension.

SUITABILITY FOR TREATMENT At the highest levels of blood pressure (diastolic greater than 115 mmHg) the risks to the individual patient are so high that treatment is mandatory. However, confirming that it is beneficial to treat patients with diastolic pressures between 90 and 114mmHg has been more difficult. The combined results of randomized trials suggest that treatment has decreased the odds of stroke significantly, by about 40%, but only managed to reduce the risk of heart attack by 10%. These large trials have provided no evidence to favour any one particular drug regimen over any other. Patients with diastolic blood pressures above 105 mmHg are recommended for treatment, particularly if they demonstrate any ventricular hypertrophy or renal impairment. For the 10-15% of the middle-aged population with diastolic pressures between 90 and 104 mmHg a more complex strategy is necessary. This is because halving the annual risk of a stroke, from 0.1 to 0.05% in an asymptomatic middle-aged man may not be worthwhile, particularly when set against the requirement for lifelong treatment. However, if the individual has already suffered a complication, such as a transient ischaemic attack, halving the annual stroke risk from 20 to 10% is a worthwhile goal. A strategy for patients with diastolic blood pressure between 90 and 104 mmHg should aim to identify those with the highest risk of complications. Treatment begins with measures to reduce associated risk factors, such as smoking, high blood lipids and obesity. Drug therapy of mild hypertension should therefore be reserved for those in whom all other measures have failed and those who carry the greatest risk of progression to severer forms of blood pressure or cardiovascular complications.

Systolic hypertension Although epidemiological studies have shown that systolic blood pressure is at least as good a predictor of subsequent complications as diastolic pressure, there is less information on the levels of systolic pressure that should be treated. In most cases, systolic and diastolic pressures follow each other closely. However, difficulties arise in elderly patients, where isolated systolic hypertension is more common. Precise guidelines are not available, but the fears that treatment might reduce cerebral blood flow and worsen cerebrovascular risks have not been borne out. Reduction of high systolic pressures (greater than 180 mmHg) might therefore be expected to accrue benefits by reducing the stroke rate, and this indeed has been the finding in the few studies of elderly hypertensive patients.

METHODS OF TREATMENT The importance of attention to associated risk factors, such as smoking and hyperlipidaemia, has already been emphasized, and is an important part of all treatment strategies. Blood pressure-lowering drugs may not be needed if other methods are successful.

Non-pharmacological methods Drug withdrawal. Withdrawal of drugs that may contribute to hypertension, e.g. NSAIDs, steroids and oestrogencontaining compounds such as the oral contraceptive pill, should be considered where possible. Patients with a high alcohol intake should be advised to abstain or reduce their intake. Dietary measures. For each kilogram loss of weight, blood pressure can be expected to fall by 2.5-3.0mmHg systolic and 1.5-2.3 mmHg diastolic. In general, salt restriction has not been found to be an effective treatment for hypertension at the levels of intake achievable with a tolerable western diet. However, high salt intakes can negate

the effectiveness of the thiazide diuretics used in treatment, so it is important to ensure patients moderate their intake. Relaxation. Teaching patients to relax - using formal programmes administered by non-medical personnel in general practices and continued at home by the patient has been shown to produce a modest sustained blood pressure reduction. Exercise. Aerobic exercise maintained for some 15-20 minutes will lower subsequent blood pressure readings by 5-10 mmHg for about 12 hours. Exercise may also improve associated cardiac risk factors, including obesity.

12

Drug treatment of essential hypertension The principles of drug treatment for hypertension are theoretically simple. Therapy with a single drug given once a day is the ideal, but combination treatment is frequently necessary. A step-by-step approach is frequently used, starting with a first-line preparation alone, and then in combination with another, before third- and fourth-line drugs are used (Table 12.45).

TABLE 12.45 Indications for specific antihypertensive drugs and their common side-effects Indication

Advantages

Symptomatic side-effects

Other side-effects

First line or adjunct to p-blockers or Ca antagonists

Effective, cheap, usually once-daily treatment

Gout, impotence

Spironolactone Amiloride Loop

First line in Conn's or adrenal hyperplasia In renal disease

Potassium-sparing

Gastrointestinal upset

Powerful diuretics

Urinary urgency

Hypokalaemia Glucose intolerance Lipids Hyperkalaemia in renal impairment As for thiazides

B-Adrenoreceptor blockers

First line

Antianginal and antiarrhythmic

Asthma in some subjects, tiredness, claudication and cold hands/feet

? lipid effect LDL HDL

Calcium antagonists

First or second line

Antianginal and antiarrhythmic

Flushing, oedema, gastrointestinal upset

Heart block and negative inotropy with some examples

Angiotensin-convertingenzyme inhibitors

First line in renal disease First line in heart failure

Well tolerated and very powerful; useful in heart failure

Cough, postural hypotension

Profound hypotension Renal impairment in renovascular disease

Angiotension II receptor antagonists

Second or third line

As ACEI, but no cough

Dizziness

As ACEI, hepatic impairment

a-Adrenoreceptor antagonists

First or second line

Largely superseded by Ca antagonists and ACE inhibitors

Postural hypotension

-

Can be used alone effectively

Sedation, impotence

Haemolytic anaemia

Hydralazine

Second/third line Methyldopa in pregnancy Third line

Headache

SLE-type reactions

Diazoxide

Third line

Oedema

Glucose intolerance

Minoxidil Bethanidine/guanethidine

Third and fourth line Fourth line

Largely superseded by Ca antagonists and ACE inhibitors Largely superseded by Ca antagonists and ACE inhibitors Extremely powerful vasodilator Rarely indicated

Hirsutism, oedema Postural hypotension/impotence

Fluid retention Hypotensive infarction

Type of drug Diuretics Long-acting thiazides

Others Methyldopa/clonidine

597

Diuretics The long-acting thiazide diuretics such as chlorothiazide and bendroflumethiazide (bendrofluazide) are effective, although symptomatic side-effects such as impotence and gout can occur. They are less powerful in renal failure, when loop diuretics such as frusemide and bumetanide have an advantage. Thiazide diuretics cause biochemical changes which have given concern about their long-term safety. In particular, mild hypokalaemia is common and may be associated with a risk of cardiac arrhythmias. Hyperuricaemia may produce symptomatic gout and, in conjunction with glucose intolerance and increases in cholesterol and triglyceride concentration, an overall worsening of the cardiovascular 'risk factor profile' is produced by these drugs. It is not at all clear whether these concerns are justified, although the lack of improvement in the incidence of myocardial infarction in the vast majority of hypertension treatment trials has been attributed in part to these unwanted biochemical effects of the thiazide diuretics. B-Adrenoreceptor blockers Chance clinical observations led to the recognition that the B-adrenoreceptor blockers are effective antihypertensives. These drugs are specifically contraindicated in obstructive airways disease, because of the production of asthma even by drugs with supposed cardiac selectivity. Other troublesome side-effects are related to the B-receptor blocking action, with bradycardia, tiredness, claudication and cold limbs being prominent. The demonstration of their protective effect following myocardial infarction, as well as their antiarrhythmic and antianginal actions, make B-blockers a rational choice when CHD complicates hypertension. They have to be used with great caution, if at all, in heart failure. Plasma lipids are altered with B-blockade, potentially worsening the cardiovascular risk. However, whether these alterations are deleterious in the long term has not been clarified. B-Blockers with diuretics have become established as the mainstays of the management of hypertension. There are many examples of these drugs. In the UK, atenolol, metoprolol, propranolol, bisoprolol and pindolol are commonly used examples. Calcium antagonists The antihypertensive effectiveness of calcium antagonists is related to the fall in peripheral vascular resistance that follows resistance vessel vasodilatation. Side-effects of nifedipine can be attributed to its vasodilator activity and include flushing, headache and oedema; in contrast, verapamil has more gastrointestinal side-effects and ought not to be combined with B-blockade because of its propensity 1

598

MCQ 12.24

to produce heart block and impair ventricular contractility. Diltiazem fills an intermediate position, with fewer peripheral effects and more cardiac effects than nifedipine, but fewer myocardial effects than verapamil. Although these drugs have not been subjected to close analysis in large trials, there is no doubt of their efficacy and popularity, even as a first-line therapy. Angiotensin-converting-enzyme inhibitors The ACE inhibitors were specifically manufactured to inhibit the renin-angiotensin system by blocking the conversion of angiotensin I to its active form, angiotensin II. These drugs are subjectively very well tolerated, and their associated efficacy, in all but the groups of patients with renal arterial stenoses, has led to their adoption as the first of the second-line treatments to be used in the more severe degrees of hypertension. Cough is a common (>10%) sideeffect of ACE inhibitors, but at higher doses severe neutropenia, rashes and taste disturbance have been seen with captopril. Their hypotensive effect may be severe in relatively hypovolaemic patients and in those with hyponatraemia (which might occur in those taking diuretics). ACE inhibitors will cause a deterioration in function or renal failure in renal arterial stenosis. Methyldopa In pregnancy methyldopa still maintains a place in the treatment of hypertension, but the associated CNS sideeffects have made it less popular now that less sedative preparations are available. Sympathetic ganglion blocking drugs In the past, the sympathetic ganglion blocking drugs (guanethidine and bethanidine) were important as early effective treatments, but their use is now limited by their troublesome postural hypotensive effects, lack of supine blood pressure control and the sexual dysfunction that they produce in men. Drug selection in therapy Despite decades of antihypertensive drug therapy, the only drugs that have been shown to improve prognosis are thiazide diuretics and B-blockers. In newly diagnosed hypertensive patients without contraindications, these drugs should be considered as first-line therapy. Newer agents have numerous claimed advantages, but whether these will translate into improved prognosis is uncertain. Their use may be more sensibly reserved for patients with contraindications for, or intolerance of, diuretic or B-blocker usage, unless they can provide specific advantages, as shown in the following conditions: • Patients with angina benefit from antihypertensive drugs which are also antianginal. • Diabetic patients with proteinuria may benefit more with ACE inhibitors as an antihypertensive agent. • Patients with systolic heart failure have a better

• •

•

•

prognosis with ACE inhibitor or vasodilating B-blocker therapy. Patients with significant left ventricular hypertrophy or diastolic ventricular failure may benefit more from ACE inhibition as antihypertensive therapy. Thiazide diuretics may worsen diabetic control and (3blockers may blunt the sensation of hypoglycaemia, and so these drugs should be reserved as second- or thirdline antihypertensive therapy in diabetics. Thiazide diuretics may result in hypokalaemia and/or hypomagnesaemia, which exacerbate or precipitate serious cardiac arrhythmia, and should be avoided in susceptible patients. In black hypertensives, (3-blockers and ACE inhibitors are less effective as monotherapy than diuretics and calcium antagonists.

Combination therapy Adequate control of hypertension with monotherapy should be sought before any combination therapy is considered. Once it has been established that monotherapy is inadequate, then combination therapy may be required. The choice of combinations is largely empirical in order to attain acceptable blood pressure levels, but, from published trials, p-blockers plus dihydropyridine calcium antagonists, ACE inhibitors plus diuretics or calcium antagonists appear to offer advantages, whereas [3-blockers plus ACE inhibitors or diuretics plus calcium antagonists appear less efficacious. Combination of a (5-blocker and verapamil should be avoided as the combination is very negatively dromotropic and inotropic. Both diuretics and ACE inhibitors are liable to alter electrolyte balance, and there is no standard formula to calculate combination dosage - there is no substitute but to monitor serum electrolytes serially.

Drug treatment of hypertension in pregnancy Pre-eclampsia The treatment of pre-eclampsia (p. 592) requires delivery of the fetus. Nevertheless, drugs may be needed to control the blood pressure before delivery can be achieved. Methyldopa (250-500 mg, 6-hourly by mouth) is used and, if necessary, can be given parenterally. It has the advantage of being effective and harmless to the fetus, although its sedative side-effects can confuse neurological assessment and its onset of action is slow (1 hour). Hydralazine, orally (<300mg daily), by intramuscular injection (1020 mg, 3-6-hourly) or by constant infusion (5-10mg/h) is a suitable alternative, although flushing, tachycardia and vomiting may be side-effects at the higher doses. Given intravenously, it has a rapid onset of action which can be maintained by a constant infusion. p-Adrenoreceptor blockers and calcium antagonists have been found to be effective in the acute situation, although experience with these drugs and follow-up of the delivered neonates has been shorter than for methyldopa and hydralazine. Diuret-

ics are used only when cardiac failure complicates the picture.

12

Chronic hypertension in pregnancy Because of its known lack of effects on the fetus, methyldopa is the drug of choice; hydralazine is the usual alternative, although (5-adrenoreceptor blockers are being used increasingly without undue hazard. Mothers who are known moderate hypertensives may be advised to cease treatment at the time of conception, although no teratogenicity has been associated with any of the commonly used antihypertensive drugs. O

Drug treatment of hypertensive emergencies Malignant or accelerated-phase hypertension This previously rapidly fatal condition is now managed by bed rest with admission to hospital and oral treatment. Parenteral treatment is avoided unless coma, fits or severe heart failure are present. Under these circumstances an effective treatment is an intravenous infusion of labetalol, starting at 25mg/h and increasing to 120mg/h, with additional diuretic treatment for pulmonary oedema. Intravenous nitrates are also effective in this situation. Hydralazine given intramuscularly has been a popular treatment, but the 20-minute delay in onset of action may lead to additional doses or drugs being given, resulting in hypotension some hours later. Methyldopa, even when given intravenously, is also too slow to be used effectively. For cases caused by phaeochromocytoma, phentolamine (2.5-10mg i.v. bolus or Img/min by i.v. infusion) produces a rapid pressure reduction by cc-adrenoreceptor blockade, but such a response is not limited to this disease. Oral treatment is optimal in most cases of malignant and accelerated hypertension, and any of the usual antihypertensive drugs can be used successfully. Recent observations show that sublingual nifedipine (10-20 mg) produces a smooth blood pressure reduction beginning within 10 minutes, although the reflex tachycardia it produces may require treatment with a (3-adrenoreceptor blocker. Additional diuretic treatment is often given. The degree and rapidity of blood pressure reduction are important. In chronic hypertensives, cerebral autoregulation is lost at mean blood pressures above ISOITS mmHg and below 90-120 mmHg. Treatment should reduce mean blood pressure to 100-120 mmHg and no lower. This should be achieved over a 12-36-hour period to allow time for the relatively slow process of reestablishing normal cerebral autoregulation. Aortic dissection Aortic dissection is an emergency requiring prompt reduction of mean blood pressure to 100-110 mmHg; this is usually achieved with an infusion of sodium nitroprusside, starting at 12.5mg/min and titrating the dose against

599

blood pressure every 2-3 min. Infusion should not continue beyond 24 hours, but the resulting thiocyanate accumulation can be diminished by concomitant use of hydroxycobalamine by injection. There may be an advantage in reducing the systolic rate of change in aortic pressure as well as mean pressure. This can be achieved with Padrenoreceptor blockade, so that labetalol infusion with its mixed, though predominantly p-receptor, effects has been advocated for use in dissection.

Hypertension in chronic renal failure A proportion of patients with chronic renal failure have hypertension; in these, blood pressure may be controlled by dietary sodium restriction and fluid removal at haemodialysis. Drug treatment is necessary in some cases, and the excretory metabolism must be taken into account when dosages are calculated. Treatment is generally similar to that in essential hypertension.

Hypertension in phaeochromocytoma Hypertensive crises in phaeochromocytoma can be treated conventionally, but it is important to avoid (3adrenoreceptor blockade before full a-adrenoreceptor blockade has been achieved, in order to avoid unopposed a effects exacerbating hypertension. Parenteral phentolamine can be used acutely, and longer-term control gained with phenoxybenzamine (10-30mgt.d.s.), with the later addition of small doses of a [3-blocker, such as propranolol (40mgt.d.s.). If a surgical cure cannot be effected, longterm treatment with oc-methyltyrosine can be used, although side-effects are often troublesome.

Adrenocortical hyperplasia and Conn's syndrome

VASCULAR DISEASE AORTIC ANEURYSM Most intrathoracic aortic aneurysms are asymptomatic unless they become large enough to produce pressure on a main bronchus, on the phrenic nerve or if they produce aortic regurgitation. They may rupture and result in sudden death. They are frequently diagnosed on chest X-ray. They are a common manifestation of collagen diseases such as Marfan's syndrome, and are a complication of atherosclerosis and hypertension.

ACUTE AORTIC DISSECTION Dissection occurs when there is haemorrhage into the wall of the aorta, causing the media to separate from the other layers (Fig. 12.70). Most dissections are seen in middleaged or elderly men with arteriosclerosis and hypertension. More than 50% of patients who present with dissection die within 24 hours. It is a particular complication of Marfan's syndrome, where cystic medial necrosis is common. In this disorder progressive aortic dilatation leads to aortic incompetence and dissection with a high mortality. Dissections can be classified clinically according to the part of the aorta that is affected. When the ascending aorta is involved, it is termed a type A dissection (De Bakey types I and II), whereas those where only the aorta beyond the left subclavian origin is involved are called type B dissections (De Bakey type III).

Clinical features The most prominent symptom is a very sudden onset

Surgical removal of the adenomas in Conn's syndrome may not always cure the electrolyte disturbances and hypertension. In these circumstances, as in idiopathic hyperplasia, treatment with amiloride (20-70 mg daily) or spironolactone (200-400 mg daily) is effective.

FURTHER READING ON SYSTEMIC BLOOD PRESSURE AND HYPERTENSION Guidelines Subcommittee 1993 Guidelines for the management of mild hypertension: memorandum from a World Health Organization/International Society of Hypertension meeting. J Hypertens 11: 905-918.

600

1

Figs 12.32-12.34

3

MCQ 12.25 4

2

Fig. 12.35

Fig. 20.36

FIG. 12.70 Sites of dissection in aortic aneurysm

of excruciating pain, often described as tearing in nature. Unlike infarction pain, the pain is often totally unheralded and at its worst initially, but may uncommonly develop to a pulsating discomfort that persists. The pain may radiate through to the back and, in type B dissection, to the left shoulder. Typical myocardial infarction pain may also coexist if the coronary arteries are involved by the dissection.

Diagnosis and investigation The diagnosis can be difficult, and its importance has considerably increased with the emphasis on early acute treatment of myocardial infarction with thrombolytic drugs. Clearly, these may have terrible consequences in dissection. The patient is frequently hypertensive or has a history of hypertension, which may or may not have been treated. Patients with Marfan's syndrome are usually easy to identify but form only a small proportion of cases. In about 10% of patients a difference in blood pressure (20mmHg or more) in the two arms may be found, and in some a prominent pulsation is visible at the root of the neck. Auscultation may reveal aortic regurgitation in type A dissection, which is typically more obvious at the right sternal edge. The ECG may be normal except in those cases with associated coronary artery involvement. In some individuals the chest X-ray provides the first clue to the diagnosis, with mediastinal widening, left-sided pleural effusions and expansion of the aortic knuckle border beyond a calcified rim of endothelium. 1 The diagnosis is confirmed at cardial catheterization by contrast aortography. Coronary angiography is also performed to allow surgical repair. CT is helpful and less invasive, often being performed prior to catheterization but rarely supplanting it totally. 2 MRI is rapidly replacing CT. Echocardiography by the standard methods may demonstrate a proximal ascending aortic dissection, but transoesophageal echocardiography is the only ultrasound method able to adequately demonstrate the descending thoracic aorta. Transoesophageal colour Doppler is particularly useful, as the flow patterns make the intimal tear more obvious.

Complications Complications of the dissection include occlusion of arterial branches, including the coronary arteries, head and neck vessels, and indeed all distal arterial branches. Stretching of the aortic annulus in type A dissection is responsible for aortic regurgitation, which may be severe and lead to left ventricular failure and shock. Progressive dilatation of the aorta may obstruct the airways or oesophagus. Finally, rupture of the aortic wall may occur, producing pericardial or pleural effusion and, occasionally, intrabronchial haemorrhage. This complication frequently, but not always, causes death.

Management Treatment is usually surgical for type A dissection and many cases of type B. However, acute management includes blood pressure reduction by parenteral treatment, preferably with drugs reducing the inotropic force of the heart, such as labetalol by infusion. Pain control and sedation are also very important in this very distressing condition, where the outlook is often very poor. 3 4

12

ARTERITIS Arteritis is an inflammatory process involving the artery wall. It can occur in vessels of various sizes. If these are small, the condition is called necrotizing vasculitis and is found in a large number of 'autoimmune' conditions (discussed in Chapter 22). Involvement of the large arteries is much less common and occurs principally in Kawasaki's disease, Takayasu's (or pulseless) disease and giant cell (or temporal) arteritis.

KAWASAKI'S DISEASE Kawasaki's disease is a disease of children described originally in Japan but now reported from other parts of the world. The aetiological agent is unknown, but as the disease occurs in cyclical epidemics an infective agent is suspected. It was originally called 'mucocutaneous lymph node syndrome' because of the prolonged fever, skin rash and marked cervical lymphadenopathy. Later it was realized that there is a high incidence of coronary artery aneurysms, occlusions and myocardial infarction which, because many of the patients are infants, were originally missed. Diagnosis of these lesions can often be made noninvasively by ECG and 2D echocardiography. Management is based on the use of antithrombotic agents and steroids, but aspirin alone has been found to result in a lower incidence of coronary artery complications than some other regimens.

TAKAYASU'S ARTERITIS

Takayasu's arteritis - a disease which predominantly affects oriental women - was first described by an ophthalmologist whose first case was a young woman with cataracts and unusual arteriovenous anastomoses around the optic papillae. It was later realized that similar ophthalmic findings were seen in patients with absent radial pulses, hence its alternative name, 'pulseless disease'. The disease can be classified into three types: • Type I involves the aortic arch and its branches, producing diminished or absent pulses over the upper body (so-called 'reversed coarctation').

601

• Type II involves the descending aorta, particularly the renal arteries and vessels supplying the abdominal viscera. • Type III is a combination of I and II. Takayasu's arteritis has also been implicated in the development of primary pulmonary hypertension. The aetiology is unknown but the disease commonly starts in teenage girls, with a sex ratio of 8:1.

Clinical features There is often a systemic illness, with fever, arthralgia and fatigue. There may be pain and tenderness over the affected arteries. This systemic phase passes and is followed by a latent period. Some time later these patients, and others who had no systemic prodrome, present with the typical features of weak or absent pulses, low or unrecordable upper body blood pressure, hypertension, heart failure and ocular damage. The hypertension is produced by an increased aortic impedance secondary to major vessel occlusion or narrowing, and also from renal artery stenoses causing renal hypertension. Bruits over affected vessels are common. Heart failure is usually secondary to hypertension, but coronary artery ostial stenoses can occur, causing angina; aortic regurgitation has also been described.

Management and prognosis The acute systemic illness is alleviated by corticosteroids, which may reduce the incidence of sequelae. Antiplatelet agents, such as aspirin and dipyridamole, and warfarin have been used to reduce the frequency of cerebral transient ischaemic attacks and possibly retard the progression of arterial lesions. Morbidity is chiefly from heart failure and cerebrovascular events. Surgical treatment of severe vascular stenoses may become necessary.

GIANT CELL (TEMPORAL) ARTERITIS

602

Although predominantly affecting the cranial arteries, giant cell arteritis can, like Takayasu's arteritis, involve the aorta and its proximal branches. The disease is described more fully in Chapter 22, page 1183. Where major vessel arteritis occurs, it may complicate the more common presentation with cranial arteritis and polymyalgia rheumatica, or may occur alone. Symptoms are similar to those of Takayasu's arteritis, including angina and exercise-related ischaemia of the upper and lower limbs. Aortic regurgitation, aneurysm and dissection can complicate giant cell arteritis of the aorta. As for the commoner forms of the disease, treatment is with corticosteroids.

PERIPHERAL VASCULAR DISEASE ARTERIAL DISEASE Thoracic outlet syndromes In the past, the thoracic outlet syndromes were thought to be due principally to arterial compression of the subclavian artery but it is now appreciated that many of the symptoms are caused by compression of the brachial plexus as well. This is from an anomalous fibrous band that stretches from a cervical rib to the first rib and, in doing so, compresses the neurovascular bundle. The patient has symptoms in the hand, which appear promptly when the arm is abducted to 90° and externally rotated. A bruit can then be found over the supraclavicular fossa. These features disappear when the arm is returned to a relaxed position. Aneurysm of, and embolism from, the subclavian artery is a comparatively rare complication of this condition, which can lead to digital gangrene.

Peripheral arterial disease and arteriosclerosis Arteriosclerotic arterial disease can be from atheroma, focal calcific sclerosis (Monckeberg's arteriosclerosis) or arteriosclerosis. Monckeberg's sclerosis leads to calcification of the media without narrowing of the vessels, but is often associated with arteriosclerosis and luminal narrowing. Most symptomatic disease is due to atheromatous plaque formation, with subsequent stenosis, occlusion or embolization of the vessel. Haemorrhage into an atheromatous plaque may precipitate a sudden vessel occlusion. Emboli may arise from atheromatous plaques (as occurs frequently from the carotid); however, major emboli are more frequently from the heart, in such conditions as mitral valve disease, left ventricular aneurysm, mural thrombus following myocardial infarction, left atrial myxoma, endocarditis, and from any enlarged heart, particularly if there is atrial fibrillation. The sites of predilection for the atheroma are not fully explained, and any one patient may have principally coronary, cerebrovascular or peripheral vascular symptoms, with risk factors very similar to another patient whose symptoms are quite different. Atheroma tends to develop in the abdominal aorta and at the origins of major branches. These may be sites of mechanical stress or low shear.

Intermittent claudication Ischaemia of the lower limbs is usually accompanied by effort-related cramp in the calves, thighs and buttocks, which disappears with rest. The site of claudication indicates the probable site of stenosis or occlusion - the higher

the claudication the higher the atheroma is likely to be found. The walking distance on the flat is usually remarkably repeatable. Claudication is often accompanied by profound weakness as well as cramp in the affected muscles, which have become acidotic from anaerobic exercise. The combination of hip and thigh claudication with impotence in males (the Leriche syndrome) is due to severe atheromatous occlusion of the lower end of the aorta. More severe ischaemia produces a continuous pain at rest, particularly in bed. The pain is most frequently in the toes or foot, with some paraesthesia. The foot is often hung over the edge of the bed and may be subject to ulceration and gangrene. In diabetics with peripheral neuropathy there may be no pain but only ulceration and gangrene. Diagnosis and investigation Peripheral vascular disease is assessed by a careful history and clinical examination for signs of ischaemia, absent pulses, audible bruits over the proximal arteries, poor skin nutrition, low skin temperature and loss of hair, and then by blood pressure measurement. This is usually done using Doppler ultrasound on the ankle arteries with a proximal cuff. The normal ankle systolic pressure is usually higher than brachial (by about 20%) and should be maintained after exercise. The ischaemic leg often has an ankle pressure of about 50% of brachial at rest, which becomes unrecordable after exercise when the blood is diverted to areas of vasodilatation elsewhere. The severity of the hypotension is related to the degree of arterial disease. Management Most patients with claudication benefit from medical measures such as weight reduction, complete cessation of smoking and an energetic training regime that increases the claudication distance. Vasodilators and viscosityreducing drugs have a limited role. Surgical intervention after angiography now includes angioplasty, as well as endarterectomy and bypass grafting. It is reserved for more severe cases.

Gangrene Sudden ischaemia of a limb (from trauma, arterial thrombosis or embolism) leads to pain, pallor, coldness, and then numbness of the limb distal to the occlusion. If perfusion is not restored, sensation is severely diminished in an hour; within 6 hours ischaemic contracture of the muscles has started. This is followed by subcutaneous haemorrhage and focal gangrene. Fixed staining of the skin indicates death of the underlying tissues. Persistent pain and numbness without loss of sensation indicate a collateral flow. Managemenf This is an emergency and treatment is surgical, by embolectomy, emergency bypass grafting or eventual amputation.

Embolectomy under local anaesthetic can often save a limb. Heparin is used to prevent further clotting, particularly if there is an obvious embolic source.

12

Buerger's disease Buerger's disease was previously thought to be a distinct disease, but is now considered to be an accelerated form of atheroma affecting young men of 20-40 years of age who smoke (or chew) tobacco heavily. It affects the small arteries of the hands and feet, with an intense inflammatory response, and often involves the veins as well. Hand and instep claudication may occur, with loss of ankle and wrist pulses and gangrene of the toes and fingers. There is often an associated Raynaud's phenomenon (see below) and a migratory nodular phlebitis. Cessation of smoking or chewing tobacco is imperative. Prognosis For patients with symptoms of limb ischaemia, the rate of limb loss is about 2% per year for non-diabetics and 7% per year for diabetics.

Raynaud's syndrome and phenomenon Raynaud's syndrome and phenomenon are vasospastic conditions affecting mainly the hands; they are often precipitated by cold and sometimes by motion, although random spontaneous attacks are common. Attacks are characterized by episodes of arteriolar spasm; these produce cold pale fingers and hands, progressing over minutes or hours to cyanosis and ending with a red flushing phase. When the condition is not associated with any underlying disorders it is termed Raynaud's syndrome, and skin changes are not seen. This is commoner in young women. Raynaud's phenomenon occurs secondary to underlying systemic conditions, the most common being scleroderma. Other collagen diseases may be associated with Raynaud's phenomenon, as may thoracic outlet compression, arteriosclerotic vascular disease, vibration trauma in pneumatic drill users, and serum protein abnormalities.

Mesenteric ischaemia Mesenteric (or small bowel) ischaemia is described in Chapter 15, page 795.

VENOUS DISEASE Varicose veins are the commonest vascular disorder affecting the legs, and venous thrombosis is a frequent complication of many medical and surgical conditions.

Varicose veins Varicose veins are dilated and tortuous veins that vary in size from large palpable bunches to mere discoloured superficial spider bursts on the skin. They arise either

603

through primary abnormalities in the venous wall leading to dilatation and incompetence of the valves, or secondarily after damage caused by previous thrombosis. In either case, factors that impede venous flow or increase venous pressure in the legs contribute to their development, as does the prolonged standing required in certain occupations. Many patients suffer no symptoms but the veins can be responsible for discomfort, usually described as heaviness or pulling. Oedema may occur and, in chronic severe disease, pigmentation of the skin, a disfiguring dermatitis, ulceration or, at its severest, gangrene may follow. These complications are more prominent in the areas subject to the highest venous pressures, where skin nutrition is consequently poorest, namely in the lower third of the leg, often more marked medially. Management Elastic stocking support, injection of sclerosants and surgical removal (or stripping) of the veins are used to treat the varicose veins. Relief of exacerbating factors, such as obesity, ascites and right heart failure, is also important. Ulcers often heal slowly, and must be kept clean and uninfected. Elevation of the limb aids healing.

Venous thrombosis Occlusion of veins by thrombus is very common. It leads to considerable morbidity and, when the deep veins of the leg and pelvis are involved, to the risk of serious illness or death through pulmonary embolism. Superficial venous thrombosis Superficial venous thrombosis causes pain and swelling over the affected vessel, which is often palpable as a tender cord or small nodule. Unless complicated by progression to deep vein thrombosis (DVT), superficial venous thromboses are essentially benign and require only symptomatic treatment. NSAIDs, such as indomethacin or aspirin, plus local applied heat, are of help. Deep venous thrombosis The incidence of DVT is low in physically active subjects. The predisposing factors are divided into: • Transient/reversible factors - surgery, trauma, bed rest, prolonged immobility (e.g. long-haul airflights), pregnancy; • Chronic factors - venous disease, hypercoagulability of blood, active or occult malignant disease, pelvic masses and marked obesity; • Idiopathic - with no known risk factors.

1

604

MCQ 13.1

The risk of recurrence of DVT is lower in patients with a reversible risk factor than in those with chronic risk factors or idiopathic DVT. The risk of pulmonary embolism (PE) is low in patients with DVT confined to the calf veins. It is the proximal DVT involving veins in the thigh and pelvis that are potential sources of large pulmonary emboli. DVT of the axillary or internal jugular veins is much less common and rarely gives rise to PE. Superior vena caval thrombosis can lead to PE, and a known precipitating factor is the presence of indwelling central venous catheters (e.g. for parenteral nutrition). Clinical features and diagnosis The diagnosis of DVT is often difficult. The main features are pain and swelling of the affected limb. The symptoms vary with the extent and type of vein involved and are often absent. Dorsiflexion of the foot may produce calf tenderness (Homan's sign) in the extended leg, but this sign is often not very helpful. The most useful diagnostic test is the venogram which, when technically adequate, will usually establish the presence and extent of any thrombus. Doppler ultrasound velocity profiles can be useful, although only for DVT affecting veins above the knee. The major conditions which may mimic DVT include muscle trauma, rupture of a popliteal (Baker's) cyst (which resembles calf vein DVT) and sciatic nerve irritation. Complications The most serious complication in DVT is PE (Ch. 13, p. 675). This is usually a complication of thrombus above the knee, but sometimes occurs with calf vein thrombosis. Other complications include vein wall and valve damage leading to long-term venous insufficiency. Management Treatment of DVT is by intravenous heparin bolus (5000 U) followed by a constant infusion (24000^8000 U/24h) sufficient to maintain a partial thromboplastin time two to three times normal. Oral anticoagulation with warfarin is then instituted, with the heparin being discontinued once a prothrombin ratio of 2.5-3.0 is achieved. Oral anticoagulation is continued for 6 weeks in patients with transient/reversible predisposing factors, and for 6 months in those with idiopathic DVT. It may need to be indefinite in patients with known active malignancy, other bedridden conditions, recurrent DVT or haematological hypercoagulable states (e.g. thrombophilia). Recurrent subacute pulmonary embolism leading to chronic pulmonary hypertension is associated with a poor prognosis, and may require insertion of an inferior vena caval filter; such patients also need indefinite anticoagulation. It has become clear that prophylactic treatment with relatively low doses of anticoagulants (heparin by subcutaneous or intramuscular injection: 5000 U 8-hourly or enoxaparin 20mgs.c. o.d.) can prevent DVT, and hence PE, in patients at risk, such as those on prolonged bed rest after major surgery.

13

FURTHER READING British Thoracic Society 2002 The burden of lung disease (www.brit-thoracic.org.uk)

Respiratory

STRUCTURE AND FUNCTION OF THE RESPIRATORY SYSTEM 1

Disease John Moxham and John F Costello

Structure and function of the respiratory system 605 Symptoms and signs of respiratory disease 611 Examination of the respiratory system 619 Imaging the thorax 623 Bronchoscopy and lung biopsy

Respiratory failure 670 Lung transplantation 674 Pulmonary hypertension 674 Pulmonary Ihromboembolism 675 Sarcoidosis 682 Connective tissue and related diseases and the lung 689

627

Vasculitides and the lung 690

Respiratory infections 631

Pulmonary arteriovenous malformations and shunts 691

Lung abscess 647 Bronchiectasis 648 Cystic fibrosis 649

Pulmonary fibrosis and diffuse interstitial lung disease 692

Asthma 651

The acute respiratory distress syndrome 697

Pulmonary eosinophilia 659

Lung tumours 700

Chronic bronchitis and emphysema 661

Pleural disease 710

Respiration during sleep 668

Respiration can be defined as those processes concerned with gas exchange between an organism and its environment. In humans respiration begins with the uptake of O2 and ends with the elimination of CO2. The adequacy of gas exchange by the lungs depends on four closely integrated processes:

Neuromuscular and skeletal disorders 718

Respiratory diseases are responsible for a large proportion of premature mortality and serious morbidity among the population, and at the same time for a large amount of minor illness. In England and Wales respiratory diseases (including lung cancer) account for approximately a quarter of all deaths. They are the second most common cause of death after cardiovascular disease. Much morbidity and many deaths are smoking related. Unfortunately, despite an overall decline in numbers smoking, the prevalence of smoking in young people in the UK is on the increase, as is that in many developing countries.

• • • •

Ventilation Diffusion of gases Pulmonary capillary blood flow The carriage of gases by the blood.

In addition, breathing is regulated in accordance with the constantly changing metabolic needs of the body by an intricate control system. The anatomical organization of the human lung is eminently suited to its role as an exchanger of O2 and CO2. A series of branching airways conducts air into and out of the alveolar spaces. Similarly, the pulmonary circulation delivers blood to, and drains blood from, the pulmonary capillaries. Ninety per cent of the alveolar surface (80m2 in an adult lung) is covered with capillaries, providing a thin but large film of blood for exposure to the gases in the alveolar spaces. The respiratory system has other functions besides gas exchange: • The upper airways filter and humidify inspired air. • It provides an elaborate defence mechanism that protects the body against inhaled pathogenic microorganisms and toxic chemicals. • The pulmonary blood vessels serve as a reservoir for the left side of the heart. • It provides O2 and substrates for its own metabolic needs. • The lung modifies the pharmacological properties of a variety of drugs.

Ventilation Ventilation is the cyclical movement of ambient air into and out of the lungs, and the distribution and mixing of that air within the lungs. This cycle is accomplished by contraction and relaxation of the respiratory muscles and the elastic recoil of the lungs. During normal quiet breathing, about two-thirds of the inspired gas reaches perfused alveoli; this fraction is known as alveolar ventilation (VA). The rest of the inspired gas fills the conducting airways and

605

alveoli that may not be well perfused, and is known as physiological dead space (VD) or 'wasted ventilation'. Therefore, total ventilation (usually in L/min) = VA + VD. The commonly measured static lung volumes are shown in Figure 13.1. The static volumes depend on the distensibility (i.e. compliance) of both the lung and chest wall, and also the capacity of the respiratory muscles, often impaired in neurological disorders, such as motor neuron disease. The distensibility or compliance of the lung is affected by the elastic recoil due to the connective tissue elements in the parenchyma and the surfactant that lines the alveolar surface. Thus the compliance of the lung is decreased in diffuse pulmonary fibrosis (tissue changes) and neonatal respiratory distress syndrome (reduced surfactant), and increased in emphysema because of the destruction of lung elastic connective tissue. Reduced compliance of the chest wall (which includes the thoracic cage and the abdominal

wall) can be caused by bony ankylosis (ankylosing spondylitis), pleural thickening, obesity and ascites.

Diffusion of gases The term 'diffusion' describes the passive movement of O2 and CO2 between alveolar gas and pulmonary capillary blood (Fig. 13.2). It depends on three major factors: • The matching of ventilation and blood flow • The diffusion of gas from the alveolus to the red blood cell • The availability of haemoglobin and the chemical reactions within the red blood cell. Although O2 is the physiological gas that is most likely to be affected by abnormalities of diffusion, the diffusion capacity of the lung is usually measured with a nonphysiological gas, carbon monoxide (TLCO, or transfer factor). If the haemoglobin concentration is normal, a decreased transfer factor means either a reduction in the

Total lung capacity (L)

Vital capacity (L)

Residual volume (L)

3.4-5.9 2.4-4.3

1.1-2.3 1.0-2.0

4.8 -7.9

2.7-5.3 2.0-4.0

1.4-2.6 1,1-2.2

4.5-7.6 3.4-6.0

2.4-4.7 1.9-3.6

1.7-2.7 1.3-2.4

4.3-7.3 3.3-5.8

Age 20 -39

Male Female

3.6-6.1

Age 40 - 59

Male Female Age 60+

Male Female

FIG. 13.1 The subdivisions of lung volume Functional residual capacity (FRC) is the resting expiratory level. (TLC = total lung capacity; TV = tidal volume; IC = inspiratory capacity; VC = vital capacity; RV = residual volume; IRV = inspiratory reserve volume; ERV = expiratory reserve volume.) Normal values are also shown.

1

606

MCQ13.2

FIG. 13.2 The terminal respiratory unit In the terminal respiratory unit A blood from the pulmonary artery is in close proximity to alveolar gas, facilitating gas exchange B and thus the uptake of 02 and the elimination of C02.

effective alveolar-capillary surface available for diffusion (thickening, loss of membrane, or inability to inflate part or all of the lung to expose it to inspired air, as in lobar collapse or respiratory muscle weakness) or obstruction or destruction of capillaries (decrease in capillary blood volume).

Carbon dioxide is transported in the blood as HCO3 in solution, or in combination with haemoglobin. The chemical reactions are reversed as blood flows through pulmonary capillaries, thus allowing CO2 to evolve into alveolar gas. Carbon dioxide forms carbonic acid in the blood by the reaction:

Pulmonary capillary blood flow The adequacy of pulmonary capillary blood flow depends upon the volume and distribution of right ventricular output. Because distribution is gravity dependent, in the upright position there is more blood flow to the bases than to the apices of the lung; in addition, local mechanisms exist to divert blood from poorly ventilated regions of the lung.

13

CO2 + H2O -> H2CO3 catalysed by carbonic anhydrase, and the carbonic acid then dissociates by the reaction: H2CO3 - H+ + HCO3Accordingly, a rise in arterial Pco2 tends to increase H+ and decrease pH; this mechanism is of fundamental importance in acid-base balance and the regulation of ventilation (p. 1114).

Carriage of gases by the blood Oxygen is transported in the blood mainly in combination with haemoglobin; a small amount is present in solution. The relationship between the percentage saturation of haemoglobin and the partial pressure (or tension) of O2 (Po2) in the blood is shown in Figure 13.3. The shape of the curve is important: blood is nearly fully saturated when the Po2 is greater than 8kPa (60mmHg), as it usually is in arterial blood; furthermore, when the prevailing Po2 is low, 5.3kPa (40mmHg), as it is in the tissues, haemoglobin readily gives up its O2, which is then available for metabolic activities. The position of the curve is not fixed, and shifts occur to the right (decreased affinity) and left (increased affinity) in many common clinical conditions. Shifts of the curve to the right, resulting from an increase in Pco2, H+ concentration, temperature and organic phosphates (particularly 2,3-diphosphoglycerate), are beneficial because they facilitate O2 transfer to the tissues and thus serve to maintain tissue oxygenation (see also p. 1203).

Control of ventilation The rhythmic pattern of breathing is driven by the spontaneous discharge of neurons located chiefly in the medulla. The activity of these cells is modulated by signals from the pons, hypothalamus and cortex, and from central chemoreceptors and receptors in the lung. Although the physiological mechanisms involved are largely unknown, there is a remarkable matching of metabolic activities and ventilation.

PATHOPHYSIOLOGY

1

It is convenient to consider generalized lung disorders under the following headings: • • • •

Airflow obstruction Restrictive lung disease Abnormalities of diffusion Abnormalities of pulmonary blood vessels.

These functional categories often overlap and, of course, do not account for localized abnormalities such as those caused by tumours, pneumonia or pulmonary infarction.

Airflow obstruction Airflow obstruction is common and occurs in acute and chronic asthma, chronic bronchitis, emphysema, cystic fibrosis and bronchiectasis. Small peripheral airways are obstructed in patients with obliterative bronchiolitis and fibrosis, and inflammation can be demonstrated in the small airways of young smokers. Localized obstruction can be due to tumours, adenopathy, inhaled material, tracheal damage and pressure from an enlarged thyroid gland. FIG. 13.3 Oxygen dissociation curve In normal subjects a large fall in arterial oxygen tension causes a small reduction in oxyhaemoglobin saturation (and oxygen carriage). In a hypoxic patient (for example, with a PaO2 of 6 kPa/45 mmHg) a small fall in oxygen tension causes marked further desaturation.

Determinants of bronchial calibre The total cross-sectional area of the tracheobronchial tree increases progressively as each generation of airways branches because the sum of diameters of the daughter branches exceeds that of the parent branch. Therefore,

607

resistance to airflow, which depends on cross-sectional area, is greatest in the large or central airways and least in the small or peripheral airways. Airway calibre is influenced by the tone of bronchial smooth muscle, which in turn is regulated by a variety of neurohumoral mechanisms. Normally, there is a small amount of resting tone because parasympathetic-bronchoconstrictor impulses predominate over sympathetic-bronchodilator impulses. In addition to bronchoconstriction, airway calibre may be decreased by mucosal oedema, muscle hyperplasia, bronchial secretions and loss of elastic recoil. Bronchial calibre will also be reduced by external compression or intraluminal tumour. Tests of airflow obstruction There are simple tests to assess the severity of airflow obstruction that can be performed at the bedside, in the doctor's surgery or in the laboratory. A simple handheld meter is used to measure the maximum or peak expiratory flow rate (PEFR). PEFR measurements are widely used to assess the severity of asthma (see Fig. 13.33) A portable spirometer can be used to measure the forced expiratory volume in 1 second (FEVi - Fig. 13.4A) and the vital capacity (VC). Normally the lung can be rapidly deflated during a maximal expiratory effort and the FEV1 is >75% of the maximum volume that can be expelled (VC), but with airways obstruction expiratory flow rates are reduced and the FEV1 becomes a smaller percentage of VC. Both tests depend on patient cooperation and effort, and predominantly reflect the function of relatively large airways. Additional useful information on the calibre of smaller airways can be obtained by recording flow against volume during a forced expiratory manoeuvre, producing a maximal expiratory flow-volume curve (Fig. 13.4B). Maximum expiratory flows (PEFR) are achieved from close to full inspiration (total lung capacity) and flow becomes progressively less as the lung volume falls. This is partly due to the narrowing of normal airways that occurs as the lungs get smaller. In patients with widespread airways obstruction (e.g. chronic bronchitis) the narrowing of bronchi means that at low lung volumes there is a marked increase in airways resistance and a fall in maximum flows. The flow-volume curve is therefore convex. Diffuse airways obstruction causes an increased volume of gas to be trapped in the lung at the end of expiration. Hyperinflation is therefore an important feature of airways obstruction, and can be demonstrated in the laboratory as an increase in residual volume (RV) and an increase in the ratio of RV to total lung capacity (Fig. 13.5). In all patients with airways obstruction, reversibility in response to an inhaled (32 agonist should be assessed. Rapid improvement after inhalation of bronchodilators is characteristic of asthma (see Fig. 13.34) and may occur to a lesser degree in some patients with chronic bronchitis, emphysema and bronchiectasis.

608

Upper airway obstruction This may be caused, for example, by tumours of the larynx

A

B FIG. 13.4

A Forced expiratory spirogram Starting from a full inspiration, the volume expired during a forced expiration (forced vital capacity, FVC) is reduced in both obstructive and restrictive pulmonary disease. However, the volume expired in 1 second (FEV1) is disproportionately reduced in patients with airways obstruction, but not in those with restriction. B Expiratory flow-volume curves In patients with airways obstruction expiratory flow is strikingly reduced as residual volume is approached. In restrictive disorders the shape of the flow-volume curve (although not the magnitude) is normal. (TLC = total lung capacity; RV = residual volume; PEFR = peak expiratory flow rate.)

or trachea, by post-tracheostomy strictures or by inhaled foreign bodies. Clinically, it can be difficult to distinguish upper airway obstruction from other causes of airflow limitation (although severe narrowing of the upper airway causes stridor). The flow-volume loop is diagnostically helpful (Fig. 13.6). In patients with asthma or chronic bronchitis expiratory flow is more reduced than inspiratory flow; when upper airway obstruction is present, inspiratory flow may also be equally reduced.

Restrictive lung disease Limitation of lung expansion can result from diseases of

IB

FIG. 13.5 The effect of obstructive and restrictive disorders on lung volumes Both obstruction and restriction reduce vital capacity (VC). With obstruction, residual volume (RV) and functional residual capacity (FRC) increase due to airway narrowing and closure during expiration, and the patient becomes 'hyperinflated'. With restrictive disorders total lung capacity (TLC) and FRC are reduced due to reduced compliance, or reduced respiratory muscle strength. (TV = tidal volume; IC = inspiratory capacity; ERV = expiratory reserve volume.)

FIG. 13.6 Flow-volume curves with airways obstruction With obstruction of the upper airway (e.g. tracheal tumour) maximal flow rates are reduced, resulting in the characteristic 'plateau' shape of the flow-volume curve. When breathing out from full inspiration (total lung capacity: TLC), peak expiratory flow rate (PEFR) is reduced. Typically PEFR is reduced to a greater extent than the forced expiratory volume during the first second (FEV1). (RV = residual volume.)

the bony ribcage, pleura or lung parenchyma, and of the respiratory muscles. A restrictive defect is therefore a feature of disorders as diverse as kyphoscoliosis, the pneumoconioses and polyneuritis. Pleural thickening or effusion also limits lung expansion. The most common causes of restrictive lung disease, however, are diffuse infiltrative processes within the lung, such as pulmonary fibrosis or extensive sarcoidosis, which reduce pulmonary compliance. Regardless of the underlying cause, pulmonary function tests show small lung volumes (Figs 13.4 and 13.5) and, often, a reduced transfer factor. Airways resistance is not increased and the ratio of FEV1 to vital capacity (VC) is not reduced.

increased blood volume and haemoglobin content of the lungs, as in left-to-right intracardiac shunts, polycythaemia, early left heart failure and pulmonary haemorrhage. The TLCO is raised in asthma, perhaps reflecting the widespread inflammation of the small airways. Transfer factor may be expressed as the carbon monoxide transfer coefficient (KCO), which is the TLCO per litre of alveolar volume. When lung volume is reduced (e.g. pneumonectomy) but the remaining lung is normal, TLCO is reduced but KCO is normal. When lung volume is small because of an inability to expand the lung (e.g. weakness, obesity) TLCO is low but KCO is normal or increased.

Abnormalities of diffusion

Abnormalities of pulmonary blood vessels

Abnormality of any of the parameters that determine diffusion will reduce TLCO (see p. 606). Tests of diffusing capacity are affected by the concentration of haemoglobin; thus TLCO is low in patients with anaemia and high in those with polycythaemia. When corrected for changes in haemoglobin concentration, decreased TLCO values are found most commonly in patients with destruction or obliteration of pulmonary capillaries and/or maldistribution of alveolar ventilation (emphysema, pulmonary vascular disease, interstitial infiltrative diseases). Increased TLCO values are not common but are found in patients with

The function of the pulmonary vascular system is to expose mixed venous blood to adequately ventilated alveoli (Fig. 13.2). Local vasoconstriction in response to hypoxia acts to divert blood flow from poorly ventilated to betterventilated regions. Conversely, if blood flow to part of the lung is compromised, ventilation of that region is reduced. Both these compensatory mechanisms tend to preserve the normal balance between blood flow and ventilation that is essential for efficient gas exchange. The most common disorder of the pulmonary circulation

609

is pulmonary embolism. Chronic obliterative disease of pulmonary arteries may cause severe dyspnoea on exertion but is difficult to diagnose at an early stage because physical examination, chest X-rays and spirometry show no abnormalities. In such disorders, gas transfer is reduced and hypoxaemia intensifies, or may only be apparent, on exercise.

Arterial blood gases and pH Arterial blood gases and pH are also discussed in Chapter 21. The prime function of the lung is to add O2 and remove CO2 from blood. Thus the ultimate test of the adequacy of ventilation, diffusion, pulmonary capillary blood flow and control of ventilation is measurement of Po2, Pco2 and pH of arterial, or arterialized earlobe, blood. 1 'Arterial Po2 Arterial oxygen tension (Pao2) is normally between 11.3 and 13.3 kPa (85 and lOOmmHg) in healthy adults breathing air, at rest and at sea level. The main causes of hypoxia are ventilation-perfusion (V/Q) abnormalities (e.g. pneumonia), hypoventilation (e.g. narcotics or sedatives), impaired diffusion (e.g. interstitial lung disease) and right-to-left shunts (e.g. pulmonary arteriovenous malformation). V/Q abnormalities are the most common cause; hypoventilation is characterized by hypercapnia, and impaired diffusion is rarely an important mechanism except on exercise. Breathing at altitude causes hypoxia (see Ch. 2, p. 49). Reduced delivery of oxygen to tissues, and therefore cellular hypoxia, occurs in severe anaemia, poor tissue perfusion, and (rarely) with respiratory chain enzyme abnormalities (e.g. mitochondrial myopathies, cyanide poisoning). Patients may have significant arterial hypoxaemia without symptoms. Central cyanosis only occurs when the saturation of arterial blood is less than 85%. Thus the clinical signs and symptoms of mild to moderate, but nevertheless important, arterial hypoxia are unreliable, and the diagnosis must be made by analysis of arterial blood gases. With chronic hypoxia patients develop secondary polycythaemia and pulmonary hypertension. Arterial Pco2 The arterial CO2 tension (Paco2) at rest is 4.6-6.0 kPa (35-45 mmHg). Regulation of arterial Pco2 is determined according to the equation: Arterial Pco2 =Kx

CO2 production Alveolar ventilation

Only two variables, CO2 production and alveolar ventilation, are involved. Deviation from normal values of arter-

1

610

MCQ 13.3

ial Pco2 can be viewed as a failure of alveolar ventilation to increase or decrease in accordance with metabolic changes. Hyperventilation, an increase in ventilation in excess of metabolic need so that arterial Pco2 decreases, occurs in patients with a variety of common lung diseases (asthma, pneumonia, pulmonary embolism), presumably from reflexes arising from the chest wall or within the diseased lung. Hyperventilation also results from stimulation of the central nervous system by drugs (aspirin), irritative lesions (cerebral tumours, infections) or anxiety. When Paco2 falls below about 20 mmHg there may be a sudden onset of dramatic symptoms such as paraesthesiae, blackouts and chest pain (owing to vasoconstriction and increased nervous irritability). Some patients hyperventilate and have symptoms due to hypocapnia, without any demonstrable cause. Symptoms are often misattributed by both the patient and the attending doctors to serious organic disease. Correct early recognition and reassurance are helpful. Inappropriate breathlessness may be a separate syndrome which is often, but not invariably, associated with excessive breathing and hypocapnia during exercise. The causes are obscure, but about one-third of such patients are depressed. There are few management strategies and response to treatment is poor. It should be stressed that, following the documentation of hyperventilation (i.e. hypocapnia), recognized disorders that cause excessive ventilation should be excluded before attributing the condition to a psychological cause. Hypoventilation can result from disorders that involve the airways (chronic bronchitis), lung parenchyma (advanced diffuse interstitial fibrosis), chest wall (kyphoscoliosis), respiratory muscles (myasthenia gravis) or central nervous system (sedative drugs). Commonly, in chronic diseases that can eventually lead to hypoventilation, hypercapnia first occurs during sleep. Primary causes of sleep-disordered breathing (e.g. obstructive sleep apnoea) can lead to profound hypoxia by night and eventually persistent hypercapnia by day. Like arterial hypoxia, mild to moderate changes in arterial Pco2 produce no or few signs and symptoms. Thus, analysis of arterial blood gases is required to determine the presence and severity of disturbances in Pco2.

Arterial pH The arterial pH (Fig. 13.7) at rest is 7.38. It can also be expressed as H+ content, and the normal value is 3545 nmol/L. The arterial pH value represents the net effects of both respiratory and non-respiratory factors that regulate Pco2 and HCO3 (and other buffers). When ventilation increases acutely, Pco2 decreases and pH increases; conversely, when ventilation decreases acutely, Pco2 increases and pH decreases. If the ventilatory changes persist, pH gradually returns towards normal values as the kidney makes appropriate adjustments in blood concentration of HCO3-. Thus the pH value depends on both the magnitude of the change in Pco2 and whether or not compensation has had time to occur.

TABLE 13.1 Indications for lung function testing

ED

Diagnostic To evaluate symptoms, signs and abnormal laboratory tests Screening 'at-risk' individuals (smokers, occupational exposure) Assessing preoperative risk Assessing prognosis Monitoring To evaluate the effects of treatment (e.g. bronchodilators, corticosteroids) To assess progress in chronic disease (COPD, pulmonary fibrosis, drug-induced lung disease) Measurement of disability Assessment as part of rehabilitation programmes Medicoiegal issues, insurance claims Epidemiological surveys (Adapted from Crapo, R. N Engl J Med, 1994, 331: 25-30).

FIG. 13.7 Acid-base diagram The relationship of arterial pH [H+], Pco2 and bicarbonate with disturbances of acid-base balance. For example, acute ventilatory failure leads to a raised arterial Pco2 and a reduction in pH; compensatory renal mechanisms then increase bicarbonate and restore pH to normal. (Modified from Flenley D C 1971 Another non-logarithmic acid-base diagram? Lancet 1961.)

Lung function testing (Table 13.1) Patterns of lung function abnormality are easily identified (Table 13.2). Airflow limitation without reversibility and with a normal TLCO is compatible with chronic obstructive bronchitis; if emphysema is present, however, TLCO and KCO will be reduced. A more than 15% improvement in FEY1 or peak flow after inhaled bronchodilator is suggestive of asthma. Restrictive lung disease due to pulmonary fibrosis will show low lung volumes with a normal or high FEV1/VC ratio and a reduced TLCO, whereas if the restriction is due to chest wall disease the TLCO may be normal or slightly reduced and the KCO raised.

lessness. Patients with lung disease stop exercising before achieving their maximum predicted heart rate and have levels of ventilation that are disproportionately high for a given oxygen uptake. Assessment of respiratory muscle strength is useful in patients with neuromuscular diseases and when evaluating selected patients with unexplained breathlessness (see p. 615). Vital capacity is a particularly useful simple test. Maximum inspiratory and expiratory mouth pressures can be measured using simple hand-held pressure meters. Specialist laboratories can undertake more complex techniques (e.g. phrenic nerve stimulation and diaphragm strength measurements). Overnight pulse oximetry and, in selected cases, polysomnography, is required to properly investigate sleep-disordered breathing (see pp. 668-670).

FURTHER READING ON STRUCTURE AND FUNCTION OF THE RESPIRATORY SYSTEM Hughes J M B, Pride N B 1999 Lung function tests. Philadelphia: WB Saunders.

Exercise testing

Simple tests can be of great clinical value in patients limited by lung disease. The 6- or 12-minute walking distance, in which the patient walks as far as he can during the time allowed, gives an indication of overall exercise capability. In some patients it may not be certain that lung function is impaired, and in such instances the monitoring of oxygen saturation, by finger probe, during treadmill or cycle ergometer exercise, is a sensitive technique for detecting minor abnormalities of oxygenation. Formal laboratory exercise testing - measuring oxygen uptake, CO2 production, heart rate and ventilation during progressive treadmill or cycle ergometer exercise - can be helpful, particularly in patients with unexplained breath-

SYMPTOMS AND SIGNS OF RESPIRATORY DISEASE The most important symptoms of respiratory disease are: Cough Sputum Haemoptysis Breathlessness Wheeze Chest pain.

611

TABLE 13.2 Patterns of abnormality in lung function tests Feature

Asthma

Chronic bronchitis

Emphysema

Pulmonary fibrosis

Chest wall abnormalities*

Airflow obstruction Reversibility Hyperinflation Reduction of lung volumes TLCO KCO Lung compliance

Yes Yes Yes No Normal or increased Normal or increased Normal

Yes Slight Yes No Normal Normal Normal

Yes Slight Yes No Reduced Reduced Increased

No No No Yes Reduced Reduced Reduced

No No No Yes Normal or reduced+ Normal or increased Normal or reduced

* e.g. kyphoscoliosis, muscle weakness. + (See p. 609)

Cough

612

Cough is initiated when irritant receptors in the mucous membrane of the respiratory tract are stimulated. Cough is by far the most common respiratory symptom, and is characteristic in heavy smokers. Frequently, cough is triggered by the presence of sputum in the respiratory tract, and is useful in helping to clear infection from the bronchial tree. A wide variety of inhaled irritants in addition to cigarette smoke (e.g. noxious gases or cold air) may stimulate coughing, and this is more likely if the airways are already irritable because of inflammation as a consequence of infection. Similarly, the irritant receptors in the bronchial tree may be stimulated by tumours, inhaled foreign bodies, allergens and the asthmatic response, pulmonary oedema and external compression by lymph nodes. In non-smokers the most frequent causes of chronic cough are asthma, sinus disease and oesophageal reflux. With neurological disease laryngeal function may be impaired or oesophageal motility abnormal (e.g. achalasia), and cough may be due to repeated aspiration. Drugs can cause cough, by damaging the lung (p. 696) or causing asthma (p. 651). A characteristic persistent dry cough can occur with ACE inhibitors. Cough after drinking can also indicate an oesophagobronchial fistula. In some patients cough is worse at night, particularly in asthma or pulmonary oedema. Prolonged coughing reduces venous return, causes a transient fall in cardiac output and cerebral oxygenation, and leads to cough syncope. Damage to the recurrent laryngeal nerve, commonly at the left hilum due to bronchial carcinoma, leads to vocal cord paralysis and an inability to produce a normal explosive cough, which becomes 'bovine'. A dry cough, sometimes following an upper respiratory tract infection and often persisting for weeks or months, for which no cause can be found, is a common clinical problem. It can present a formidable management difficulty, especially as patients' expectations of 'cure' are high (Table 13.3, Figure 13.8).

TABLE 13.3 Investigation of chronic cough Chest X-ray/HRCT Pulmonary function testing Sinus X-ray/CT Skin prick testing Blood/sputum eosinophils Sputum culture/AFB Oesophageal function - Barium swallow, pH, manometry, Speech therapist evaluation Bronchoscopy (when occult endobronchial lesion suspected) Common causes of refractory cough Asthma Chronic bronchitis/smoking Post viral URTI Gastro-oesophageal reflux Chronic sinusitis/post nasal drip Bronchiectasis Endobronchial - tumours foreign body exposure to irritants Parenchymal - pulmonary fibrosis sarcoidosis Cardiac - left ventricular failure Drugs-ACE inhibitors Inhaled irritants Neuromuscular disorders Psychogenic

Sputum In healthy subjects the bronchial tree produces approximately 100 mL of mucus each day; this is carried upwards by ciliary action and is then unconsciously swallowed. This

TABLE 13.4 Important causes of haemoptysis Pulmonary infections Bronchiectasis (particularly upper lobe bronchiectasis from past pulmonary tuberculosis) Lung abscess Pneumonia Tuberculosis Aspergilloma Pulmonary infarction Tumours Carcinoma Adenoma (Haemoptysis is unusual with pulmonary metastases) Pulmonary oedema Particularly due to mitral stenosis Pulmonary haemorrhage Goodpasture's syndrome Idiopathic pulmonary haemosiderosis Systemic lupus erythematosus Pulmonary vasculitis

Trauma Needle biopsy Transbronchial biopsy Pulmonary contusion Vascular abnormalities Arteriovenous malformation Hereditary haemorrhagic telangiectasia Bleeding disorders Usually severe thrombocytopenia from any cause Miscellaneous Pulmonary endometriosis (catamenial haemoptysis)

13

Disorders of the upper airways Epistaxis Tumours

FIG. 13.8 Algorithm for investigation of chronic cough

'escalator' is a normal part of the mechanism for clearing debris and pathogens from the bronchial tree. In disease processes causing the production of excess mucus, irritant receptors are stimulated and sputum is coughed up. Sputum is not described reliably by patients and it is always best to inspect it. Sputum may be clear, white or mucoid, as in chronic bronchitis, or purulent, in which case pus is mixed with mucus and the sputum is yellow or green. Sputum may contain blood, which may be bright red (e.g. pulmonary infarction), a rusty colour (acute pneumonia) or pink (pulmonary oedema due to left heart failure). In asthma the sputum may contain mucus plugs. Microscopically, sputum may contain bacteria, pus cells, eosinophils (as in asthma and pulmonary eosinophilia) or malignant cells. It is helpful to know the volume of sputum produced each day, and this can be particularly large: greater than 20 mL in bronchiectasis, cystic fibrosis, and lung abscess when there is a bronchopulmonary fistula. Clinical progress can be monitored by documentation of sputum volume. Occasionally patients with alveolar cell carcinoma produce very large volumes of clear watery sputum

(bronchorrhoea). Anaerobic infection results in foulsmelling sputum.

Haemoptysis Patients may have difficulty in being sure that blood has been coughed rather than vomited. Coughed blood may be from a lesion of the nose, nasopharynx and vocal cords. When infection causes haemoptysis, blood will be mixed with purulent sputum, whereas in non-infective causes (e.g. pulmonary infarction) there is usually frank blood and no sputum. In patients with chronic bronchitis it is common for the sputum to contain occasional specks of blood, but more substantial haemoptysis is unusual (Table 13.4). A cause for haemoptysis should be carefully sought in all cases, but a definite diagnosis is only achieved in 60-70%. Occasionally haemoptysis can be severe and lifethreatening. This occurs most frequently in cavitating tuberculosis and bronchial carcinoma. In those patients in whom no definite cause for haemoptysis is found, careful follow-up is necessary, with repeated chest X-rays. The importance of haemoptysis as the presenting feature of lung cancer must be emphasized. In particular, frank haemoptysis in a smoker, even in the presence of a normal plain chest X-ray, may be the earliest warning of a

613

CASE STUDY: 13.1 PERSISTENT SEVERE HAEMOPTYSIS A 45-year-old Afro-Caribbean woman was admitted to hospital with persistent severe haemoptysis. She had originally attended the hospital 10 years previously with sharp central chest pain, bilateral hilar adenopathy, and enlargement of the mediastinal lymph nodes. At that time a lymph node biopsy, via a mediastinoscopy, demonstrated non-caseating granulomata, consistent with a diagnosis of sarcoidosis. Subsequently the patient developed bilateral pulmonary infiltrates and increasing breathlessness. She was treated with oral prednisolone, with substantial symptomatic improvement. Over the next few years, on varying doses of prednisolone, the adenopathy resolved but the interstitial lung disease persisted. The patient developed a moderately severe restrictive defect on lung function testing. She found the adverse effects of steroid therapy, particularly the weight gain and hirsutism, troublesome, and for a period of several years did not attend regularly for follow-up and took her steroid treatment intermittently. Eighteen months prior to her current admission she first experienced haemoptysis. At that time bronchoscopy showed blood throughout the bronchial tree, but probably originating from the left upper lobe bronchus. Chest X-rays and a CT scan showed extensive bilateral upper lobe fibrosis with spaces within the lung and cavity formation (Case Fig. 13.1.1).

Questions 1. Should the diagnosis of sarcoidosis be reconsidered? 2. What is the cause of the haemoptysis? 3. What would be the future management 01 this patient?

614

CASE FIG. 13.1.1 [70 Chest X-ray showing bilateral upper zone volume loss, fibrosis and cavitation (best seen on the right). B CT scan There are large spaces in the upper zones of both lungs.

The original presentation is typical of sarcoidosis. It is not uncommon for patients with extensive intrathoracic lymphadenopathy to experience chest pain, often with a sharp quality. The histology of the lymph node makes the diagnosis virtually certain. Tuberculosis was possible and culture of the biopsy material was essential. In this patient the culture did not yield mycobacteria and a certain diagnosis of sarcoidosis was therefore made. Although sarcoidosis can effect all racial and ethnic groups it can be particularly persistent and severe in Afro-Caribbeans. A reasonable question is whether the development

of upper zone spaces and cavitation should lead one to suspect an additional diagnosis, particularly tuberculosis. Occasionally patients with sarcoidosis do develop tuberculosis (and vice versa); however, extensive sarcoidosis also leads to cavity formation, usually in the mid and upper zones. The poor control of this patient's disease has led to extensive fibrosis and lung destruction, which can also cause a degree of bronchiectasis. Thus, although tuberculosis should be considered as a cause for the haemoptysis, so should bacterial infection within damaged and bronchiectatic lung. Crucial investigations would include sputum examination and culture to exclude tuberculosis and a tuberculin skin test, usually positive in TB and characteristically negative in sarcoidosis. This patient had bronchial lavage, which showed no evidence of TB, and a Heaf test which showed no reaction. She was treated with antibiotics and steroids and the bleeding settled. Over the next 18 months there were several episodes of substantial bleeding. Bronchial arteriograms showed abnormal bronchial artery circulations in both upper lobes, and on two occasions she was treated by bronchial artery embolization. This treatment produced only a temporary reduction in haemoptysis. In the months leading up to the current admission the patient had coughed blood during most weeks, often up to a cupful 'daily, and was highly symptomatic because of a persistent disabling cough for much of the day and night. Question 4. What other complication of her cavitatpry disease could explain her haemoptysis?

13

CASE STUDY 13.1 CONTINUED A repeat CT scan demonstrated soft tissue shadowing within the cavitatory systems in the upper lobes, particularly on the left, strongly suggesting aspergilloma formation (Case Fig 13.1.2). The patient underwent surgery and the complex mycetoma cavity in the left upper lobe was cleaned out and the space packed with intercostal muscle. She did well postoperatively, with resolution of both her cough and haemoptysis. If bleeding returns the management would be to operate to remove the mycetoma cavities within the right lung. Question 5. How should severe haemoptysis be managed? The main danger of severe haemoptysis is hypoxia, as the bronchial tree becomes flooded with

blood. Large-volume bleeding, combined with coughing, rapidly fills the large and small airways. It is helpful to identify which lung is bleeding and for the patient to lie on their side with the bleeding lung dependent, thereby limiting soiling of the good lung. If the PaCO2 is not raised mild sedation is appropriate. The patient should be discussed with anaesthetic colleagues because urgent intubation may be necessary. Intubation should be with a doublelumen tube to enable the bleeding lung to be isolated. Definitive treatment of haemoptysis will depend on the cause. For example, bleeding endobronchial tumour can be treated by laser therapy; bleeding from bronchiectatic lung by bronchial artery embolization; arteriovenous malformation by coil insertion; and, if necessary, refractory bleeding can be treated by lung resection.

CASE FIG. 13.1.2 IT] Chest X-ray. Note that compared with Figure 1 the left upper zone is more opaque, as a result of mycetoma formation. B CT scan. Note the large round opacity within a cavity in the left lung (arrowed). This is a large mycetoma.

tumour in a large airway which has not yet occluded the bronchus.

Breathlessness Breathlessness is an unpleasant awareness of the effort of breathing. This sense of effort usually reflects an increased ventilatory load (e.g. bronchoconstriction), a reduced breathing capacity (e.g. respiratory muscle weakness due to motor neuron disease) or both (e.g. severe COPD, in which respiratory muscle function is impaired because of hyperinflation and load is increased). The history should document severity: How far can the patient walk on the flat? How many steps can the patient climb without stopping? Breathlessness may have characteristic

features: in left heart failure breathlessness is worse when lying flat; in asthma it is frequently episodic, worse at night. With psychogenic breathlessness the patient complains of difficulty in taking air in, difficulty in fully filling the lungs with air and an inability to take a deep breath. These patients are often breathless at rest but not on exercise. Such 'behavioural breathlessness' causes a low arterial CO2, and symptoms of the hyperventilation syndrome (e.g. faintness, chest pains) can be reproduced by asking patients to breathe deeply for 20-30 seconds. The period over which breathlessness has developed is of great importance (Table 13.5). The patient with chronic bronchitis who continues to smoke has gradually progressive breathlessness; an acceleration in the pace of the symptom may reflect a bronchogenic

615

CASE STUDY 13.2 CHRONIC BREATHLESSNESS ON EXERTION IN A YOUNG PROFESSIONAL FOOTBALLER The patient was a 29-year-old man from Kenya who had come to the UK 4 years previously. He was a professional football player. For many years he had noted breathlessness on exertion, gradually becoming more severe, and despite his footballing talents he was now finding it difficult to complete a full game. Three months previously he had been admitted to hospital with cough, purulent sputum and fever. He was diagnosed as having a chest infection, treated with antibiotics, and made a full recovery, albeit slowly, after a period of 6-8 weeks. There was no relevant past medical history, and particularly no history of respiratory or cardiac illnesses. He was a non-smoker. On examination the patient was comfortable at rest; BP 125/80, heart rate 64, JVP not raised, apex beat not displaced, no murmurs or added sounds heard. There was no cyanosis, no clubbing, the trachea was central and the respiratory rate 14. Expansion of the lower chest on the left was reduced. The breath sounds were slightly reduced over the left lower zone anteriorly, laterally and posteriorly. Vocal resonance and tactile vocal fremitus were reduced at the left base. Occasional inspiratory crackles were audible at the left base.

Question 1. What is the likely explanation for the patient's physical signs and breathlessness?

Discussion There are no signs of cardiac or pericardial disease. The signs indicate chest pathology. The reduced expansion of the lower left hemithorax, combined with reduced breath sounds, tactile fremitus and vocal resonance, would be consistent

616

with pleural effusion, collapse or pleural thickening. The normally positioned apex beat and the easily audible breath sounds make a substantial pleural effusion unlikely. The central position of both the upper and lower mediastinum is against collapse. The pathology best explained by the physical signs is extensive pleural thickening. The chest X-ray is illustrated in Case Figure 13.2.1. What abnormalities are shown? Note the reduced transradiancy of

linear, mainly vertical opacities, typical of pleural thickening and calcification. There is dense calcification of the left diaphragmatic pleura. A CT scan of the thorax is shown in Case Figure 13.2.2 and demonstrates a rim of thickened

CASE FIG. 13.2.1 Chest X-ray. (See text for description).

the left lower zone compared to the right. The vascular pattern is nevertheless detectable, indicating that the abnormal shadowing is outside the lung (i.e. pleural). Pleural thickening is apparent, running down the inner border of the left chest wall and over the left hemidiaphragm, making the diaphragm indistinct. The mediastinum is central. Note the reduced space between the left lower ribs compared to the right, indicating that the changes are chronic. In the left lower zone, adjacent to the heart and behind it, are irregular

CASE FIG. 13.2.2 CT scan of thorax. A Scan through lower thorax, showing pleural thickening, calcification and reduction in volume on the left side. B Scan at the level of the diaphragm; on the left there is gross thickening, and calcification of the pleura. pleura with calcification, surrounding the left lung. Note the reduction in the area of the left hemithorax compared to the right. The cause of

this patient's chronic pleural thickening and calcification is almost certainly tuberculosis, which would have caused a tuberculous pleurisy when he was a child. Adult patients often have no knowledge of having been unwell when young. A traumatic haemothorax can cause a similar picture, but there was no history or signs of severe thoracic trauma. The lung function test results for this patient are shown in the table below. Measured

FEV! (L) VC (L) FEVj/VC % PEFR(L/min) TCO (mmol/ kPa/min) KCO

2.2 2.3 95% 455 7.8 2.3

Predicted

4.5 5.4 606 12.3 1.6

Question 2. What is the interpretation of these results?

The marked reduction of vital capacity, combined with the high FEV\/VC ratio, indicates a severe restrictive ventilatory defect. The small reduction in PEFR reflects the reduced TLC, and there is no airways obstruction. The difficulty in expanding the lung and recruiting alveoli for gas exchange causes the low overall gas transfer (TCO), but the lungs themselves are normal and relatively well perfused, and when TCO is corrected for the small lung volumes to derive the transfer coefficient (KCO) this is supranormal. A reduced TCO and normal or increased KCO is typical

13

of extrapulmonary restriction (e.g. pleural disease, obesity, muscle weakness).

Question 3. Why is the patient breathless?

The causes of breathlessness are multiple and poorly understood. Hypoxia and hypercapnia can contribute, but changes in blood gases per se are often not of great importance. Breathlessness is an unpleasant awareness of the difficulty of breathing and is most commonly the consequence of disproportionate respiratory effort. Whenever the load on the respiratory system is increased (e.g. lung fibrosis, airways obstruction) and/or ventilatory capacity is reduced (e.g. respiratory muscle weakness), neural respiratory drive is increased and becomes a larger proportion of maximum available respiratory drive. Thus, breathing is perceived as being a great effort. Pleural thickening encasing the lung, as in the present case, dramatically reduces chest wall compliance, imposes a high load on the respiratory system, and makes ventilation difficult despite great efforts. Case Figure 13.2.3 shows ventilation and perfusion lung scans of the patient. Note the marked reduction in ventilation of the left lung. Perfusion of the left lung is also reduced, but less so than ventilation, and such ventilation/perfusion mismatch contributes to hypoxia, especially on exercise. Treatment for chronic pleural thickening is limited. In selected cases it is possible for a thoracic

carcinoma, narrowing a bronchus or causing a pleural effusion. Most causes of breathlessness are relatively easily diagnosed by history, physical examination, chest X-ray, blood

ANT

VENT

CASE FIG. 13.2.3 Lung ventilation and perfusion scan. Both ventilation and perfusion are reduced in the left lung. The left lung is receiving 23% of total ventilation and 34% of perfusion.

surgeon to dissect the thickened pleura off the lung. This is a difficult operation, often only partially successful, and not appropriate in most cases. In the present case the patient was advised that surgery was not appropriate and he resolved to seek educational training for a different profession. The case emphasizes how important it is that acute pleural disease (particularly empyema) is correctly managed to avoid whenever possible the development of chronic extensive pleural thickening.

gas analysis and lung function tests. In difficult cases an exercise test may be helpful. Although orthopnoea and paroxysmal nocturnal dyspnoea suggest left heart failure, many patients with

617

TABLE 13.5 Time course for the development of breathlessness Immediate Pneumothorax Pulmonary oedema due to cardiac arrhythmias Pulmonary embolism Inhalation of foreign body

Weeks Pleural effusion Anaemia Muscle weakness Months Tumours Pulmonary fibrosis Thyrotoxicosis Muscle weakness

Hours Asthma Left heart failure Pneumonia Laryngeal oedema Days Pneumonia Acute respiratory distress syndrome Left heart failure

Years Muscle weakness Chronic airways obstruction Pulmonary fibrosis Primary pulmonary hypertension Chest wall disorders

asthma are worse at night and many patients with severe airways obstruction are more comfortable sleeping with several pillows. Diaphragm weakness can cause profound orthopnoea.

Wheeze Wheeze is generated by increased turbulence of airflow through a narrowed airway. Patients with airways obstruction frequently complain of wheeze, which in asthma, chronic bronchitis and emphysema is predominantly expiratory. Patients may notice that a wheeze is inspiratory, indicating narrowing of the larynx, trachea or main bronchi, and occasionally they will remark that the wheezing sound is from only one side of the chest, as when a tumour narrows a major bronchus. Diffuse expiratory wheezes can occur in left heart failure.

Chest pain The many causes of chest pain can usually be distinguished by careful history taking, paying particular attention to the duration of the pain, its site and radiation, and the relationship to movement, breathing and exercise (see Ch. 12, p. 467). The quality of pain is also diagnostically important, especially if it is pleuritic in nature. Diseases within the

618

1

MCQ 13.4

3

Fig. 10.48

2

Figs 13.1-13.3

lung are usually painless, as there are no pain receptors in lung tissue. Pleural pain This usually indicates involvement of the parietal pleura by inflammation or malignancy, but is also a feature of pneumothorax. Inflammation of pleura can be caused by: • • • • • • •

Pneumonia Pulmonary infarction Lung abscess Tuberculosis Rheumatoid arthritis Systemic lupus erythematosus Uraemia.

Pleural pain is sharp, often well localized, worse on breathing, particularly with deep inspiration, and causes the patient to catch his breath. Pain from inflammation of the central diaphragm is referred to the shoulder; pain from inflammation of the lateral diaphragm is referred to the lower lateral chest wall and upper abdomen. Mediastinal pain Central chest pain may occur with primary carcinoma of the lung, particularly when there is involvement of mediastinal structures including lymph nodes. Adenopathy due to other disorders, such as sarcoidosis, can cause pain. The differential diagnosis will include angina, pericarditis, oesophagitis and bone pain. Severe sudden central chest pain can occur with pulmonary embolism (p. 682). Chest wall pain With rib fractures there is local pain and tenderness. Occasionally coughing leads to fracture, particularly of the middle ribs posterolaterally. Osteoporotic ribs and those involved by tumour fracture easily. Cough fractures in patients with chronic asthma treated with long-term oral steroids are relatively common, as are fractures in the elderly. Infiltration of the chest wall by tumour causes pain and is an important feature of pleural malignancy, particularly mesothelioma. Perichondritis causes pain and swelling of the costochondral junction, often of the second ribs bilaterally (Tietze's syndrome). Pain within the distribution of thoracic nerves occurs with herpes zoster, the pain frequently preceding the vesicular rash. Nerve root pain, often described as burning, also occurs with an intervertebral disc or malignant and inflammatory diseases of the spine. Local pain and tenderness of the muscles of the chest wall is caused by Coxsackie B viral infections (Bornholm's disease) which may sometimes involve the diaphragm. Upper chest wall and shoulder pain is a feature of neuralgic amyotrophy, and phrenic nerve involvement can lead to unilateral or bilateral diaphragm paralysis. Rib pain occurs in sickle cell crisis, and these patients are more likely to develop sickle lung. Localized muscle pain without an important underlying pathological process is common in breathless, coughing patients. 1

13

EXAMINATION OF THE RESPIRATORY SYSTEM In thoracic medicine the history, physical examination, chest X-ray and pulmonary function tests all contribute to diagnosis and assessment in a complementary manner. The chest X-ray gives accurate information on thoracic anatomy, whereas the clinical examination is much more useful in assessing pathophysiology. Before the clinical examination, many relevant observations will have been made while talking to the patient concerning breathlessness, pain, cough and general appearance. In addition, the history will have served to focus attention on particular diagnostic probabilities and possibilities that subsequent examination will confirm or refute. If already available, the information from the chest X-ray will similarly direct the physician's examination of the patient.

Examination of the hands Examination of the hands may give valuable information: • Finger and toe clubbing (Fig. 13.9) is an important physical sign. The four criteria for documenting finger clubbing are: increased sponginess of the nailbed; loss of the usual acute angle between the nail and the nailbed; increased nail curvature; and increased bulk of the soft tissues over the terminal phalanges. The majority of patients with clubbing have pulmonary disease (Table 13.6). Clubbing is occasionally familial in otherwise normal subjects. It can develop very rapidly, within a few weeks (e.g. with an empyema), but its onset is usually gradual. The mechanism of clubbing is not understood. The bones of the fingers and toes are normal. If the underlying cause is successfully treated clubbing usually resolves. • Hypertrophic pulmonary osteoarthropathy (HPOA). Virtually all patients with this unusual condition have clubbing, but only a few patients with clubbing have HPOA. There is arthralgia and joint swelling affecting particularly the wrists and ankles. X-rays show subperiosteal new bone formation in the long bones of the lower limbs and forearm, the bone scan demonstrates increased activity, and often the serum alkaline phosphatase is raised. Ninety per cent of cases are associated with bronchogenic carcinoma, particularly peripheral squamous cell tumours. • Examination of the nails. This may show that some or all of them are thickened and yellow or greenish in appearance. This occurs in the yellow nail syndrome, 3 in which clubbing is associated with lymphoedema, exudative pleural effusions, bronchiectasis and sinusitis. • Carbon dioxide retention. When there is carbon dioxide retention, peripheral vascular dilatation greatly

FIG. 13.9 Finger clubbing

TABLE 13.6 Important causes of clubbing Bronchial carcinoma (particularly squamous cell) Mesothelioma Benign intrathoracic tumours

Cyanotic congenital heart disease Bacterial endocarditis Atrial myxoma

Pulmonary arteriovenous fistula

Cirrhosis

Lung abscess Empyema Bronchiectasis Fibrosing alveolitis Asbestosis Advanced pulmonary fibrosis of any cause

Ulcerative colitis Crohn's disease

Familial

enhances blood flow and the hands are strikingly warm. With severe and acute carbon dioxide retention there may also be an irregular flapping tremor of the outstretched hands. The fingers may be blue (peripheral cyanosis), and if the limb is warm this is likely to reflect central cyanosis rather than poor peripheral perfusion.

Cyanosis In patients with a normal haemoglobin concentration, oxygen saturation must fall to 85%, corresponding to an arterial oxygen tension of 7-8 kPa (approximately 5560mmHg), before cyanosis can be clinically detected. As a consequence of the sigmoid shape of the oxygen dissociation curve (Fig. 13.3), further small falls in oxygen tension then produce large, dangerous falls in saturation. In anaemia, severe hypoxia is required to produce the quantity of reduced haemoglobin (about 1.5g/dL) necessary to cause cyanosis, whereas patients with polycythaemia become cyanosed at higher arterial oxygen tensions. The bluish colour of central cyanosis is best appreciated by examination of the tongue in adequate daylight.

619

Cyanosis is a difficult physical sign, and estimates of severity are so unreliable that measurement of the arterial oxygen tension is the next logical step. Cyanosis cannot be reliably detected in many black and Asian patients, in whom severe hypoxaemia may go undetected. In most cases central cyanosis is due to lung disease. Right-to-left shunts within the heart or within the lungs (arteriovenous malformation) cause central cyanosis, as do pulmonary shunts in advanced cirrhosis. Rarely patients appear cyanosed due to methaemoglobinaemia or sulphaemoglobinaemia, most commonly caused by drugs.

Pulsus paradoxus The curious term 'pulsus paradoxus' is used to describe a greater than normal fall in blood pressure during inspiration. In normal subjects the systolic blood pressure falls by a few mmHg during inspiration. This fall is greater (not paradoxical) when venous return to the right heart is impaired, e.g. in patients with hypovolaemia, cardiac tamponade or massive pulmonary embolism. In respiratory disorders pulsus paradoxus is seen when there are large pressure swings within the thorax during the respiratory cycle. This is particularly likely to occur in severe acute asthma, when the extent of paradox is related to the severity of airways obstruction, and systolic blood pressure may fall by 40 mmHg on inspiration. The degree of paradox is best documented using a sphygmomanometer: as the pressure of the cuff is reduced, the systolic sound is initially only audible on expiration, but with a further reduction in cuff pressure it becomes audible throughout inspiration too. The pressure difference between the measurement taken when the systolic sound is first detected and the measurement taken when it is present throughout the breathing cycle is termed the degree of paradox (recorded in mmHg). In asthma, paradox is dependent on respiratory muscle contraction and is reduced when hyperinflation is so severe that the generation of large inspiratory pressures is no longer possible because of exhaustion, as well as with improvement in airways obstruction.

Jugular venous pressure The jugular venous pressure is raised in right heart failure, which is itself frequently due to pulmonary disorders such as chronic bronchitis. The venous pressure is also elevated if the resting pressure in the thorax is raised, as with a tension pneumothorax. When there is severe airways obstruction the pressure swings within the thorax are large; intrathoracic pressure is positive during expiration, owing

620

1 Fig. 13.4

2

Fig. 13.5

3

Fig. 13.6

4

5

Fig. 13.8

6

Fig. 13.9

Fig. 13.7

to active recruitment of the expiratory muscles and the positive recoil pressure of the respiratory system when expiration is halted by airway collapse at a lung volume well above normal FRC. This positive intrathoracic pressure elevates the jugular venous pressure, which then falls during inspiration. Thus, interpretation of the jugular venous pressure in the tachypnoeic patient, particularly with severe airways obstruction, is difficult. The jugular venous pressure is raised and not pulsatile in superior vena cava obstruction, most commonly due to malignant nodes or a tumour mass compressing the vein. 1 The most common cause is an extending bronchial carcinoma, but lymphoma, thymic tumours and metastases can all cause compression. Very rarely non-malignant lesions are responsible, e.g. mediastinal fibrosis, or aortic root aneurysm or dissection. Superior vena caval obstruction produces drowsiness, a sense of fullness in the head, swelling and cyanosis of the face, neck and arms, epistaxis, and sometimes papilloedema.

Peripheral oedema Peripheral oedema associated with severe pulmonary disease (cor pulmonale) is relatively common, and is most often seen with chronic bronchitis, in which there is hypercapnia as well as hypoxia. Patients with hypoxic normocapnic respiratory failure seldom have oedema. In some patients with oedema the jugular venous pressure is not elevated, the cardiac output is normal and fluid accumulation is related to factors other than right heart failure, perhaps mediated through the action of hypoxia and hypercapnia on renal blood flow and function.

Examination of the chest Examination of the chest (Fig. 13.10) may reveal that the chest wall shows an abnormality of shape (kyphoscoliosis, a barrel chest reflecting hyperinflation or pectus deformity) or an abnormality of symmetry (a reduction in the volume of one hemithorax reflecting underlying chronic fibrosis, or perhaps an increase in volume reflecting a pneumothorax). Much can be learned about respiratory function by observing the patient's breathing pattern. An increase in respiratory rate is a sensitive index of cardiorespiratory disorders and is not sustained above 14-20 breaths per minute in normal adults. Disorders that cause widespread functional impairment, such as asthma, pulmonary fibrosis and pulmonary oedema, invariably increase respiratory rate. Conversely, respiratory rate is reduced by central nervous system injury or central depressant drugs. Visible or palpable contraction of the accessory muscles is abnormal. Scalene and sternomastoid activity is particularly obvious in severely hyperinflated patients, whereas in patients who are not overinflated, accessory muscle activity is an indication of a greatly increased respiratory effort. In severely breathless patients abdominal muscle activity is vigorous.

Abnormality

Consolidation 2

Pleural effusion 3

Lobar collapse 4

Pneumothorax

5

Yes away from effusion

Yes towards collapse

No without tension Yes with tension

Pneural thickening 6

13

Chest radiograph

Mediastinal shift

No

Chest wall movements

Normal or

Breath sounds

(Bronchial)

Added sounds

Crackles

Percussion note duller =more resonant) Tactile fremitus Vocal resonance

No

Normal or

Occasional rub

No

Occasional click

No

Normal or

| most distinguishing sign FIG. 13.10 Signs of most important focal abnormalities

Chest wall movements Both sides of the thorax should be seen to expand equally during tidal and full inspiration. Hands placed flat on either side of the sternum can appreciate the predominantly forward movement of the sternum and upper ribs, whereas hands placed around the lower thorax can best detect the normal outward and upward movements of the mid and lower ribs. Overall movement of the ribcage is reduced by hyperinflation (e.g. emphysema), reduced pulmonary compliance (e.g. lung fibrosis), reduced chest wall compliance (e.g. ankylosing spondylitis) and weak inspiratory muscles (e.g. myasthenia gravis). Local chest wall movement is reduced in pleural effusion, pleural thickening, pulmonary collapse, and to a lesser degree in consolidation and pneumothorax. The pattern of breathing is altered by disease. For example, patients with stiff lungs, a poorly compliant ribcage or weak respiratory muscles breathe rapidly and shallowly. It must be stressed that clinical assessment of the depth of respiration is notoriously inaccurate. Patients with airflow obstruction have a prolonged expiratory phase and patients with left heart failure may have marked periodic respiration. With severe weakness or paralysis of the diaphragm the anterior abdominal wall moves paradoxically inwards during inspiration when supine (see Fig. 13.76).

Trachea and apex beat The trachea is an important anatomical landmark indicating the position of the upper mediastinum. It is best pal-

pated by one finger gently placed in the suprasternal notch. Loss of volume in one hemithorax, as with pulmonary flbrosis, pulls the trachea to the same side. Expansion of a hemithorax (as with tension pneumothorax) or local masses (e.g. an enlarged thyroid) pushes the trachea towards the opposite side. When examining the trachea, the distance between the suprasternal notch and the cricoid cartilage is noted and is normally 3-4 fingers, and a reduction in this distance is a reliable sign of hyperinflation. Assessment of tracheal shift is difficult in patients with hyperinflation because the displacement is within the thorax. A shift of the lower mediastinum displaces the cardiac apex. However, a displaced apex must be interpreted with caution because ventricular enlargement also moves the apex beat. In some patients the apex beat is not palpable, usually because of obesity or hyperinflation of the lungs. Less commonly the apex is difficult to palpate because of pericardial disease or poor left ventricular function, and the apex will be located on the right side with dextrocardia. Palpation at the left sternal edge may elicit the heave of right ventricular hypertrophy, common in patients with severe pulmonary disease causing pulmonary hypertension. In patients with hyperinflation heart sounds are much reduced at the apex, but are well heard over the lower sternum in the midline.

Percussion When percussing the chest, comparison should be made between corresponding areas on both sides in an attempt to detect differences in percussion note. The percussion note is resonant over aerated lung, and hyperresonant with emphysema, large bullae or pneumothorax. Percussion is

621

dull over solid organs such as the liver and heart (except with hyperinflation, when the aerated lung covers the heart). In the normal subject the upper level of liver dullness when supine is at the sixth rib in the midclavicular line. The percussion note becomes dull when there is fluid in the pleural space, when there is collapse of a lobe or lung, and when there is extensive consolidation. Less marked dullness occurs with pulmonary fibrosis, pleural thickening and large peripheral tumours. Basal dullness as a consequence of diaphragm elevation is easily confused with pleural fluid. When the palm or side of the examiner's hand is placed on the chest and the patient talks (e.g. says ninetynine) vibrations are easily felt (tactile fremitus); and if a stethoscope is used the voice sounds are well heard (vocal resonance). Tactile fremitus and vocal resonance are increased with pulmonary consolidation and decreased by pleural fluid, but are relatively crude physical signs.

Auscultation Listening to the chest is of the greatest clinical importance: a chest X-ray can subsequently provide detailed structural information, but auscultation is crucial for providing functional data. At the bedside the breathing of patients with airways obstruction is noisy, and this correlates well with the degree of obstruction documented by pulmonary function tests. This noisy breathing is particularly prominent in chronic bronchitis and asthma, but is less marked in emphysema because inspiratory resistance is not increased. On auscultation by contrast, breath sounds are generally reduced in patients with airways obstruction, particularly emphysema, and there may be local areas of greatly reduced breath sounds over bullae, a pneumothorax, pleural effusions, and when a lobe or lung is poorly ventilated, as when there is substantial bronchial obstruction due to tumour. Expiratory breath sounds are prolonged with airways obstruction, whereas expiration is rapid in patients with severe restrictive disease. In normal subjects expiration is followed by a pause before the next inspiration; this disappears as respiratory rate increases. Bronchial breathing is when the breath sounds become harsh and high-pitched, owing to enhanced transmission of sound through the abnormal lung. Whispered sounds are then easily heard as high-pitched 'whispering pectoriloquy'. Bronchial breathing is most commonly heard over areas of consolidated lung or over large peripheral cavities. In addition to alteration of the normal ('vesicular') breath sounds the stethoscope may detect additional sounds:

622

Crackles Crackles (crepitations or rales) are short explosive sounds thought to represent the equalization of intraluminal pressure as collapsed small airways open during inspiration. Crackles therefore occur in disease processes causing small airway closure, whether due to airway damage (chronic bronchitis) or to increased interstitial volume (pulmonary oedema, pulmonary fibrosis). With diffuse disease the effect of gravity makes crackles more prominent in dependent parts of the lungs. In immobile patients small airways at the lung bases close, giving rise to crackles that clear once deep breaths are taken; the same also occurs postoperatively. The crackles of chronic bronchitis and emphysema are characteristically early in inspiration and are transmitted to the patient's mouth, where they can be easily heard by the unaided ear. In bronchiectasis the crackles are maximum in midinspiration, and in pulmonary fibrosis or oedema they occur during mid and late inspiration; in both circumstances the crackles are not transmitted to the mouth. Wheezes Wheezes are musical sounds reflecting airway narrowing, such that the airway oscillates between being open and closed like the reed of an oboe. In asthma, chronic bronchitis and emphysema, multiple polyphonic wheezes are heard in expiration. A localized area of narrowing due to tumour or foreign body produces a single monophonic wheeze. Wheezes caused by intraluminal secretions often disappear following coughing. The causes of wheeze are listed in Table 13.7. Narrowing of the upper airway (larynx and trachea) causes a wheeze in inspiration as well as on expiration, which may be audible at the bedside and is then described as stridor. Stridor is an important clinical sign that indicates severe, potentially critical, narrowing of the upper airway. Pleural rub A pleural rub is always pathological and is the sound produced when the two layers of pleura move over one another in a jerking motion, generating a creaking sound similar to that produced by bending stiff leather. Rubs are sometimes palpable, and the rubbing sensation is often appreciated by the patient. Pericardial sounds are discussed in Chapter 12.

SUMMARY 1 Common causes of crackles

• Crackles • Wheezes

1

• Pleural rub • Clicks.

MCQ 13.5

• • • • • • •

Bronchiectasis Pneumonia Left heart failure Acute respiratory distress syndrome Pulmonary fibrosis Chronic bronchitis and emphysema Bronchiolitis

TABLE 13.7 Causes of wheeze

Viewing the chest X-ray 1

Generalized

Localized

A normal chest X-ray is illustrated in Figure 13.11. The following points should be considered.

Asthma Chronic bronchitis Emphysema Pulmonary oedema

Tumour Extrinsic compression Foreign body Bronchial secretions

SUMMARY 2 Clinical examination: points to note • • • • • • • • •

Hands Cyanosis Respiratory rate Peripheral oedema Trachea and apex beat Jugular venous pressure Chest wall structure and movements Percussion note Breath sounds

Clicks A clicking sound with each heartbeat is occasionally heard with a small left pneumothorax.

FURTHER READING ON EXAMINATION OF THE RESPIRATORY SYSTEM Braman S S (ed). Pulmonary signs and symptoms. Clin Chest Med 1987; 8:2.

IMAGING THE THORAX THE CHEST X-RAY In respiratory medicine the chest X-ray is of fundamental importance in demonstrating anatomy and is a direct extension of the physical examination. Chest X-rays should be of good quality. Underexposure of films is a much greater problem than overexposure because it cannot be compensated for by any viewing conditions. Digital imaging systems have improved the quality of films, especially in difficult clinical situations (e.g. supine patients in ICU). Digital films can be viewed in depth and online, and may greatly change respiratory clinical practice. All available films should be studied in sequence: past films are frequently the most important in understanding the patient's clinical problems. The chest X-ray should be taken with the patient properly centred and the medial end of the clavicles equidistant from the spinous processes. The film should be taken in full inspiration, in which case the dome of the diaphragm will normally be at the level of the sixth rib anteriorly and the tenth rib posteriorly in the midclavicular line.

13

Trachea The translucent tracheal air column is readily seen in the neck and superior mediastinum; it is placed centrally or slightly to the right. The tracheal air column may be narrowed by intraluminal disease (e.g. tumour), external compression (e.g. enlarged thyroid), or displaced to the right or left (e.g. upper lobe collapse or fibrosis). The paratracheal regions of the superior mediastinum are common sites of pathology (e.g. lymphadenopathy, apical pulmonary tumours). Diaphragm Downward displacement of the diaphragm indicates largevolume lungs (e.g. emphysema); upward displacement suggests small-volume lungs (e.g. pulmonary fibrosis), diaphragm dysfunction (commonly unilateral) or abdominal pathology. Both hemidiaphragms have a curved shape and a clear edge, and can be seen throughout their length. The right hemidiaphragm is 2 cm higher than the left. The level of the left hemidiaphragm, however, is influenced by the volume of gas in the bowel beneath it. The diaphragm becomes flattened by hyperinflation, best appreciated on the lateral chest X-ray. Costophrenic angle This sharp angle is obliterated and the lateral aspect of the diaphragm obscured by small pleural effusions. Chronic pleural abnormalities (e.g. from past infection) also blunt the costophrenic angle. Cardiophrenic angle The cardiophrenic angle is blunted by pleural effusion or thickening, and also obscured by pericardial fat pads. Sub diaphragmatic region Posteriorly the lung extends below the level of the diaphragm as viewed on the PA film, and pathology in this recess is easily missed; this area is best inspected on the lateral chest X-ray (see Fig. 13.13). On the left, gas in the stomach is commonly seen immediately below

SUMMARY 3 Viewing the chest X-ray Points to consider • Trachea (check the patient is not rotated). • Diaphragm • Costophrenic angle • Cardiophrenic angle • Subdiaphragmatic region • Cardiac silhouette • Behind the heart • Hilar shadows

• • • • • • • •

Horizontal (minor) fissure Vessels Lymphatics The lung fields Behind the clavicle, and lung apex Pleura Ribcage Soft tissues

623

cardiac silhouette may be enlarged because of a pericardial effusion, as well as dilatation of cardiac chambers. The cardiac diameter is usually less than 50% of the transthoracic diameter, but this is a crude measure and absolute measurements are more valuable; the cardiac diameter is less than 14.5cm in females and less than 15.5cm in males on a standard size X-ray, and a change in diameter of more than 1.5cm between two X-rays is significant. Emergency chest X-rays taken in the AP plane magnify the cardiac silhouette, as do supine films. Behind the heart This is a notorious blind spot on the PA chest X-ray, particularly if the film is underpenetrated. Digital techniques have improved the imaging of this region. The retrocardiac space is well seen on the lateral chest X-ray. A retrocardiac shadow can be caused by a hiatus hernia, which may contain a fluid level.

Fig. 13.11 Normal chest X-ray Careful inspection will include: 1. Trachea; 2. Diaphragm; 3. Costophrenic angle; 4. Cardiophrenic angle; 5. Subdiaphragmatic region; 6. Cardiac silhouette; 7. Behind the heart; 8. Hilar shadows; 9. Horizontal (minor) fissure; 10. Vessels; 11. Lung fields; 12. Behind the clavicle, and lung apex; 13. Ribcage (Fig. 13.65); 14. Soft tissues (Fig. 13.65).

the diaphragm and provides some indication of the position of the diaphragm even when it is obscured by pleural or pulmonary pathology. Perforation of an abdominal viscus causes free gas below the diaphragm, as may abdominal surgery. Cardiac silhouette The mediastinum requires careful inspection to confirm that it is centrally placed and that the cardiac silhouette is of normal configuration and size (see Fig. 12.16, p. 481). The heart may be displaced by pleural effusions, collapse of a lung or lobe, extensive fibrosis or pneumothorax. The heart is displaced to the left by a depressed sternum, obvious on clinical examination and confirmed by lateral chest X-ray. Many disorders alter the configuration of the heart and mediastinum (e.g. cardiac and aortic disease, mediastinal tumour and lymphadenopathy) and the

1 Fig. 13.10 2 Fig. 13.11 624

Hilar shadows The left hilum is always higher than the right. The hilar shadows are vascular and the midpoint (the 'hilar point') is where the upper lobe veins appear to cross the basal pulmonary artery. The lateral border of the hilum is concave. Both hila are similar in size, shape and density. Adenopathy (e.g. tumour) is a common cause of hilar enlargement, with increased density and convexity of the lateral border. Volume loss in the lung (e.g. fibrotic shrinkage) distorts the hilum and moves the hilar point towards the loss of volume. Any rotation of the chest X-ray will make the hila asymmetrical. Horizontal (minor) fissure This fissure is visible in 60% of normal chest X-rays, running from the centre of the right hilum, laterally and horizontally, to meet the sixth rib in the mid-axilla. The fissure is more easily visible when there is increased pleural fluid (e.g. cardiac failure), and encysted pleural fluid, within any fissure, can mimic a mass lesion. Localized volume loss in the right lung (e.g. collapse of the right upper lobe) moves the fissure from its normal position, and the position of the fissure is therefore an important landmark when assessing the chest X-ray for possible obstructing lesions in the right bronchial tree. Vessels On the X-ray of normal lungs virtually all pulmonary shadows, apart from a few central bronchi, are vascular. Arteries can be traced back to the hilum, but this is more difficult with pulmonary veins. As expected, vessels branch asymmetrically and taper towards the periphery. In the erect posture blood flow to the base of the lungs is greater than to the upper zones, and vessels at the bases are therefore larger. Vessels in the second intercostal space are usually less than 3 mm in diameter on a standard size Xray. The right basal artery measures approximately 15mm. If vessels are enlarged, this implies that either pressure or flow is increased. Loss of vascularity (e.g. in emphysema) makes the affected area of lung abnormally transradiant. Following

13

lobar collapse the remaining lobes of the lung enlarge, the vessels per unit volume of lung are reduced, and the chest X-ray shows increased transradiancy. The vascular pattern of the lung is therefore an excellent index of localized volume loss. It is not possible to see any vessels in a large bulla, cyst or pneumothorax (see Fig. 13.36). Lymphatics These are not normally visible on the chest X-ray, but when distended by fluid and surrounded by oedema (left heart failure) or infiltrated by tumour (lymphangitis carcinomatosa, see Fig. 13.64) they are seen as linear shadows. In left heart failure the distended lymphatics are seen as 1-2 cm horizontal lines extending inwards from the pleura, best seen just above the costophrenic angles (termed Kerley B lines, or septal lines). On the lateral film, septal lines are best seen immediately behind the sternum. The lung fields Pathological processes within the lung increase its density (with the exception of pulmonary embolism, emphysema and bullous disease). The abnormality may be localized (e.g. tumour mass, pneumonic consolidation) or generalized (e.g. diffuse fibrosis, pulmonary oedema). Abnormal shadowing is often characteristic of a particular condition or group of disorders, e.g. the bilateral apical shadowing with cavitation characteristic of pulmonary tuberculosis. Two particularly important radiological signs arise from increased pulmonary shadowing: the air bronchogram and the silhouette sign. • Air bronchogram. An air bronchogram (Fig. 13.12) is visible when alveoli are filled with exudate, transudate or other substance and the bronchi remain patent and filled with air. Important causes of an air bronchogram are pneumonia, cardiogenic pulmonary oedema, acute respiratory distress syndrome and alveolar cell carcinoma. • The silhouette sign. The outline of many structures on the chest X-ray is visible because of an interface between opaque tissue (e.g. heart or diaphragm) and air in the lung. If the lung adjacent to such structures also becomes opaque the 'silhouette' is lost (Fig. 13.13). Thus, collapse of the left lower lobe obscures the left or right hemidiaphragm and collapse of the lingula or right middle lobe obscures the left and right heart borders. 1

Fig. 13.12 Air bronchogram The chest X-ray shows bilateral air-space consolidation in a patient with pneumocystis pneumonia and the acute respiratory distress syndrome. Note that the major airways (arrowed) are seen as an 'air bronchogram' against the background pulmonary shadowing. 0

Behind the clavicle, and the lung apex The lung apex is an important site for pulmonary tumours and tuberculosis. If necessary, an apical view should be taken. Apical lesions are clearly demonstrated by CT scans. Pleura Pleural effusions usually obliterate the costophrenic angle, but loculated effusions and pleural thickening (e.g. mesothelioma, Fig. 13.74) are easily missed unless care is taken to look all around the inner aspect of the thoracic cage. In some cases the pleura may calcify (asbestos plaques, tuberculosis, past empyema). With pneumothorax the line of the visceral pleura is seen.

Fig. 13.13 Silhouette sign On the normal chest X-ray (Fig. 13.11) there is a clear silhouette of the structures making up the left mediastinum. Collapse of the left upper lobe (including the lingular division) causes an opacity adjacent to the mediastinum and the silhouette is lost. Note the volume loss of the left lung and shift to the left of the trachea.

625

Ribcage The ribcage should be symmetrical. All bones should be inspected, looking particularly for fractures and destructive processes (Fig. 13.65). Destruction of ribs in the axillary line is easily missed unless care is taken to follow the curve of each bone - the 'skyline' - down the outer aspect of both sides of the thoracic cage. Soft tissues Inspection of the soft tissues not uncommonly provides valuable information which is helpful to the interpretation of lung field abnormalities. Thus, a mastectomy may be noted in a patient with a pleural effusion.

The lateral chest X-ray A lateral chest X-ray is necessary in many patients with cardiorespiratory problems in whom a comprehensive assessment of thoracic structure is required. On the lateral film some areas are better seen than on the PA view, and three-dimensional localization of pathology is possible (Fig. 13.14). Retrosternal area This area is transradiant and should be of the same density as the retrocardiac area. With hyperinflation (e.g. emphysema) there is an increase in the size and transradiancy of the retrosternal space. Conversely, there is increased retrosternal shadowing with anterior mediastinal masses (see Fig. 13.70).

SUMMARY 4 Viewing the lateral chest X-ray Points to consider • • • •

Retrosternal area Retrocardiac area Vertebral bodies Diaphragm

Greater (oblique) fissure Horizontal fissure Trachea Cardiac silhouette

Retrocardiac area Careful inspection of this area, which should be transradiant, is useful because much of the retrocardiac space is not easily visible on the PA chest X-ray. Vertebral bodies Superimposed structures make it difficult to visualize the upper thoracic vertebral bodies, but they are seen progressively more clearly towards the lower thoracic spine. If this is not the case this suggests that there is pathology (e.g. consolidation or tumour) overlying the spine (see Fig. 13.72). Diaphragm Both hemidiaphragms should be visible. The gastric gas bubble is beneath the left hemidiaphragm (see Fig. 13.72). Greater (oblique) fissure The greater fissure passes from where the anterior quarter of the diaphragm meets the posterior three-quarters, through the hilum, and up to the level of the fourth thoracic vertebral body. Displacement of the fissure is an important sign of volume changes within the lobes of the lungs. Horizontal fissure The horizontal fissure between the right upper and middle lobes passes anteriorly and horizontally from the level of the hilum. Its position is altered by volume loss, particularly of the right upper and middle lobes. Trachea The tracheal air column is frequently better visualized on the lateral chest X-ray than on the PA film. Cardiac silhouette Intracardiac, aortic and pericardial calcification is often better seen on the lateral than the PA chest X-ray.

Supplementary radiological examination of the chest

626

FIG. 13.14 Lateral chest X-ray 1. Retrosternal area; 2. Retrocardiac area; 3. Vertebral bodies; 4. Diaphragm; 5. Greater (oblique) fissure; 6. Trachea; 7. Cardiac silhouette.

Films taken in expiration are useful for demonstrating small pneumothoraces not easily visible on the PA film at full inspiration and gas trapping in bullous disease or severe airways obstruction. Screening of diaphragm movement can detect paralysis (p. 718). For the differential diagnosis of pleural effusion from pleural thickening it can be useful to request a chest X-ray taken in the lateral decubitus position, in which free fluid will run along the lateral wall of the thorax.

CT scanning Many patients with cardiorespiratory disorders can be satisfactorily investigated with simple radiological techniques. However, CT scans of the thorax can provide remarkably detailed information of thoracic anatomy (Figs 13.15, 13.28, 13.35, 13.38, 13.53, 13.59, 13.70, 13.75). CT scanning is particularly useful for imaging the mediastinum and demonstrating pleural and chest wall disease. For many patients with bronchogenic carcinoma CT scanning is necessary to stage the disease adequately, particularly when surgery is being considered. Scanning is also valuable in the diagnosis and assessment of emphysema, bullous disease, bronchiectasis, lymphangitis carcinomatosis, alveolar proteinosis, interstitial lung disease, and pulmonary fibrosis. Spiral CT is often useful for the diagnosis of pulmonary emboli. Magnetic resonance scans are particularly good for

demonstrating lesions of bone lymphadenopathy.

and

for

detecting

13

Ultrasound Ultrasound is not generally a suitable technique for visualizing the lung but it is a good method for imaging pleural disease, both pleural fluid and pleural thickening, subdiaphragmatic regions (particularly when investigating a subphrenic abscess), and for assessing diaphragm movement.

Positron emission tomography (PET) The key role of PET is in differentiating malignant from benign lesions, at presentation or following therapy with surgery or radiotherapy. PET can avoid alternative difficult and invasive procedures. Tumours are metabolically active and will take up an appropriate tracer, the most commonly used being fluorodeoxyglucose (FDG). The tracer emits positrons, and therefore tumour issue can be imaged by the PET technique. Clinical studies of PET in the diagnosis of pulmonary lesions, including the common problem of evaluating solitary pulmonary nodules, demonstrate near 100% sensitivity and high specificity for the detection of malignancy. Occasional false positive scans are because non-malignant processes can have high metabolic activity (e.g. tuberculous granulomas). Some tumour types (carcinoid, bronchoalveolar cell) can have a relatively low metabolic activity and are not always identified by PET. PET may not diagnose very small nodules. Overall, PET is probably superior to CT in staging nodal disease. It is useful for detecting distant metastases and identifying that some abnormalities demonstrated by CT are benign rather than malignant. It is a good technique for the follow-up of patients after surgery and radiotherapy, both to detect residual disease and to diagnose tumour recurrence early.

FURTHER READING ON IMAGING THE THORAX Armstrong P A, Wilson A G, Dee P, Hansell D M (eds) 2000 Imaging of diseases of the chest, 3rd edn. London: Mosby.

BRONCHOSCOPY AND LUNG BIOPSY Fibreoptic bronchoscopy FIG. 13.15 CT scans of the thorax [A] Scan demonstrating the normal pulmonary vascular pattern. B CT scan in arantitrypsin deficiency. Bullous air spaces occupy most of the left lung and much of the right posteriorly: the peripheral vascular markings are lost and the mediastinum is shifted to the right.

Fibreoptic bronchoscopy is an invaluable technique for diagnosis, assessment and therapy in respiratory medicine. For most patients the procedure is undertaken using local anaesthesia and on a day-case basis. As skills in the technique have increased, so the risks have become small and the indications for fibreoptic bronchoscopy have widened

627

TABLE 13.8 Major indications for fibreoptic bronchoscopy • Investigation of symptoms, signs or radiological appearances suggesting pulmonary tumour (e.g. haemoptysis, hoarseness, mass, persistent consolidation, or volume loss on chest X-ray) • Diagnosis of pneumonias, tuberculosis and pulmonary shadowing in the immunocompromised host • Removal of mucus, pus and aspirated material from the bronchial tree (frequently on the intensive care unit) • Diagnosis and assessment of activity in diffuse interstitial lung disease (e.g. sarcoidosis)

(Table 13.8). Fibreoptic bronchoscopy is of great value in many patients with seemingly inexplicable respiratory problems. The development of video-assisted bronchoscopy has improved the views obtained, and the hardcopy colour images help in management (Fig. 13.16). Bronchogenic carcinoma Bronchogenic carcinoma is commonly visible as an obvious endobronchial tumour mass (Fig. 13.16B), in which case bronchial biopsy will confirm the diagnosis in 90% of cases. The flushing of saline over the surface of tumours and the subsequent aspiration of this fluid, or the brushing of the surface of tumours, provides cells for cytological examination, and these investigations have a diagnostic yield of 60-80% for visible endobronchial disease. In assessing the extent of malignant disease fibreoptic bronchoscopy is of great importance, allowing visualization of the proximal spread of the tumour, which is one of the factors that determines the feasibility and likely extent of any surgical resection. In patients with extrabronchial tumours situated adjacent to a large airway, a needle can be passed through the bronchial wall to acquire a cytological specimen. In selected patients endobronchial tumour can be partially destroyed by a laser beam, transmitted through a fibre passed down the biopsy channel of a fibreoptic bronchoscope. In patients with large central tumours causing breathlessness or haemoptysis that has not responded to radiotherapy, such palliative therapy can be of substantial benefit. In appropriate patients bronchoscopy can be used to introduce stents for the dilation of narrowed airways (usually caused by extrinsic compression by tumour) or treatment with local intrabronchial radiotherapy. Haemoptysis All patients with haemoptysis who are at risk of carcinoma of the bronchus should be bronchoscoped; approximately 10% of patients with a chest X-ray that does not obviously

1

628

Fig. 13.12

2

Fig. 13.13-13.17

B FIG. 13.16 Bronchoscopy

1

A View of right and left main bronchi at fibreoptic bronchoscopy; the appearance is normal. B View of right main bronchus, occluded by bronchogenic carcinoma. ©

suggest tumour will nevertheless have malignancy. When bleeding is not due to tumour (for example in bronchiectasis) bronchoscopy is useful to determine the site of bleeding, which is important should bleeding become a serious problem and bronchial arterial embolization or thoracic surgery be considered. Interstitial lung disease Sarcoidosis is the most common cause of interstitial lung disease and in the majority of cases the diagnosis is easily confirmed by bronchoscopy. Overall, a bronchoscopic

antibiotics, bacterial culture is frequently negative and staining techniques give the most valuable information. Similarly, patients with pulmonary tuberculosis do not always have sputum, and in these cases bronchoscopy and lavage of the affected lobes is an efficient technique for confirming the diagnosis. In miliary tuberculosis BAL is commonly positive, and TBB will confirm the diagnosis in virtually all cases. Bronchoscopy with BAL and TBB is of great value in the investigation of pulmonary shadowing in the immunocompromised host (p. 638). The importance of these techniques has increased since the advent of AIDS (p. 638).

13

Research indications for bronchoscopy Much important research data can be acquired through bronchoscopy. Bronchial biopsy and lavage can yield information on inflammatory changes in the airway and lung parenchyma and its response to treatment.

Bronchoscopy on the intensive care unit FIG. 13.17 Bronchoscopy Using the flexible fibreoptic bronchoscope, biopsy forceps can be passed into peripheral bronchi and samples of lung obtained by transbronchial biopsy.

transbronchial biopsy has a positive yield of 90% (p. 683). With the technique of transbronchial biopsy (TBB) a small fragment of lung tissue is obtained when the forceps removes the bronchial wall of the carina of two small peripheral bronchi (Fig. 13.17). In pulmonary fibrosis, including cryptogenic fibrosing alveolitis (p. 695), TBB is less helpful. The small samples obtained are often not representative, and if histological confirmation of the diagnosis is required, thoracoscopic or open lung biopsy is preferable. TBB can be diagnostic in the uncommon conditions of alveolar proteinosis, histiocytosis X and pulmonary disease due to inorganic dusts. Bronchoalveolar lavage At fibreoptic bronchoscopy saline is instilled into the periphery of the lung, with the tip of the bronchoscope wedged in the segmental or subsegmental bronchus, and the fluid is subsequently gently aspirated. The lavage fluid contains thousands of cells (macrophages, lymphocytes, neutrophils and eosinophils), the numbers of which, and their differential count, give information on the nature of the interstitial lung disease. In occasional circumstances bronchoalveolar lavage (BAL) can be diagnostic, as in alveolar proteinosis, pulmonary haemosiderosis and histiocytosis X. Pulmonary infections Some patients with pneumonia fail to show an adequate response to chemotherapy, yet do not have sputum for microbiological analysis. In such cases BAL provides helpful information, although in patients already receiving

Many patients on the intensive care unit, as well as other ill patients, often in the immediate postoperative period, develop sputum retention, basal atelectasis or pulmonary collapse. Bronchoscopy and BAL are effective at removing impacted secretions. BAL is frequently superior to conventional catheter suction in the diagnosis of intensive care unit pneumonias. Bronchoscopy is also valuable for the inspection of the bronchial tree, particularly the trachea, for catheter or endotracheal tube trauma, or stenosis.

Rigid bronchoscopy As a diagnostic technique, rigid bronchoscopy, which requires a general anaesthetic, has been largely superseded by fibreoptic bronchoscopy, but remains of great value in the following situations: • In the bronchoscopy of smaller children; • In patients who have a foreign body impacted in the bronchial tree; • For the biopsy of vascular tumours, particularly adenomas; • For obtaining a large biopsy sample to confirm submucosal malignancy not demonstrated by the small superficial biopsies that can be taken using the fibreoptic bronchoscope.

Transthoracic needle biopsy Several techniques are available to biopsy the lung and diagnose pathological processes, most commonly tumours, within it. The usual technique is fine needle aspiration or biopsy under radiological control. If it is performed using local anaesthesia, patients seldom experience pain and major complications are rare. A small pneumothorax is common (30%), particularly in patients with emphysema,

629

and a few patients (<5%) will require tube drainage. The overall mortality is very low and the technique is now used extensively in the investigation of pulmonary disease. In practice, the most common indication is to provide confirmation of malignancy for tumours that are not visible at fibreoptic bronchoscopy. With experience and expert laboratory support a yield of 90% or more is obtained in malignant disease. A further important use of the technique is in the investigation of pulmonary infections not diagnosed by more conventional techniques, particularly fungal, protozoal and mycobacterial infections in the immunocompromised host.

Mediastinoscopy and mediastinotomy These techniques are mainly used to diagnose and stage bronchogenic carcinoma, but occasionally are required to confirm other malignancies (e.g. lymphoma) or to diagnose the causes of adenopathy (e.g. sarcoidosis or tuberculosis). At mediastinoscopy an incision is made above the suprasternal notch, the tissues anterior to the trachea are dissected and the mediastinoscope is inserted. Hilar, subcarinal and paratracheal nodes can be inspected, biopsied or removed. The involvement of mediastinal nodes in bronchogenic carcinoma is common (40% of all cases) and a positive diagnosis is achieved in sarcoidosis in more than 80% of cases. Mediastinotomy is occasionally required to evaluate areas of the mediastinum not accessible at mediastinoscopy, particularly the subaortic fossa on the left side, and most commonly the surgeon gains access through the second left costal cartilage. The development of CT scanning has reduced the need for mediastinoscopy and mediastinotomy. In carcinoma of the bronchus a CT scan which demonstrates a normal mediastinum obviates the need for surgical evaluation prior to thoracotomy. However, if the CT scan demonstrates adenopathy, such an evaluation may still be required to confirm that the nodes are involved by tumour and are not enlarged due to reactive hyperplasia.

RECENT ADVANCES IN THORACOSCOPY WITH VIDEOSCOPIC IMAGING Thoracoscopy allows direct inspection of the pleura overlying the lung and lining the chest wall (Fig. 13.18). Videoscopic imaging and the insertion of surgical instruments through additional small incisions greatly extends the thoracoscopic technique and allows 'keyhole' surgery in the thorax. Patients are treated under general anaesthesia and with a double-lumen endotracheal tube to allow controlled deflation of the appropriate lung. Common indications for this type of surgery include pleural biopsy (most often in the diagnosis of pleural effusion), lung biopsy, treatment of pneumothorax, fashioning of a pericardial window, sympathectomy, and biopsy of mediastinal masses. The indications continue to expand and the technique can be used for staging of lung cancer, lobectomy, management of bullous disease and volume reduction surgery for chronic obstructive pulmonary disease. Some surgeons have performed pneumonectomy and oesophagectomy via keyhole surgery. The advantage of minimally invasive thoracic surgery is reduced pain postoperatively. It remains to be clearly established whether time in hospital, time off work, morbidity and mortality are reduced. Disadvantages of the technique include longer operation times, high cost of equipment, and the need for specialist training.

Thoracoscopy The development of video-assisted thoracoscopy has resulted in a great increase in 'keyhole' thoracic surgery. FIG. 13.18 Thoracoscopy

Open lung biopsy Open lung biopsy is most commonly undertaken for diffuse interstitial lung disease, atypical tumours and, occasionally, serious pulmonary infections. A limited submammary thoracotomy allows biopsy of the anterior segments of all lobes, as well as inspection of the pleura and 1

630

Fig. 13.18

2

MCQ13.6

1

View of lung (left) and parietal pleura (right) at thoracoscopy; there are multiple metastatic tumour deposits on the pleura.

mediastinum. Complications (approximately 5%) include infection and chronic pneumothorax, and there is a small mortality (less than 4%), reflecting for the most part the serious nature of the underlying illness in many of the patients. For peripheral lesions in the lung, or diffuse lung abnormalities, thoracoscopic minimally invasive techniques allow access not just to the pleura but to the underlying lung.

RESPIRATORY INFECTIONS UPPER RESPIRATORY TRACT INFECTIONS Most upper respiratory tract infections (URTIs) are viral and the majority of these are the results of infection with rhinoviruses of the picornavirus family. However, adenoviruses, coronaviruses, coxsackie viruses, echoviruses, influenza viruses, parainfluenza viruses and the respiratory syncytial virus (RSV) can all cause upper respiratory tract infections. URTIs represent the most common of all illnesses and are responsible for approximately half of all time lost from work. In addition to URTI the influenza viruses, parainfluenza viruses 1 and 2, adenoviruses and RSV can cause acute infection of the larynx, trachea and major bronchi, and are causes of croup. URTIs are most common in the late winter months and early spring. The most frequent manifestation of these viral infections is the common cold (acute coryza), which may be complicated by secondary bacterial infection and subsequently by sinusitis, otitis media, obstruction to the eustachian tubes and infection of the lower respiratory tract. Viral laryngotracheobronchitis, particularly in children, may be complicated by severe laryngeal oedema and life-threatening croup. Croup is also a feature of acute epiglottitis in children, and occasionally in adults, and is most commonly due to Haemophilus influenzae type B infection. The swelling of the epiglottis and surrounding soft tissues can rapidly produce respiratory distress. Unlike in laryngitis, the voice is not hoarse. The epiglottis is hugely swollen and attempts to examine the throat can precipitate total upper airway obstruction. In most patients with URTI the illness is self-limiting and symptoms subside after a few days. Antibiotics are not generally required unless there is acute epiglottitis due to H. influenzae, in which case intravenous chloramphenicol, or a cephalosporin such as cefotaxime or ceftriaxone, is indicated. However, in patients in whom secondary bacterial infection is both likely and potentially serious (e.g. in chronic bronchitis and emphysema) immediate treatment with antibiotics is justified. Oral amoxycillin, co-amoxiclav or tetracycline in standard doses are suitable. The development of stridor requires careful observation in hospital, with anaesthetic support for possible emergency intubation. Patients with stridor, particularly children with croup, adopt the posture that facilitates adequate ventilation and they should be allowed to do so. Inhalation of warm, humidified air is helpful. Given early, high-dose corticosteroids reduce the need for intubation, and in those who are intubated they shorten the time for which a tube is necessary. Nebulized corticsteroid therapy is helpful.

Sinusitis Sinusitis (infection of the paranasal sinuses) is a common complication of URTI and the bacteria most frequently

involved are H. influenzae and Streptococcus pneumoniae. Amoxycillin, trimethoprim or an oral cephalosporin in standard doses may be given. Facial pain and tenderness, headache, nasal discharge and a postnasal drip are the usual features. Radiology may demonstrate fluid levels within the maxillary sinuses or show sinus mucosal thickening. Sinusitis may complicate the nasal obstruction of allergic rhinitis, in which case topical steroid therapy is useful, in addition to oral antibiotics. Topical steroid can be administered by nasal spray (e.g. beclometasone or fluticasone) but if the nose is obstructed steroid nasal drops, incorporating an antibiotic, are usually more effective (e.g. betamethasone plus neomycin). Nasal drops are best instilled with the head held forward (between the knees) so that the fluid is retained within the nostril and the drugs able to be absorbed into the oedematous nasal mucosa. Persistent or recurrent sinusitis may require surgical drainage. Sinusitis is a common feature of patients with bronchiectasis (occasionally, both are manifestations of the immotile cilia syndromes) and it is equally important to control infection at both sites.

13

LOWER RESPIRATORY TRACT INFECTIONS Acute bronchitis Acute bronchitis is common in smokers and in patients with asthma, chronic bronchitis and emphysema. However, it also occurs in otherwise healthy individuals, usually due to viral infections. In both groups, when bacterial infection occurs, the bacteria most often incriminated are Strep, pneumoniae or H. influenzae. These infections cause an acute inflammation of the trachea and major airways, and as a consequence of this there is chest pain, commonly experienced as a raw feeling maximal on deep inspiration, as well as chest tightness. The patient may wheeze. In patients with pre-existing chronic airflow limitation the development of acute bronchitis can cause severe breathlessness. Acute bronchitis causes an irritating, persistent dry cough, although after 1 or 2 days patients produce small amounts of mucoid thick sputum which subsequently becomes more plentiful and purulent. Appropriate antibiotic therapy is amoxycillin, erythromycin or tetracycline. In patients with chronic airflow limitation, aggressive therapy with bronchodilators is important, as acute bronchitis can precipitate a worsening of respiratory failure.

PNEUMONIA 2 Pneumonia remains an important clinical problem, causing many more deaths than any other infectious disease in the UK, both in the community and in hospitals. In recent years the aetiology of the pneumonias has changed with, for example, an increased incidence of opportunistic pneumonias in the immunocompromised host, particularly patients with HIV infection.

631

TABLE

13.9 Differential diagnosis of a presumed pneumonia

Bronchial obstruction with distal collapse/consolidation Pulmonary infarction Empyema Pulmonary vasculitis or haemorrhage Alveolar cell carcinoma Leukaemia, lymphoma, Kaposi's sarcoma and other pulmonary tumours Pulmonary eosinophilia Allergic alveolitis Pulmonary drug reaction Radiation injury

TABLE 13.10 Important causes of community-acquired pneumonia Streptococcus pneumoniae Mycoplasma pneumoniae Haemophilus influenzae Staphylococcus aureus

Legionel/a pneumophila Viruses Mycobacterium tuberculosis

Although pneumonia is the most common cause of fever and pulmonary shadowing, other conditions must be considered in the differential diagnosis (Table 13.9). Having arrived at a clinical diagnosis of pneumonia, the next decision is what therapy to institute. For the adult patient, the clinical background of the pneumonia serves to reduce the range of possible pathogens. The most useful classification of the pneumonias is therefore a clinical one.

Community-acquired pneumonia Community-acquired pneumonia is commonly due to a limited and predictable group of pathogens (Table 13.10). Patients with pneumonia usually have fever, cough, breathlessness, abnormal chest signs and X-ray shadowing. Specific pneumonias may have additional features and Mycobacterium tuberculosis remains an important pathogen to consider (p. 645). Pneumococcal pneumonia Pneumococcal pneumonia is more common in the winter months and upper respiratory tract viral infections are a predisposing factor. Fever is often high and there may be rigors. Pleurisy is common and the cough, which is initially dry and painful, subsequently becomes productive of rusty sputum. Altered blood from the congested lung tissue gives

1

632

Fig. 13.19

SUMMARY 5 Classification of pneumonia • • • • • •

Community-acquired pneumonia Hospital-acquired (nosocomial) pneumonia Recurrent pneumonia Aspiration pneumonia Pneumonia in the immunocompromised host Unusual pneumonias

SUMMARY 6 Typical features of pneumococcal pneumonia • • • • • •

More common in winter Rapid onset High fever/rigors Dry cough initially Pleurisy common Labial herpes common

the sputum its characteristic colour. Labial herpes simplex is common (30% of cases). In some cases the onset of symptoms in pneumococcal pneumonia can be very rapid and patients may be critically ill within a few hours. In the usual case clinical examination demonstrates crackles more than classic signs of consolidation, and a pleural rub is common. The chest X-ray shows hazy shadowing, often with an air bronchogram, in any lobe, although lower lobes are the most frequently involved. In the majority of cases the consolidation does not involve the whole lobe. 1 Mildly or moderately ill patients can be managed adequately at home and are satisfactorily treated with oral amoxycillin or erythromycin. Severely ill patients should be transferred to hospital for the adequate treatment of pain, dehydration and hypoxaemia. A Gram stain of the sputum will show typical Gram-positive diplococci, and blood cultures will be positive in 25-40% of untreated cases. If the specific diagnosis is established, benzylpenicillin is the treatment of choice, but patients also do well with amoxycillin. In the UK about 10% of pneumococci are partially or fully resistant to penicillin. Currently, penicillin or amoxycillin are adequate for mild to moderate infections, unless penicillin resistance is detected, in which case cefotaxime or ceftriaxone is required, especially in the more severely ill patient who requires hospitalization. Recovery from pneumococcal pneumonia is usually rapid, although the X-ray may take several weeks to return to normal, as is the case with most pneumonias. In patients who are more severely ill and who are bacteraemic, mortality may be as high as 25%. Of all pneumonias, pleural effusion is most common in association with pneumococcal infection. Empyema complicating the pleural effusion is now relatively rare and occurs in 3% of cases with positive blood cultures. In such blood culture-positive patients pneumococcal pneumonia can be complicated by pericarditis, endocarditis, septic arthritis, peritonitis, cellulitis and, on rare occasions, meningitis.

Haemophilus influenzae pneumonia It is unusual for H. influenzae to cause pneumonia in previously fit individuals. However, it is probably the most common cause of infection in patients with pre-existing lung disease, particularly in those with chronic bronchitis and emphysema. It is the organism most often responsible for the exacerbations that occur in chronic bronchitis, sometimes complicating an initial viral infection. The chest X-ray distinguishes haemophilus bronchopneumonia from simple infective bronchitis by showing shadowing, usually as nodules 0.5-3.0 cm in diameter at both bases, and on examination there may be bronchial breathing. The same picture of bronchopneumonia is sometimes seen with staphylococcal and pneumococcal infection. Bronchopneumonia is very common. Patients are often wheezy and progressively more breathless. There is usually fever, but in elderly and debilitated patients the temperature can be normal. Cough and purulent sputum are prominent features. As a consequence of diffuse airway inflammation and intraluminal sputum, patients with chronic bronchitis are frequently precipitated into respiratory failure and cor pulmonale. The first-line treatment of H. influenzae bronchopneumonia is intravenous cefuroxime unless the isolate is known to be sensitive to amoxycillin. Staphylococcal pneumonia Staphylococcal infection causes more necrosis than other organisms responsible for pneumonia, and there is a high incidence of abscess formation. In the community staphylococcal pneumonia is not common (about 1% of cases), but it is an important and serious complication of influenza and is therefore more common during influenza epidemics and during the winter months. Staphylococcal lobar pneumonia can be fulminant and rapidly fatal. In most cases the clinical picture is similar to that of pneumococcal pneumonia, but haemoptysis is more common. Cavitation is unusual in community-acquired pneumonias, with the exception of staphylococcal or particularly virulent serotype 3 pneumococcal infections. The chest X-ray typically shows bilateral consolidation, usually basal, with abscesses that are thin-walled and cyst-like. Staphylococcal lung abscesses may rupture into the pleural cavity, resulting in pneumothorax or pyopneumothorax. In the community, the development of pneumonia as a complication of influenza should be treated with flucloxacillin in addition to amoxycillin. Following admission to hospital, a Gram stain of the sputum is helpful in confirming the diagnosis and blood cultures are frequently positive. With a definite diagnosis treatment should be with intravenous flucloxacillin (l-2gqds) and fusidic acid. Although the majority of patients make a good recovery from staphylococcal pneumonia, the combination of influenza A infection and staphylococcal pneumonia still carries an appreciable mortality. Important complications include the haematogenous spread of infection to brain, bone and other organs, and occasionally patients can develop acute bacterial endocarditis.

Legionella pneumonia Legionella pneumonia was first recognized in 1976 after an outbreak of pneumonia with a high mortality at a Legionnaires' conference in Philadelphia. The pathogen is now known as Legionella pneumophila, a small Gramnegative coccobacillary organism. Infection is acquired from contaminated water, usually in air-conditioning systems and showers in hotels and hospitals, and transmission does not occur from person to person. Most cases have been in middle-aged or elderly males and have occurred in the summer months. Legionella is a cause of opportunistic chest infection in immunocompromised patients. Clinically, legionella pneumonia resembles viral or mycoplasma infection. There is a cough but little sputum, which is mucoid, not purulent. Patients are frequently severely ill, with high fever, rigors, confusion, myalgia, abdominal pain, vomiting and diarrhoea. Hyponatraemia, hypoalbuminaemia and haematuria (50% of cases) are common. The white cell count may be normal or modestly elevated, but is seldom above 15000 x 106/L. Gram stain of the mucoid sputum reveals no organisms. The chest X-ray most commonly shows patchy shadowing, which can be bilateral and which progresses to lobar consolidation. The diagnosis is usually retrospective and based on a greater than fourfold increase in the indirect fluorescent antibody titre. Legionella antigen can be detected in the urine, and this has proved to be a more useful test in the acute setting. The organism can be isolated from lung tissue, pleural fluid and blood. Treatment is with erythromycin (1 g 6-hourly) for a period of 3 weeks or clarithromycin (SOOmgbd). If the diagnosis is strongly suspected, or confirmed, and particularly in patients who fail to respond, or who are severely ill, rifampicin should be added. Ciprofloxacin may also be effective. The overall mortality is 15%, some patients dying despite appropriate antibiotic therapy.

13

Viral and viral-like pneumonia Viral and viral-like pneumonias (Table 13.11) are characterized by fever, systemic symptoms (e.g. myalgia, headache) and a normal or near-normal white cell count. Respiratory syncytial virus infection is important in children. Influenza and measles are frequently complicated by serious bacterial infection, particularly staphylococcal. Cytomegalovirus is a cause of pneumonia in the immuno-

TABLE 13.11 Organisms causing viral and viral-like pneumonias Mycoplasma pneumonias Respiratory syncytial virus Influenza Measles Cytomegalovirus Adenoviruses

Parainfluenza virus Rhinovirus Varicella (chickenpox) Coxie/la bumetii (Q fever) Chlamydial pneumonias (psittacosis, ornithosis)

633

compromised host (p. 638). Of the remaining causes, Mycoplasma pneumoniae, Chlamydia psittaci and Coxiella burnetii are the most important. Mycoplasma pneumonia M. pneumoniae is the most important agent causing the socalled 'atypical' pneumonias. In the past, this organism was the cause of major outbreaks of pneumonias in the armed forces; many cases still occur in clusters, and epidemics typically occur every 3 or 4 years. Infection is caused by an organism of the mycoplasma group, the smallest known free-living organism. Most patients are aged between 15 and 30 years. After an incubation period of between 1 and 3 weeks the patient develops symptoms suggestive of viral pneumonia, with systemic upset, arthralgia and myalgia being particularly common. Typically, the white cell count is not raised. The appearances on chest X-ray are very variable, and although segmental and subsegmental shadows are most common, a lobar pattern can also be seen. In mycoplasma pneumonia cold agglutinins are present in approximately 50% of cases and serological investigations demonstrate antibodies to mycoplasma in most instances. Mycoplasma titres may be raised for several years following infection. The most effective therapy for mycoplasma infection is erythromycin, tetracycline or doxycycline. Complications can include a haemolytic anaemia and renal failure, owing to the presence of cold agglutinins, as well as meningism, central nervous system involvement and myocarditis. For most patients the prognosis is excellent. Psittacosis and ornithosis Chlamydia psittaci infection is transmitted from parrots to humans (psittacosis) or from other birds to humans (ornithosis). The organism (intermediate between a virus and a rickettsial organism and an obligate intracellular bacterium) is in the dust derived from excreta and feathers. The birds do not always appear to be ill. There is usually a history of close contact with birds (e.g. pigeon racers). The ensuing illness is rather like influenza. There is a cough, mucoid sputum, and sometimes haemoptysis. The chest X-ray shows patchy consolidation. A complicating bacterial pneumonia is common. Diagnosis is best made by a rising titre for chlamydia antibodies, and effective treatment is tetracycline or doxycycline. Erythromycin may be used if these drugs are contraindicated. Chlamydia pneumoniae causes a pneumonia, usually in young adults within the community, similar to mycoplasma pneumonia. Diagnosis is serological and therapy is the same as for psittacosis.

634

Rickettsial pneumonia (Q fever) The rickettsial organism (Coxiella burnetii) is transmitted from animals (most commonly cattle and sheep) to humans by dust inhalation. Slaughterhouse workers are most often infected. There is an abrupt onset of a viral-like illness with headache and meningism. Respiratory symptoms are less prominent. The chest X-ray shows patchy consolidation.

Diagnosis is confirmed by a rising antibody titre. A prolonged course of tetracycline (or erythromycin, combined with rifampicin in severe cases) is effective, as is chloramphenicol. Rickettsia can cause endocarditis (p. 574). Treatment of community acquired pneumonia Because of the likely pathogens, patients respond well to a macrolide antibiotic (e.g. erythromycin or clarithromycin) plus, if systemically unwell, amoxycillin.

Hospital-acquired (nosocomial) pneumonia Up to 5% of patients admitted to hospital for other causes subsequently develop a pneumonia; the mechanisms of infection are depicted in Figure 13.19. Particularly important predisposing factors are cigarette smoking, chronic lung disease, advanced age, obesity, the effects of anaesthesia and surgery, and prior use of broad-spectrum antibiotics. The causative organisms for hospital-acquired pneumonias are different from those responsible for communityacquired pneumonia and, in particular, Gram-negative organisms are responsible for 50% of cases. Anaerobic infections are also important (Table 13.12). The prior use of broad-spectrum antibiotics and the consequent colonization of the oropharynx with Gramnegative bacilli reduces the value of sputum culture and makes it difficult to find the cause of hospital-acquired pneumonias. The management of these pneumonias involves close collaboration between clinical staff and microbiologists. Gram stain of the sputum may be more useful than culture. Blood cultures will be positive in up to 25% of cases. For potentially serious pneumonias clinical management is helped by a precise diagnosis, and bron-

FIG. 13.19 Factors predisposing to hospital-acquired pneumonia

TABLE 13.12 Organisms responsible for hospital-acquired pneumonia Organism

% of cases

Gram-negative bacteria (especially Kiebsiel/a spp. and Pseudomonas spp.) Staphylococcus aureus Anaerobes Streptococcus pneumoniae Others (legionella, fungi etc.)

50 15 9 6 20

TABLE 13.13 Important causes of recurrent pneumonia Diffuse bronchopulmonary disease Bronchiectasis Cystic fibrosis Chronic bronchitis, emphysema and chronic asthma

Immotile cilia syndrome Immune deficiency states HIV infection Hypogammaglobulinaemia Myeloma

TABLE 13,14 Factors predisposing to aspiration pneumonia Altered consciousness Drug overdose Anaesthesia Epilepsy Cerebrovascular accident Alcoholism Dysphagia and oesophageal disease Stricture Achalasia Oesophageal pouch Fistula Hiatus hernia Reflux

13

Neurological disorders Pseudobulbar palsy Motor neuron disease etc. Nasogastric tubes Severe dental and upper airways sepsis Terminal illness

Local bronchial obstruction Tumour (benign or malignant) Adenopathy Foreign body

Local bronchopulmonary disease Bronchiectasis Congenital abnormalities (sequestration) Recurrent aspiration pneumonia

choscopy can yield useful information. The wide range of potential pathogens demands the use of broad-spectrum antibiotics, including good Gram-negative cover, in the initial therapy of hospital-acquired pneumonia. Mild to moderately ill patients can be treated with intravenous cefuroxime, with added metronidazole if aspiration is likely, with augmentin as an alternative. More severely ill cases require intravenous piperacillin-tazobactam plus gentamicin, or ceftazidime and amoxycillin. In the ICU appropriate treatment regimens would be: augmentin and clarithromycin ± rifampicin; cefuroxime (or cefotaxime, or ceftriaxone) and clarithromycin ± rifampicin; or a fluoroquinolone and gram-positive cover.

Recurrent pneumonia Prior to making a diagnosis of recurrent pneumonia, the possibility of a non-infective cause of recurrent pulmonary problems should be considered. Alternative diagnoses will include pulmonary infarction, pulmonary eosinophilia (including bronchopulmonary aspergillosis) and asthma. Recurrent pneumonia is unusual without a predisposing factor (Table 13.13). Any cause of obstruction to a bronchus can result in recurrent or persistent pneumonia, but malignant tumours are most important (p. 700). In elderly or debilitated patients, or those with impaired cough reflex or vocal cord

function, repeated episodes of pneumonia are commonly caused by aspiration of gastric contents (see below).

Aspiration pneumonia Many patients with aspiration pneumonia have dental sepsis, a predisposition to aspiration, or both (Table 13.14). Infection is usually with anaerobic organisms derived from the upper respiratory tract. Aspiration pneumonia may be acute, extensive and progressive, or it may run a more subacute course and progress to lung abscess formation. When cavitating anaerobic pneumonias communicate with the bronchial tree the sputum is both copious and foul-smelling. The site of the pneumonia or lung abscess will depend on the position of the patient at the time of aspiration. Aspirated material enters the right lung more easily than the left, and will enter the lower lobes when the subject is standing, and the apical segment of the lower lobes or the posterior segment of the upper lobes when supine. An important cause of anaerobic pneumonias and lung abscess is bronchial obstruction (e.g. with bronchogenic carcinoma), and if the diagnosis of aspiration is in doubt bronchoscopy is indicated. Radiological studies of swallowing will frequently be rewarding. Anaerobic infection causes considerable tissue destruction, with abscess formation, and empyema and metastatic abscesses can also occur. Prompt treatment avoids such problems. Most upper respirating tract anaerobes are sensitive to penicillin. Early aspiration pneumonias are adequately treated with amoxycillin, but more severe infections and lung abscesses require parenteral cefuroxime and metronidazole initially, followed by oral amoxycillin and metronidazole (depending on cultures). Postural drainage is important, and antibiotic therapy should be continued for 6 weeks or more to minimize lung destruction.

635

CASE STUDY 13.3 PLEURITIC CHEST PAIN AND NIGHT SWEATS IN A 54-YEAR-OLD IRISH BUILDING LABOURER A 54-year-old Irishman was referred to hospital by his GP because of leftsided pleuritic chest pain and cough. The patient had become increasingly breathless over the previous week and had experienced drenching night sweats. He had lost 1 stone in weight. Although he had a cough, there was no sputum or haemoptysis. He had worked all his adult life as a building labourer, but was off work because of osteoarthritis of the left hip. A heavy smoker, he also drank 5-6 pints of beer on most days. On examination his weight was 80kg and temperature 37.7°C. There were extensive dental caries and poor oral hygiene. The hands were nicotine stained and there was equivocal finger clubbing. There was bronchial breathing and inspiratory crackles over the midzone of the left lung anterolaterally. The trachea was central, there was no lymphadenopathy, and neither liver nor spleen were palpable. Blood tests showed a white cell count of 7.9, Hb 10.9 and albumin 27. The chest X-ray was reported as showing air-space consolidation in the left lung, with a fluid level (Case Fig. 13.3.1).

Question 1. What is the most likely diagnosis, and what would you include in the differential? Discussion The differential diagnosis would include a cavitating pneumonia/lung abscess, tuberculosis, a cavitating tumour and vasculitic disorders (e.g. Wegener's granulomatosis). The development of a lung abscess seems most likely. The chest X-ray findings, fever and night sweats are consistent with this diagnosis.

636

CASE FIG. 13.3.1 Note the extensive air-space shadowing in the left lung. There is increased density in the left upper and midzones, and loss of clarity of the left heart border, demonstrating extensive involvement of the left upper lobe. There is a fluid level within a large cavity in the left midzone. There is no volume loss and the right lung is clear.

Although most patients with a lung abscess and a fluid level have sputum - often copious - this is not invariably the case. Similarly, the white cell count is raised in most but not all cases. A lung abscess can occur as a result of primary infection or secondary to other conditions (see p. 647). The air-space consolidation in this case suggests the abscess is a complication of a pneumonia. Cavitating pneumonia leading to lung abscess formation can occur particularly with Klebsiella and staphylococcal infection. A common predisposing factor for a secondary lung abscess is bronchial obstruction, for example by a tumour. Infected pulmonary infarcts can also cavitate. In this case, where the lung abscess was almost certainly preceded by a pneumonia, this could have followed aspiration as a result either of excessive alcohol intake or poor

dental hygiene. Tuberculosis is a possibility, with the patient's ethnic origin, alcohol intake and normal white cell count making the diagnosis worth considering. The radiology, however, is not typical. In cavitatory TB the chest X-ray often shows bilateral abnormalities, and it is usually the apex of the upper or lower lobes that are most affected. Malignant disease must be considered. The patient is a smoker and in his work may have been exposed to asbestos. Cavitation is most common in squamous cell carcinoma, in which case finger clubbing might be expected. Although the fever and night sweats strongly suggest infection, it is possible that with a tumour there could be infection distal to obstruction. However, the radiology does not indicate volume loss. The patient has no clear

predisposing factors for pulmonary embolic disease or clinical features to suggest a vasculitic disorder such as Wegener's. Question 2. The patient had a CT scan of the thorax: how does this investigation help support a particular diagnosis (Case Fig. 13.3.2) The CT scan shows consolidation and cavitation (in keeping with lung abscess complicating a pneumonia), but no mass lesion, bronchial obstruction or volume loss. Nevertheless, because a tumour or tuberculosis were possible diagnoses, the patient was bronchoscoped. No endobronchial lesion was seen and bronchial lavage demonstrated no acid-fast bacilli. Question 3. Given the likely diagnosis of an acute pneumonia complicated by abscess formation and cavitation, what are the most likely pathogens? 4. Which antibiotics would you choose and for how long would you give them? 5. What is a rare but important complication of lung abscess?

CASE FIG. 13.3.2

lAl There is confluent consolidation with a well defined air bronchogram. [B] This scan, taken from just below scan A, shows the large cavity with its fluid level, and an air bronchogram. There is no mass lesion or evidence of bronchial obstruction.

Inhalation pneumonias Aspiration pneumonia is the term used to denote aspiration of organisms into the lower respiratory tract (usually anaerobes); inhalational pneumonias refer to the consequences of inhaling non-infected particulate matter, fluids and irritant gases. The inhalation of gastric contents during anaesthesia or other causes of reduced consciousness, or in childbirth (Mendelson's syndrome) produces a chemical pneumonitis and respiratory distress. When severe, the features are those of the acute respiratory distress syndrome (p. 697). Management includes antibiotics, and metronidazole provides good cover for anaerobic organisms, but corticosteroids are not helpful.

Common pathogens causing lunj abscess with pneumonia include Klebsiella, Pseudomonas, Staphylococcus and anaerobes, and therefore unless a specific organism can be identified broad-spectrum antibiotics are required. In this patient no organisms were identified from culture of blood or bronchial lavage fluid (perhaps because antibiotic therapy had been started by the GP prior to hospital admission). The possibility of aspiration when drunk, and the poor dental hygiene, suggested anaerobic infection was possible. The patient was treated with intravenous cefuroxime, metronidazole and flucloxacillin. Within days there was substantial symptomatic improvement, but it was 12 days before he became afebrile. It is not uncommon for resolution of fever and radiological changes to be slow with a lung abscess. After 2 weeks' therapy he was changed to oral antibiotic treatment, and in total he was treated for 6 weeks. Gradually, over a period of 2 months, the chest X-ray cleared. This patient made an excellent recovery, but a rare and serious complication of lung abscess is cerebral abscess.

Inhalation of food produces the clinical picture of a chronic, often recurrent pneumonia. The inflammatory infiltrate in the lung may progress to fibrosis. Lipoid pneumonia follows the aspiration of liquid paraffin laxatives, oily nosedrops or oily seawater. The lung pathology shows a foreign body giant cell reaction and fibrosis. Chronic pneumonias due to inhalation may mimic lung tumours and the diagnosis may only become evident following surgical resection. Repeated inhalation may eventually produce bilateral fibrosis and bronchiectasis. Hot smoke inhalation causes respiratory burns. There is oedema of the upper airways and bronchial tree, causing wheeze and croup. Severe injury is unlikely if there is no blistering of the mouth. Several hours or days after inhalation,

13

637

pulmonary oedema may develop. Early corticosteroid therapy may be beneficial. Patients may progress to develop the acute respiratory distress syndrome (p. 697).

Pneumonia in the immunocompromised host The increased use of immunosuppressive agents and the emergence of HIV infection have greatly increased the prevalence of pneumonia in the immunocompromised host - so-called 'opportunistic' infections. These infections are particularly common in certain well-recognized groups (Table 13.15). In patients who are immunosuppressed, particularly following organ transplantation or cytotoxic therapy for malignancy, the development of fever and pulmonary infiltrates is not always due to infection (Table 13.16). Occasionally bleeding into the lung occurs in patients who are thrombocytopenic. Radiation pneumonitis is most intense between 4 and 6 weeks after treatment. In a small number of patients a non-specific pneumonitis develops for which no cause is found, even when the patient is submitted to open lung biopsy. Infection can be due to a wide variety of agents. Immunodeficiency facilitates infection with organisms seldom encountered in immunocompetent individuals and the clinical picture is often of a rapid, extensive and life-threatening pneumonia (Table 13.17). 1 TABLE 13.15 Groups at risk of opportunistic pneumonias Primary immunodeficiency B-cell deficiency (agammaglobulinaemia) T-cell deficiency (Di George syndrome) T-cell and B-cell deficiency (combined immunodeficiency)

Secondary immunodeficiency HIV infection (AIDS) Leukaemias and lymphomas Cytotoxic agents (particularly following organ transplantation and the treatment of haematological malignancies) Corticosteroid therapy Malnutrition, general debility, uraemia, liver failure etc.

TABLE 13.16 Fever and pulmonary infiltrates in immunocompromised patients Malignant infiltration (e.g. leukaemia, Kaposi's sarcoma) Pulmonary emboli Non-specific pneumonitis

Infection Pulmonary oedema Pulmonary haemorrhage Drug-induced pneumonitis Radiation pneumonitis

1

638

MCQ13.7

2

Fig. 13.20

TABLE 13.17 Important causes of pneumonia in the immunocompromised patient Bacterial Gram-negative organisms as well as Gram-positive Nocardia Mycobacteria Legionella Mycoplasma Ch/amydia

Viral Cytomegalovirus Herpes simplex Fungal Candida Aspergillus Cryptococcus Pneumocystis

Management of pneumonia in the immunocompromised patient Pneumonia in the immunocompromised host requires immediate investigation and treatment. All patients should have blood cultures and appropriate serological investigations for likely pathogens. The chest X-ray is of limited value, with the exception of pneumocystis pneumonia in patients with HIV infection. Any pleural fluid should be sampled. Localized shadows are more common in bacterial infection; diffuse infiltration suggests an opportunistic organism. In all patients the likely value (and the risks) of invasive techniques to establish a definite cause for pneumonia must be considered. Most patients with diffuse pulmonary infiltration do not have sputum, and invasive investigations are therefore required to demonstrate the infective agent. Transtracheal aspiration, percutaneous needle aspiration biopsy, sputum induction with nebulized hypertonic saline, bronchoscopy and open lung biopsy can all yield useful information. Bronchoscopy is the most widely available technique. Bronchoalveolar lavage is a valuable technique for diagnosing pneumocystis and cytomegalovirus infection and is safe in patients with thrombocytopenia and bleeding diatheses. Transbronchial biopsy, although effective, can produce dangerous haemorrhage in patients with bleeding problems (uraemic patients are particularly prone to haemorrhage) and pneumothorax is a complication, particularly hazardous in patients receiving positive-pressure ventilation. Specific pneumonias in immunocompromised patients Gram-negative pneumonias Bacterial pneumonias, particularly Gram-negative infections, are common in neutropenic patients (e.g. following bone marrow transplantation, or chemotherapy for leukaemia), particularly if the granulocyte count is less than 500 x 106/L, and may be associated with lifethreatening septicaemia. The chest X-ray usually shows localized shadowing, rather than diffuse infiltration. Following blood cultures empirical therapy should be started. Common organisms are Pseudomonas aeruginosa, Escherichia coli, Klebsiella, Enterobacter and Serratia spp.

The use of broad-spectrum antibiotics favours the emergence of resistant organisms and also the development of fungal infections. Treatment of Gram-negative infections commonly requires aminoglycosides, cephalosporins and antipseudomonal pencillins, for example either ceftazidime or piperacillin-tazobactam with gentamicin. Ciprofloxacin or a carbapenem (imipenem or meropenem) should be reserved for the treatment of resistant strains. Klebsiella pneumonia The most common clinical setting for this Gram-negative pneumonia is the elderly male patient, often with chronic lung disease, whose 'immunosuppression' is due to general debility, associated with chronic illnesses such as diabetes or alcoholism. The illness can be severe (mortality 20-50%), with high fever, rigors and pleuritic pain, and haemoptysis occurs more often than in most bacterial pneumonias. The upper lobes are commonly involved, with considerable necrosis and cavitation, often bilateral, and with bulging of the fissure adjacent to the consolidated lung. There may be diagnostic confusion with tuberculosis. Patients with klebsiella pneumonia can develop empyema. Treatment is most effective with gentamicin plus cefuroxime, but resolution is often slow and substantial residual pulmonary damage is not uncommon. Pneumocystis carinii pneumonia (PCP) Pneumocystis pneumonia occurs in patients receiving steroids and other immunosuppressive agents, and is particularly common in AIDS. The clinical picture is of fever, a dry cough and progressive breathlessness. On auscultation the lungs often sound remarkably normal. The chest X-ray is often clear when symptoms first develop, but there

FIG. 13.20 Pneumocystis carinii pneumonia (PCP) Note that the diffuse hazy consolidation is bilateral and maximal in the mid and lower zones. 2

is progressive bilateral pulmonary infiltration, commencing in the perihilar regions, without effusions or adenopathy (Fig. 13.20). PCP is often marked by profound hypoxaemia which is worse than might be suggested from the physical signs on chest X-ray. Untreated, all patients become incapacitated by dyspnoea and die of hypoxaemia. In patients with AIDS, pneumocystis pneumonia characteristically evolves over several weeks, whereas in other patients at risk the time course is more rapid. In AIDS, pneumocystis is the most common cause of diffuse pulmonary shadowing. Because pneumocystis pneumonia is so common in this group it is appropriate in some circumstances to initiate empirical therapy. Whenever there is doubt about the correct diagnosis, or if response to treatment for presumed pneumocystis infection is slow, a definite diagnosis is required; for this, bronchoscopy with BAL and, if lavage is negative, TBB, has a very high diagnostic yield. The diagnosis can also be achieved by inducing sputum production with nebulized hypertonic saline. When treated early, pneumocystis pneumonia responds well to therapy with high-dose co-trimoxazole (20 mg of trimethoprim and 100 mg of sulfamethoxazole per kg body weight per day) given for 3 weeks. Mortality has fallen from 30% to 5% in recent years. Trimethoprim and dapsone is an effective combination, as is clindamycin plus primaquine, whereas pentamidine, although effective, is more toxic. Progressive hypoxia may require CPAP therapy (p. 670), and in selected patients intubation and mechanical ventilation is appropriate. Unfortunately, recurrence is common. Co-trimoxazole and inhaled pentamidine provide effective prophylaxis. Because PCP is unusual in HIV until the CD4 lymphocyte count is less than 250, it is usual to initiate primary prophylaxis at that level. In AIDS patients with severe pneumocystis pneumonia and respiratory hypoxia (Pao2 < 9.0 kPa) corticosteroids, given early, improve survival. Aspergillus pneumonia Aspergillus fumigatus is a rare cause of pneumonia and patients usually have severe granulocytopenia or are on corticosteroid therapy, and have frequently received broad-spectrum antibiotics. Patients receiving therapy for leukaemia are particularly at risk. The clinical picture is of fever, dyspnoea, pulmonary infiltrates that resemble pulmonary infarction, sometimes with cavitation, and pleural involvement. There is frequently evidence of aspergillus infection at other sites, notably the brain, bones and endocardium. Diagnosis is most reliably achieved by open lung biopsy or transbronchial biopsy. Bronchoalveolar lavage in immunosuppressed patients frequently demonstrates aspergillus, but biopsy material is required to confirm invasive pulmonary infection. Treatment is with intravenous amphotericin, but mortality is high. High-dose lipid formulations of amphotericin B are commonly used in aspergillus pneumonia to reduce toxicity, but they are very expensive. Newer agents such as itrazonazole and voriconazole have activity against aspergillus, but their role in treating

EH

639

invasive aspergillosis, as an alternative to amphotericin, is not established. Some reports suggest that G-CSF in addition to amphotericin may improve outcome, but this has yet to be demonstrated unequivocally. Candida albicans is an unusual cause of pneumonia, occurring in the same high-risk group as those who develop aspergillus pneumonia and presenting a similar clinical picture. Blood cultures are frequently positive. Treatment is with amphotericin and flucytosine. Fluconazole may be an alternative. Nocardia pneumonia Nocardia asteroides can rarely cause a chronic suppurative pneumonia in otherwise normal individuals, but more commonly it is a cause of an acute pneumonia in the immunocompromised host. In addition to lung involvement, in which there may be a single lesion or extensive consolidation with cavitation, pleural disease, empyema and metastatic spread also occur. A CT head scan is an important investigation. The sputum shows Gram-positive hyphae which are acid fast and grow rapidly on aerobic culture. Co-trimoxazole (thrice normal dose) is effective; if there is no improvement in 4-5 days treatment can be changed to ciprofloxacin plus a cephalosporin such as cefotoxime or ceftriaxone. Other agents that have been successful include amitracin, imipenem and meropenem. Cytomegalovirus pneumonia CMV pneumonia is an important problem in immunosuppressed patients, particularly following transplantation (see p. 288).

Unusual pneumonias A number of unusual infections (bacterial, viral, fungal and protozoal) can cause pneumonia and are particularly important in certain parts of the world (Table 13.18). They are described in Chapter 9.

TUBERCULOSIS

TABLE 13.18 Unusual causes of pneumonia Bacterial Salmonella typhi and paratyphi (typhoid) Brucella abortus and melitensis (brucellosis) Pasteurella pestis (plague) Pasteurella tularensis (tularaemia) Bacillus anthracis (anthrax) Leptospira ictemhaemorrhagicae (leptospirosis) Rickettsial Typhus

Viral Varicella (chickenpox) Herpes zoster Epstein-Barr virus (infectious mononucleosis) Protozoal, yeast and fungal Actinomyces israelii (actinomycosis) Coccidioides immitis (coccidioidomycosis) Histoplasma capsulatum (histoplasmosis)

can infect anyone, but there are well-recognized 'at-risk' population groups in the UK, and these include the elderly, migrants and refugees from countries with a high incidence of the disease, as occurs in Africa and the Indian subcontinent, those in poor social circumstances, alcoholics, immunosuppressed patients and hospital employees. Tuberculosis is a common complication of HIV infection, and as a consequence the incidence has increased substantially in some areas of the world. Mycobacterium bovis was once a common cause of human infection but has now been almost eradicated in developed countries by pasteurization of milk and skin testing of cattle, with tuberculin-positive animals being slaughtered. Mycobacteria other than M. tuberculosis or M. bovis may cause infection in humans, for example M. kansasii, M. xenopi, M. avium and M. intracellulare. The source of infection, unlike the droplet route of M. tuberculosis, is generally thought to be exposure of susceptible individuals to organisms in water, soil, dust and animals (particularly pigs, chickens, birds and monkeys). The organisms are not transmitted between humans.

1

Infactivity, immunology and pathology Epidemiology In 19th-century England one person in five died of Mycobacterium tuberculosis infection. The dramatic decline in notifications and deaths reflects one of the great medical advances of that century. However, tuberculosis remains a common cause of death in the developing world. Eight million new clinical cases occur each year, leading to 2 million deaths. One-third of the world's population is infected with the bacterium that causes tuberculosis. TB

640

1

MCQ13.8 2 Fig. 13.21

3

Figs 13.22,13.23

4

5

Figs 13.26-13.28

6

Figs 13.24-13.25 Fig.13-29

The response to a primary infection is shown in Figure 13.21. The usual result of this is a small subpleural lesion which heals with fibrosis; the glandular lesion also heals, and within about 18 months both may calcify. However, some lesions, especially in children, may erode a blood vessel (Fig. 13.22) and progress to 'miliary' tuberculosis, so named because of the 'millet'-sized lesions on the chest X-ray. Alternatively the lung lesion may progress, and cavitate or involve the pleura, causing a pleural effusion. Finally the lymph node involvement may progress, especially in Asians and Black Africans, and occasionally compress, obstruct or erode a bronchus, particularly that of the right middle lobe. Of those that heal, some subsequently develop postprimary tuberculosis (Fig. 13.21), often many years later,

13

FIG. 13.21 Primary and postprimary (reinfection) tuberculosis In primary tuberculosis there is pneumonic consolidation (Ghon focus) which is usually subpleural and affects the mid or lower zones. [A] Subsequently, infection involves the draining hilar nodes [B] which may cause bronchial compression. Reinfection (post-primary) tuberculosis [C] is usually apical and progresses to cavitation, fibrosis, volume loss and pleural thickening [D].

sometimes referred to as reinfection adult tuberculosis. Whether the patients become reinfected exogenously or whether their existing lesions become active again is often not clear. Postprimary tuberculosis is usually apical, often bilateral, and can lead to cavitation, fibrosis, volume loss, bronchiectasis, and pleural thickening. Patients can develop a devastating tuberculous bronchopneumonia or miliary TB. In many countries children are immunized against TB using BCG vaccination (see below). Immunization confers only partial protection. Tuberculous infection in the immunized individual leads to postprimary disease.

Clinical features Some patients with tuberculosis are asymptomatic, their disease being noted on routine chest X-ray. Many patients have cough, often with sputum, haemoptysis, fever and weight loss. Tuberculosis may present insidiously, particularly in the elderly.

Investigation Tuberculin testing, in which protein derived from tuberculous bacilli is introduced into the skin, with either the Mantoux technique or the multiple puncture Heaf test, is useful in supporting the diagnosis, and for assessing contacts and children before BCG vaccination. A positive

FIG. 13.22 Miliary tuberculosis In miliary tuberculosis multiple organs are involved. The diagnosis is most commonly suggested by the chest X-ray. 4 Meningitis is relatively common.

Mantoux test should show 5 mm of induration using up to 10 tuberculin units. An alternative is the Heaf test. When performed correctly, a negative tuberculin test makes active tuberculosis infection relatively unlikely, except in patients who are profoundly immunosuppressed (e.g. those with advanced HIV disease). A normal chest X-ray is rare in active pulmonary tuberculosis. Characteristic appearances include: • Primary tuberculosis with hilar gland enlargement • Miliary tuberculosis (Fig. 13.23) 5 • Postprimary (reinfection) tuberculosis with cavitation (Fig. 13.24) 6 • Tuberculous bronchopneumonia. Fluorescent microscopic techniques and/or staining with the Ziehl-Neelsen stain are usually used to screen sputum to confirm that bacilli are present. Routine culture of the organism takes 2-8 weeks, and is important for excluding atypical mycobacteria and determining drug sensitivities. If the patient has no sputum, fibreoptic bronchoscopy with bronchial lavage is a useful diagnostic tool. In adults gastric washings or laryngeal swabs are seldom necessary when bronchoscopy is available. The sedimentation rate is usually raised, and in advanced disease there may be an anaemia with a raised white cell count, occasionally with a lymphocytosis. Liver function tests may be abnormal. Extrapulmonary tuberculosis can involve most organ systems (Table 13.19).

641

CASE STUDY 13.4 COUGH AND WEIGHT LOSS IN A 39-YEAR-OLD WOMAN FROM AFRICA A 39-year-old female nurse presented to outpatients with reduced appetite, weight loss and cough of 4 months' duration. She had lost 3 stone in weight. The cough was productive of greenish sputum flecked with blood. Born in Tanzania, the patient had moved to the UK 18 months previously. She smoked 20 cigarettes daily. Preliminary investigations arranged by her family doctor had demonstrated an abnormal chest X-ray, a blood sugar of 22 mmol/L, albumin 34g/L, Hb 10.7 g/dL and WBC 9.4 (normal differential). Clinical examination showed a temperature of 37.8°C, respiratory rate 22 and reduced chest wall movements over the left upper chest, with bronchial breathing on auscultation. The chest X-ray showed extensive consolidation in the left mid and upper zones, with a large cavity. The lateral chest X-ray confirmed that the abnormalities were largely confined to the left upper lobe (Case Fig. 13.4.1).

Questions 1. What is the likely diagnoses? 2. What is the explanation of the weight loss? 3. What would be your first investigation?

Discussion The most likely diagnosis is pulmonary tuberculosis. The ethnic origin of the patient, the duration of symptoms, fever and X-ray appearances all favour this diagnosis. It is slightly unusual to have such extensive disease in the left upper lobe with no abnormalities at all in the right lung. The patient also has diabetes, probably exacerbated by her illness. Further questioning revealed that the patient had had transient gestational diabetes 15 years previously. It is likely that her

642

CASE FIG. 13.4.1 [A] The PA chest X-ray shows extensive consolidation in the left lung, with a large cavity. [B] The lateral chest X-ray confirms that the disease is largely confined to the left upper lobe; note the large retrosternal cavity.

diabetes was contributing to the weight loss. Although tuberculosis is an important cause of weight loss, when this happens the patient is usually also hypoalbuminaemic. HIV requires consideration, although the white cell count is normal, without a lymphopenia. A key investigation is immediate examination of the patient's sputum, and this demonstrated numerous acid-fast bacilli.

Question 4. What would be your initial management? The patient has 'open' pulmonary tuberculosis and is infectious. Clearly, she should stop working as a nurse and the infection control team needs to be informed: TB is a notifiable disease and careful contact tracing will be needed.

The patient's TB requires treatment, as does her diabetes, and in view of the complexity and severity of her illness admission to hospital is appropriate. Drug resistance is an increasing problem in the therapy of tuberculosis, and resistance is relatively common in patients from some parts of Africa. Therefore, therapy is best started with four drugs, usually rifampicin, isoniazid, pyrazinamide and ethambutol. The patient was started on these four drugs and her diabetes was controlled with insulin. After appropriate counselling she was tested for HIV, and this proved negative. The patient responded rapidly to therapy. After a few weeks her diabetes was well controlled with oral hypoglycaemic agents. After 4 weeks of therapy the patient, who had been feeling much improved up to that point, developed nausea, tiredness and abdominal discomfort. She contacted the hospital and was immediately reviewed. On examination she was jaundiced and the liver was enlarged. Liver function tests showed a bilirubin of 84mmol/L, y-GT 564 U/L, aspartate aminotransferase 840 U/L and alkaline phosphatase 326 U/L.

incriminated. All drugs should be stopped and liver function monitored. This usually leads to improvement, but not always, and occasional patients develop established liver failure; a number of cases have been successfully managed with liver transplantation. When liver function tests have returned to normal, it is best to reintroduce therapy one drug at a time. The first-line drugs used in the treatment of tuberculosis are the best available, and it is desirable that as many of them as possible be used. Because rifampicin is the most likely to cause hepatitis, many physicians introduce either pyrazinamide or isoniazid first. If patients tolerate one drug for a week, the second drug is added, and subsequently an alternative third or fourth drug. If drug-induced problems are less severe than hepatitis (for example

skin rashes), and the treatment is crucial (for example in patients with extensive tuberculosis or TB meningitis), it may be possible to continue therapy with the addition of corticosteroids. In this patient the liver function tests improved off treatment, and therapy was continued with isoniazid, pyrazinamide and ethambutol. Sputum culture confirmed the diagnosis of mycobacterium tuberculosis. Because of the unconventional drug regimen treatment was for 9 months rather than 6.

13

Question 6. With therapy, what is the likely outcome for this patient in terms of lung function and chest radiology?

Question 5. What should be done?

Patients with TB commonly have minor abnormalities of liver function before starting therapy. They also often develop modest elevation of liver enzymes on rifampicin treatment. However, the development of jaundice and the markedly abnormal liver function tests indicate drug-induced liver disease. Rifampicin, isoniazid and pyrazinamide are all hepatotoxic, and rifampicin is the most commonly

CASE FIG. 13.4.2 Chest X-ray after 8 months' treatment. Cavitation is no longer visible and the consolidation has resolved, leaving some linear fibrotic change in the left upper zone.

643

CASE STUDY 13.4 CONTINUED If treatment for tuberculosis is started early, patients usually do remarkably well. Consolidation clears completely and, perhaps surprisingly, cavities also disappear. Patients do less well if there is established fibrosis. Such fibrosis usually causes volume loss, and it is noteworthy that in this patient there was no evidence

of volume loss on her initial X-rays (Case Fig. 13.4.1). In this patient, after 6 months' therapy the large cavity had resolved and the extensive consolidation had largely disappeared, leaving relatively minor linear shadowing in the left upper zone (Case Fig. 13.4.2). The patient became symptom free and was not

FIG. 13.23 Miliary tuberculosis Thousands of small nodules, 'millet-size', all of similar size and distributed throughout the lungs.

Differential diagnosis In any febrile illness it may be relevant to consider tuberculosis, particularly if there is chest X-ray shadowing or lymph gland enlargement. It must be distinguished from simple pneumonia, and in such cases the usual clinical feature that raises suspicion of tuberculosis is the indolent nature of the illness.

Complications These may be due to dissemination of the disease, extensive lung damage or non-specific metabolic disturbances.

1

644

Case 13.1

breathless. The FEVi and vital capacity were within the normal range and gas transfer was approximately 80% of normal. The Po2 of 10.4 and Pco2 of 4.7 demonstrated minor abnormalities of gas exchange. As a result of her illness, the patient managed to stop her cigarette smoking.

FIG. 13.24 Postprimary (reinfection) tuberculosis The widespread tuberculous consolidation involves predominantly the upper and mid zones. The translucent areas in the upper zones are cavities. There is calcification in the right upper zone of the right hemidiaphragm, suggesting that the patient has reactivation of old disease.

TABLE 13.19 Diagnosis of extrapulmonary tuberculosis' Three early morning urine specimens Liver biopsy Lymph node biopsy Bone marrow aspiration and trephine biopsy Lumbar puncture Laparoscopy and biopsy Joint aspiration Bone biopsy Stool specimens in immunosuppressed patients * All specimens should be cultured.

• Dissemination (extrapulmonary tuberculosis). Tuberculosis may involve lymph nodes, pleura, bones, joints, kidneys, adrenals, liver, peritoneum, bowel, epididymis, meninges and brain. 1 • Extensive lung damage. This may result in massive haemoptysis, bronchiectasis, breathlessness, and eventually respiratory failure with cor pulmonale. • Associated metabolic conditions. These can include hypokalaemia, hyponatraemia, hypercalcaemia and hypoalbuminaemia.

Identification of the organism is best achieved by culture from blood, bone marrow, liver and bronchi. In patients with advanced HIV disease the mycobacterial infection is usually disseminated. The organism is resistant to pyrazinamide and isoniazid, and treatment is difficult. Recommended drugs include ethambutol with rifampicin or rifabutin, quinolones, the newer macrolides, and amikacin. An in vivo response often occurs despite in vitro resistance, and is due to the synergistic action between combinations of drugs. However, prognosis is poor.

Atypical mycobacterial infections

Management of M. tuberculosis

Occasionally mycobacteria other than M. tuberculosis or M. bovis cause pulmonary disease. The most common of these 'atypical' mycobacterial pathogens are M. kansasii, M. xenopi, M. avium intracellulare (MAI) and M. malmoense. These are environmental organisms, found particularly in water and soil. In patients with HIV disease MAI is relatively common. In other patients M. kansasii infection is most frequent in the UK. Patients with atypical mycobacterial infection do not pose a significant infectious risk to their close contacts or carers. Except for patients with HIV disease, atypical mycobacterial infection occurs predominantly in middle-aged or elderly males. Often there is pre-existing lung disease. The onset of symptoms is gradual, most commonly with cough, sputum, haemoptysis and weight loss. The chest X-ray usually mimics classic postprimary M. tuberculosis infection. Treatment of these organisms is indicated only when it is clear that they are causing significant disease and is made on an empirical basis. M. kansasii is usually sensitive to rifampicin and ethambutol. M. xenopi is usually treated with rifampicin, isoniazid and ethambutol for 18-24 months; however, 9 months' therapy will eradicate the disease in 90% of cases.

Organisms in the walls of pulmonary cavities, exposed to a high oxygen tension, tend to replicate rapidly, whereas those in caseous areas or within macrophages, in which the pH is low, divide relatively slowly. There are therefore three populations of organisms to be treated:

Mycobacterial infection in HIV/AIDS Patients with the acquired immunodeficiency syndrome (AIDS) are frequently infected with mycobacteria, either typical M. tuberculosis or atypical mycobacteria, commonly Mycobacterium avium intracellulare. Infection may result in pulmonary or extrapulmonary disease, and unusual clinical presentations may occur. Although M. tuberculosis infection is usually a reactivation of previous tuberculosis, typical upper lobe disease is not so common in patients with AIDS. M. tuberculosis in AIDS is usually responsive to conventional antituberculous treatment. Prevention of tuberculosis is possible with isoniazid prophylaxis, and this should be considered in HIV-positive patients who also have a positive tuberculin skin test.

13

• The rapidly dividing extracellular organisms • The slowly dividing extracellular organisms • The slowly dividing intracellular organisms (persisters). The immediate target for chemotherapy is to kill the large numbers of organisms in the 'rapidly dividing' group and to prevent the survival of 'mutant' strains. The slowgrowing intracellular and extracellular organisms will be completely eliminated only by a prolonged course of effective drugs, and if there are 'persisters' from these two groups they will cause late relapse of the disease.

Antituberculous drugs and treatment regimens The four most commonly used drugs in developed countries are listed in Table 13.20. With the use of isoniazid and rifampicin, both of which have activity against intracellular organisms, tuberculous lesions should be sterilized quickly. Effective chemotherapy of pulmonary tuberculosis is 6 months' isoniazid, and rifampicin, plus pyrazinamide and ethambutol given for the first 2 months. Ethambutol can be omitted in patients with a low risk of resistance to isoniazid.

'First-line' drugs for tuberculosis Drug

Daily dose

Side-effects

Rifampicin

10 mg/kg, usually 450-600 mg

Hepatitis, nausea, vomiting, fevers, thrombocytopenia, renal failure (on intermittent therapy), drug interactions (e.g. oestrogens)

Isoniazid (prescribed with pyridoxine 10 mg daily)

Children 6 mg/kg Adults usually 300 mg

Peripheral neuropathy, hepatotoxicity, hypersensitivity reactions

Mycobacterium avium intracellulare infection in AIDS

Pyrazinamide

20-30 mg/kg usually 1-2 g

Hepatotoxicity, hyperuricaemia, arthralgia

Infection with Mycobacterium avium intracellulare is the commonest atypical mycobacterial complication of AIDS.

Ethambutol

15-25 mg/kg

Optic neuritis, rashes

645

In patients with minor liver disease standard therapy can be given, but liver function should be closely monitored. In severe liver disease, second-line non-hepatotoxic drugs should be used (e.g. ciprofloxacin, streptomycin). In renal failure treatment of tuberculosis is with rifampicin, isoniazid and pyrazinamide, and if ethambutol or streptomycin is used the dosage should be reduced and drug levels monitored. Dialysis can greatly affect drug clearance. Rifampicin, isoniazid, pyrazinamide and ethambutol are safe in pregnancy, but streptomycin should be avoided. Rifampicin reduces the effectiveness of the combined oral contraceptive as well as other drugs, and approximately halves the pharmacological effect of concurrent oral prednisolone. Rifampicin makes urine a red-orange colour. Drug resistance Drug resistance is due to infection with an organism from an ineffectively treated patient (primary or initial resistance) or occurs during therapy (secondary or acquited resistance). Resistance to isoniazid is most common. Resistance to rifampicin and pyrazinamide is unusual, but in the Far East and Africa initial resistance to isoniazid or streptomycin, or both, is high. Multiple-drugresistant (MDR) TB is an increasingly important problem, usually in partially treated patients, for example from Africa and Russia. Such patients require isolation facilities for their treatment, and pose an important public health problem, not least for their medical carers. Immunosuppressed patients are at particular risk. In patients with resistant strains the use of 'second-line' drugs (e.g. streptomycin, capreomycin, cycloserine, quinolones, prothionamide), particularly in the first 2^4 months of treatment, may be necessary. Four or even five drugs may be required in this initial period, and to achieve cure a more prolonged, 18-24-month, regimen is necessary. Drug resistance can be determined using bacteriological or molecular tests on appropriate specimens. Drug resistance, for example to rifampicin, can be rapidly established using molecular techniques. Drug side-effects Short-course regimens are relatively non-toxic (Table 13.19), although rarely standard drug therapy can cause fulminant liver failure and liver function should be tested before starting therapy. If adverse reactions occur (most commonly skin rashes) all the drugs should be stopped and subsequently, when the reaction has subsided, one drug reintroduced. If all goes well, a second drug can then be started. Cost Rifampicin is an expensive drug and pyrazinamide is also costly. Cost limits their use in developing countries. Overall, however, short course chemotherapy remains one of the most cost-effective interventions worldwide. 646

Patient non-compliance with treatment The treatment of tuberculosis requires disciplined drug-

taking, usually on a daily basis, for 6-9 months in most drug regimens. This may be difficult to achieve, particularly in the rural populations of developing countries and in the vagrant populations of the inner cities of western countries. For this reason supervised twice-weekly regimens, ultrashort courses which accept less than 100% cure, or even supervision in a residential hostel, may have to be considered. An effective three-times-weekly treatment regimen is rifampicin, isoniazid and pyrazinamide, with either streptomycin or ethambutol, for 2 months, followed by rifampicin and isoniazid for 4 months. Supervision can include direct observations of pill-taking (directly observed therapy-DOT), urine tests for isoniazid, and urine inspection for the red discoloration of rifampicin. Combination tablets of rifampicin and isoniazid, with or without pyrazinamide, aid compliance. An advantage of combination tablets is that red urine or a positive urine test for isoniazid confirms that all drugs are being taken. Treatment of extrapulmonary tuberculosis Treatment of renal, bone and joint, lymph node, liver, pericardial, peritoneal and bowel tuberculosis is in general similar to that of pulmonary tuberculosis. The treatment of tuberculous meningitis deserves special mention, as treatment failure may be lethal. Treatment should include isoniazid, pyrazinamide and rifampicin, with a fourth drug such as streptomycin or ethambutol. Intrathecal drugs are rarely given. Treatment should be given for 12 months. Similarly, patients with miliary tuberculosis should be treated for 1 year. The use of corticosteroids in tuberculosis The administration of corticosteroids, in addition to antituberculous drugs, may sometimes be beneficial. Their use should be considered in the very wasted patient, in tuberculous meningitis and intracerebral tuberculomata, as well as in pleural and pericardial effusions. Where a patient develops hypersensitivity to a drug that essential and irreplaceable in a treatment regimen, steroids may allow therapy to continue.

Prevention The dramatic decline in tuberculosis in developed countries is due to several factors. Of great importance is the control of infection from human sources by effective chemotherapy. Isolation of infectious patients (those with smear-positive sputum) for 2 weeks is helpful. Contact tracing is crucial and immediate contacts, especially children, should have a tuberculin test and chest X-ray. Host defences are improved by better housing and social conditions and the use of Bacille Calmette-Guerin (BCG) vaccination. This vaccine, using an attenuated strain of tubercle bacillus, gives a harmless primary infection and therefore a large measure of immunity. It is given to teenage children in many developed countries and to infants elsewhere. To avoid the development of tuberculous disease, chemoprophylaxis (preventative therapy) is

commonly recommended for individuals with a strongly positive tuberculin skin test but without other evidence of infection who have not received BCG. Many are immigrants from parts of the world where tuberculosis is very common. Because of the increased risk of drug side-effects in older individuals, chemoprophylaxis is usually limited to children and young adults. A suitable drug regimen is isoniazid plus rifampicin in standard doses, for 3 months, or isoniazid alone for 6 months. In many developing countries the incidence of TB is high and rising because of HIV infection, and this poses a formidable public health problem.

FURTHER READING ON RESPIRATORY INFECTIONS American Thoracic Society 1994 Treatment of tuberculosis and tuberculosis infection in adults and children. Am J Respir Crit Care Med 149:1359-1374 Baughman R P 1994 Use of bronchoscopy in the diagnosis of infection in the immunocompromised host. Thorax 49:3-7 British Thoracic Society 2001 Guidelines for the management of community-acquired pneumonia in adults. Thorax 56 (suppl. iv) British Thoracic Society 2001 Management of opportunist mycobacterial infections. Thorax 55: 210-218 Fein A M, Niederman M S 1995 Guidelines for the initial management of community-acquired pneumonia: savory recipe or cookbook for disaster. Am J Respir Crit Care Med 152:1149-1153 Joint Tuberculosis Committee of the British Thoracic Society 1998 Chemotherapy and management of tuberculosis in the United Kingdom: recommendations 1998. Thorax 53:536-548 Joint Tuberculosis Committee of the British Thoracic Society 2000 Control and prevention of tuberculosis in the United Kingdom. Code of practice 2000. Thorax 55:887-901 Nunn P P, Elliott A M, McAdam K P W J 1994 Impact of human immunodeficiency virus on tuberculosis in developing countries. Thorax 49:511-518 Ormerod L P, Skinner C, Wales J 1996 On behalf of the Joint Tuberculosis Committee of the British Thoracic Society. Hepatotoxicity of antituberculous drugs. Thorax 51:111-113

LUNG ABSCESS

13

A lung abscess is an infected, cavitating lesion within the lung parenchyma. Abscesses can occur as a result of primary infection, or secondary to other conditions (Table 13.21). The two factors critical to the pathogenesis of lung abscesses are necrotic tissue and infection. Apart from septic emboli, infection is via the tracheobronchial tree and seeds in an area of necrotic lung tissue. This leads to further breakdown of lung tissue and communication with a bronchus. The presentation will vary with the cause: patients with acute lung abscesses secondary to bacterial pneumonia (e.g. Klebsieila) are acutely unwell, febrile and coughing foul, bloodstained sputum, whereas the presentation of a tuberculous abscess will be much more indolent. The patient's breath and sputum may smell faeculent in the presence of anaerobes. There may be finger clubbing. Lung abscesses are more common in alcoholics and elderly debilitated patients, who are prone to aspiration, and in intravenous drug abusers. Lung abscess may follow the aspiration of infected material in patients with dental sepsis and sinusitis, and may complicate bronchiectasis. The chest X-ray initially shows a homogeneous opacity, and only when bronchial drainage is established does an air-fluid level appear (Fig. 13.25). Sputum should be sent for Gram stain, culture and sensitivity and examined for acid-fast bacilli. Fibreoptic bronchoscopy will provide useful information, both for bacteriology (especially the

TABLE 13.21 Causes of lung abscess Necrotizing infection Pyogenic bacteria (Staphy/ococcus aureus, Klebsieila, anaerobes, Pseudomonas aeruginosa) Mycobacteria (M. tuberculosis or atypical mycobacteria) Fungi (Histoplasma capsulatum, Coccidioides immitis) Parasites (Entamoeba histolytica, Paragon ium wester mani) Secondary to bronchial occlusion Bronchial carcinoma or adenoma Foreign body aspiration Bronchial stenosis

Pulmonary infarction Pulmonary thromboembolism Pulmonary foreign body embolism (e.g. i.v. drug abusers) Septic pulmonary embolism Cavitation in malignant tumour Bronchogenic carcinoma (esp. squamous cell) Metastases Lymphoma Others Wegener's granulomatosis Infected cysts, bullae, sequestrations Pneumoconiosis (silica, coal)

FIG. 13.25 Cavitating lung abscess The abscess is in the lingular segment of the left upper lobe. The patient had a cough productive of large volumes of purulent sputum from the abscess cavity; note the fluid level.

647

isolation of anaerobes) and in identifying neoplasms or foreign bodies. Patients with staphylococcal or Klebsiella pneumonias tend to develop abscesses. The commonest radiological appearances with staphylococcus are of multiple, thinwalled cystic spaces with little fluid. Anaerobic abscesses due to Entamoeba histolytica usually involve the liver as well as the lung. The patient may have haemoptysis or cough up 'anchovy paste' or 'chocolate sauce' sputum. The treatment of lung abscess depends on the cause. Foreign bodies and some cavitating tumours can be surgically removed. When there is no surgically resectable lesion treatment centres on frequent physiotherapy and appropriate antibiotics, guided by the bacteriology of sputum, blood and bronchial washings, and given for 6 weeks or more. Needle aspiration to obtain samples of pus for microbiology may be helpful and percutaneous catheters positioned during radiological imaging can facilitate resolution. Surgical resection is occasionally necessary.

BRONCHIECTASIS

Bronchiectasis is chronic dilatation of bronchi, usually associated with deficient local clearance of secretions. The disease may be localized to a lobe or segment, or may be generalized throughout the bronchial tree. Clinically, bronchiectasis presents with intermittent or constant cough, and sputum with or without haemoptysis.

Aetiology Congenital factors • Cystic fibrosis (p. 649). This multisystem disorder is associated with bronchial obstruction due to inspissated mucus and thus recurrent infection, the consequence of which is bronchiectasis. • Immune deficiency syndromes. Bronchiectasis is particularly frequent in hypogammaglobulinaemia. • Kartagener's syndrome. This is the triad of dextrocardia, sinusitis and bronchiectasis. The sinusitis and bronchiectasis are due to disorders of ciliary function. The Chandra-Khetarpal syndrome is similar, but with laevocardia and no demonstrable ciliary abnormality. • Pulmonary sequestration. Bronchiectasis frequently occurs in congenitally sequestrated lung tissue. Inflammatory factors A variety of infective processes in the lung may produce bronchiectasis. These include infection with necrotizing

1

648

MCQ13.9

organisms such as Klebsiella pneumoniae, staphylococci, Pseudomonas aeruginosa and tuberculosis. The damage caused to the bronchial wall by these organisms may lead to dilatation and permanent ciliary dysfunction. Some viral infections, particularly adenovirus and influenza virus, may predispose to bronchiectasis, as may an episode of bronchiolitis or whooping cough in childhood. MacLeod's syndrome is a unilateral emphysema-like condition which follows early childhood infections and may be associated with bronchiectasis. The other major inflammatory process predisposing to bronchiectasis is immunological. This may be important in bronchiectasis of any aetiology, but the classic example is allergic bronchopulmonary aspergillosis (p. 660). It is likely that antigen/antibody complexes in the bronchial wall fix complement, causing an inflammatory reaction, tissue damage and bronchiectasis, characteristically in proximal large airways. Bronchial obstruction and retraction Proximal obstruction of an airway by foreign body, tumour, enlarged lymph gland or mucus plug causes secretions to accumulate distally; these then become infected, causing bronchial wall damage. The result of this process is often an area of bronchiectasis localized to a segment or lobe, which may be amenable to surgical resection. Retraction of lung tissue can occur with healing of tuberculosis or with extensive pulmonary fibrosis, causing distortion and widening of adjacent bronchi. Idiopathic Probably the largest single group of patients are those in whom a cause is never found. They usually have bilateral lower lobe disease.

Clinical features The usual presentation is with cough and sputum. In the advanced stage of the disease patients may produce up to 40-50 mL of sputum per day, most of it in the morning. Haemoptysis is common, usually as blood-streaking of the sputum. Massive haemoptysis does occur, often in association with a recurrence of infection, and is occasionally fatal. With widespread disease the patient is breathless, with malaise, weight loss and recurrent fevers. The physical signs depend on the extent and stage of the disease. In the patient with severe widespread bronchiectasis general health may be poor, finger clubbing is common, there are coarse crackles and wheezes, and there may be central cyanosis and cor pulmonale. Complications of bronchiectasis include pneumonia, lung abscess, pleural effusion, empyema, brain abscess and amyloidosis. With effective antibiotic therapy these complications are unusual. 1

Investigation The chest X-ray may be normal. However, the more characteristic radiographic changes (Fig. 13.26) include ring,

tive/restrictive defect develops, with variable response to bronchodilators. Younger bronchiectatic patients require a sweat test to exclude cystic fibrosis, and all patients should have a routine estimation of their serum immunoglobulins. A blood eosinophilia is common in allergic bronchopulmonary aspergillosis. Ciliary function may require investigation, including a nasal mucosal biopsy with motility studies.

13

Management

FIG. 13.26 Extensive bilateral basal bronchiectasis

FIG. 13.27 Bronchiectasis: CT scan There is extensive disease, with bronchial wall thickening and airway dilatation, easily seen in the left lower lobe (arrow).

line and parallel ('tramline') shadows - the appearances of thickened bronchial walls seen end-on or laterally. Thinsection CT scanning is a sensitive technique for demonstrating dilated bronchi (Fig. 13.27). If the disease is largely peripheral it usually has a cystic appearance, whereas when the major bronchi are affected they take on a fusiform or saccular appearance. Bronchography is seldom performed. The most common infecting organism is H. influenzae, but pneumococcus is also a frequent pathogen. Colonization with Ps. aeruginosa is not uncommon. Lung function tests reflect the extent and severity of disease. As the condition worsens a mixed obstruc-

The long-term management of bronchiectasis includes regular, effective physiotherapy. Patients and their relatives are taught postural drainage. In the majority of patients continuous long-term antibiotics are not appropriate and predispose to colonization with Ps. aeruginosa. Prompt treatment of intercurrent and symptomatic infection with appropriate antibiotics is essential. Regular inhaled antibiotics may be helpful (e.g. gentamicin, piperacillin, colomycin), especially in patients with cystic fibrosis. Severe episodes of infection require hospitalization. Antibiotic therapy should be guided by microbiological information whenever possible. If no organism can be identified, broad-spectrum cover, including activity against Pseudomanas, is required; for example, long-term venous access may be necessary in patients with frequent infective episodes (usually those with cystic fibrosis) and many such patients can be managed in the community. Patients with agammaglobulinaemia require regular intravenous replacement therapy and can be kept remarkably free of infection. Inhaled |3 agonists are useful in those patients with airways obstruction. Oral steroids are required in bronchopulmonary aspergillosis and can be helpful in some other bronchiectatic patients in whom bronchial damage is immunologically mediated. In severe cases patients eventually develop cor pulmonale and require diuretic and domiciliary oxygen therapy. A proportion of patients with cystic fibrosis have been successfully treated by heart-lung transplantation. Surgical resection for bronchiectasis is indicated in only a small number of cases, usually for localized disease caused by previous bronchial occlusion or lobar pneumonia. It may also be necessary to resect bronchiectatic lung for massive haemoptysis. Haemoptysis can sometimes be successfully treated by embolization of the relevant branch of the bronchial artery with foam. With the exception of patients with generalized disorders, such as cystic fibrosis or immune deficiency, the long-term outlook for the majority of patients with bronchiectasis is good.

CYSTIC FIBROSIS Cystic fibrosis is the most common potentially lethal genetic disease in Caucasians. It is an autosomal recessive

649

gene which is carried by 1: 20-25 of the Caucasian population and its incidence in Britain is about 1:2000-2500 live births. It affects more than 50000 people world wide. There have been important advances in the understanding of the genetics and pathogenesis of cystic fibrosis over the past 20 years and life expectancy has been greatly improved by a holistic approach to therapy. Mucus in cystic fibrosis patients is no more viscid than in chronic bronchitis. However, mucus-secreting glands throughout the body show morphological changes and the ducts of the glands are obstructed. The organs affected include the gut, pancreas, liver and reproductive tract. However, the most severe and life-threatening expression of the condition is within the lungs, which show evidence of small airway inflammatory change in infancy and subsequent susceptibility to infection with organisms such as Staphylococcus aureus and Haemophilus influenzas. Eventually most patients are infected with Pseudomonas aeruginosa. The cystic fibrosis gene was identified in 1989. It is located on the long 'arm' of chromosome 7 and encodes for the cystic fibrosis transmembrane conductants regulator (CFTR) protein. This protein is a member of the adenosine triphosphate (ATP)-binding cassette (ABC) family, which has two functions characteristic of this family of molecules: 1. It is a cyclic AMP-regulated chloride channel in the apical membrane of epithelial cells; mutations in the gene result in CFTR being either dysfunctional or wrongly localized; this impairs chloride transport across epithelial cells in the affected organs. CFTR mutations are amiloride-sensitive epithelium sodium channels causing increased sodium absorption. The epithelial cells of the respiratory tract and sweat ducts are relatively impermeable to chloride ions with a secondary increase in sodium transport; chloride transport in other cells such as erythrocytes, are normal. The ion transport abnormality leads to impaired local defence mechanisms in the respiratory tract and abnormal pancreatic secretions. The defect in the lung causes bacterial colonization, damage to the bronchial wall and bronchiectasis. Defective pancreatic secretion leads to a malabsorption syndrome, and blockage of the ducts with autodigestion of the pancreas results in diabetes mellitus in 10-20% of the adult patient population. The molecular mechanisms implicated in the lung disease are still a matter of debate. The 'high salt/defence' hypothesis suggests that missing or defective CFTR results in reduced transepithelial chloride conductants, allowing airway surface liquid (ASL) salt levels to remain close to plasma levels. The high salt concentrations may interfere with their epithelial cell-derived natural antibiotics, such as defensins, lactoferrin or 1

650

MCQ13.10

2

MCQ 13.11

TABLE 13,22 Clinical consequences of cystic fibrosis Neonates Meconium ileus Rectal prolapse Failure to thrive Recurrent bronchopulmonary infections

Older children and young adults Bronchiectasis Malabsorption Meconium ileus equivalent Infertility Cirrhosis and portal hypertension

lysozymes, and there is consequent bacterial colonization of the airway. 2. The low-volume hypothesis postulates that ASL is salt depleted because the abnormal CFTR is unable to inhibit EnaC. Therefore, there is depletion of the ASL volume leading to reduced mucociliary clearance and retention of mucus on the airway surface, which predisposes to chronic infection.

Clinical features and diagnosis The clinical presentation differs between neonates and older children and young adults (Table 13.22). Clinical examination of adolescents and adults usually shows finger clubbing secondary to chronic respiratory infection, retarded growth and development, and a hyperinflated chest with coarse crackles and wheezes. Gynaecomastia is occasionally seen, and there may be hepatosplenomegaly and other signs of chronic liver disease. The chest X-ray shows hyperinflated lungs and widespread bronchiectasis (parallel line and ring shadows, with multiple nodular opacities, mainly in the mid and upper zones; the hila are usually prominent). Lung function testing shows a progressive obstructive defect. The single most important diagnostic test is the sweat sodium, which is consistently greater than 70mmol/L in children with the disease and greater than 90mmol/L in affected adults, although the results in adults are less reliable. Sputum bacteriology will often yield S. aureus, and later in life Ps. aeruginosa becomes a major problem. Haemophilus may be an important pathogen, and patients occasionally develop tuberculosis. Prenatal diagnosis can be achieved by the measurement of elevated levels of enzymes such as alkaline phosphatase in the amniotic fluid or by DNA analysis of chorionic villus biopsies; these techniques are only of value for 'at-risk' pregnancies. Neonatal screening for immunoreactive trypsin in the blood is simple and inexpensive. The differential diagnosis includes hypogammaglobulinaemia, immotile cilia syndromes, asthma, coeliac disease and Schwachman's syndrome (pancreatic insufficiency, neutropenia, metaphyseal chondrodysplasia, growth retardation and frequent infections). 1

Management The outlook for patients with cystic fibrosis has improved, and predicted average survival for children born in the

1990s is now 40 years. These improvements have been achieved by aggressive physiotherapy, antibiotic treatment, pancreatic enzyme supplementation and dietary support. The cornerstone of treatment of respiratory disease is regular chest physiotherapy, particularly postural drainage. In the first year of life prophylactic antistaphylococcal drugs may be helpful. Severe infective exacerbations require intravenous antibiotics, usually antipseudomonal agents (e.g. gentamicin, ceftazidime, piperacillin, ciprofloxacin). As the disease progresses regular nebulized antibiotics are of value. To facilitate home care, some patients with repeated flare-up of infection have indwelling central venous lines and administer their own antibiotics. Bronchodilators, and occasionally oral corticosteroids, may be useful in patients with airflow obstruction. Pneumothoraces are not uncommon and should be treated conservatively whenever possible. Malabsorption is treated by oral pancreatic enzyme supplements. Abdominal emergencies occur, including adult meconium ileus equivalent, intussusception, acute pancreatitis and bleeding from oesophageal varices. Diabetes is common. Adult patients will need counselling about fertility, as 98% of the males are infertile (owing to maldevelopment of the vas deferens) and for females pregnancy can be hazardous. Patients and their family need constant moral, psychological and practical support. Genetic counselling of the parents of children with cystic fibrosis and of the patients themselves is essential. Gene therapy There are now many studies reporting gene therapy in cystic fibrosis patients. These have viral or non-viral vectors to deliver CFTR-cDNA topically to the nose or the lungs by aerosol or by liquid instillation. Although in theory the transfer of normal CFTR-cDNA into affected CF cells in early life should cause the production of enough normal CFTR to prevent the clinical manifestations of cystic fibrosis, technical problems with the vectors and the clinical response have meant that gene therapy is not yet curative. In particular, the target cell type, the target airways (large or small) and the barrier created by viscous mucus are proving difficult to overcome. Furthermore, it is not clear how much CFTR mRNA is necessary to get into cells to prevent disease. These topics are still the focus of intensive research and are likely to lead to important therapeutic advances in the relatively near future. The infected secretions in patients with cystic fibrosis contain abundant highly polymerized DNA released from dead and dying neutrophils. This contributes significantly to the increased viscosity of the secretions, and the development of recombinant human DNase (rhDNase) for inhalation helps to reduce sputum viscosity and assist clearance in selected patients.

Lung transplantation in cystic fibrosis Lung transplantation programmes around the world now

incorporate a significant number of patients with cystic fibrosis. Transplantation is reserved for those with the most advanced lung disease, and presents special difficulties. In particular because of the level of infection present the procedure is often complicated by postoperative sepsis, and this can be very difficult to manage. Furthermore, previous pleural disease or tracheostomies can either be contraindications or make the surgical techniques difficult. Live related lobar donor transplantation has also been carried out, with mixed success. Patients with cystic fibrosis who have advanced liver disease have had successful liver transplantations in recent years. Clearly, however, the best strategies for the treatment of cystic fibrosis and all its complications are preventative, and in the long term it is in this arena that the most success is likely.

13

FURTHER READING ON CYSTIC FIBROSIS AJRCCM 1995 Cystic fibrosis - from the gene to the cure. Am J Respir Crit Care Med 151(3) suppl. Range S P, Knox A J 1995 rhDNase in cystic fibrosis. Thorax 50:321-322. Wilmott R W, Fiedler M A 1994 Recent advances in the treatment of cystic fibrosis. Pediat Clin North Am 41:431-451.

ASTHMA

2

Bronchial asthma has been recognized since ancient times. Despite a recent improved understanding of its pathogenesis and consequent new treatments, morbidity and mortality from the condition remain a substantial clinical problem.

Definition The American Thoracic Society defines asthma as 'a disease characterized by increased responsiveness of the bronchi to various stimuli, manifested by widespread narrowing of the airways that changes in severity either spontaneously or as a result of treatment'. Asthmatics may be said to have 'twitchy' airways, and it is an essential feature of the resulting bronchoconstriction that it is usually reversible, at least in part, with the administration of inhaled bronchodilators. There are a number of non-specific trigger factors that can start an attack (Table 13.23).

TABLE 13.23 Non-specific 'trigger' factors in asthma Upper respiratory tract infection Exercise Cold air Laughter

Irritants: smoke, paint or chemical fumes Drugs, e.g. (3-blockers Industrial causes

651

TABLE 13.24 Sensitizing chemicals and Industrial agents as a cause of asthma Material

Industry

Toluene diisocyonate (TDI) Soldering flux Cotton dust (byssinosis) Urine, serum, animal dander

Polyurethane industry Electrical, engineering industry Textiles Veterinary, agricultural and laboratory workers Bakers, millers Carpenters Chemical industry

Flour, grain Wood dust Sulphur dioxide, ozone, chloride

There are three components to asthma: the acute attack, the hyperresponsiveness of the airway to a heterogeneous variety of provocative stimuli, and persistent inflammation of the airways. Links between inflammation, hyperresponsiveness, and airflow obstruction are almost certainly central to the pathogenesis of asthma, but the mechanisms are not yet clearly defined. Much new information about the nature of the inflammatory infiltrate has been acquired from bronchial biopsies and lavage studies via the fibreoptic bronchoscope. Some late-onset asthmatics who have smoked may also have chronic bronchitis and are therefore difficult to classify, falling into the untidy category of 'asthmatic bronchitis'. Asthma caused by sensitizing chemicals and industrial agents is increasingly recognized (Table 13.24).

Epidemiology

652

Asthma is common, affecting between 5% and 7% of people in Europe and the United States. The peak age of onset is under 5 years, and in this age group it is more common in boys by a ratio of 3:2; over 5 years there is an equal sex ratio. There is an allergic component in one-third to half of all cases. About 2000 patients die from asthma in the UK each year, and it is likely that many of these deaths are preventable. There is a worldwide increase in the prevalence of allergy and asthma, especially in the developed countries. An international survey of asthma and allergy in childhood which assessed children aged 13-14 years demonstrated an over 20-fold difference in reported asthma symptoms worldwide. Symptoms were quite uncommon in eastern European countries and in Asia, and the highest prevalence was in the UK, Australia and New Zealand. There was a poor relationship between the incidence of allergic symptoms such as eczema and rhinitis and those of asthma, which might suggest that the risk factors are different or that they are following different time trends. Local environmental factors may play a part. In addition to the increasing prevalence there appears to be a trend towards increasing severity: in particular, in some countries such as New Zealand and Australia there is a

greatly increased hospital admission rate, and worldwide there is a steady increase in the use of anti-asthma medications. It is possible that environmental pollution and the consequent damage to protective epithelium may be contributing to this, allowing allergens to penetrate the airway lining. Furthermore, there have been disturbing epidemics of asthma deaths in recent years in England, New Zealand and elsewhere. It is possible that misuse and overreliance on inhaled (3-agonist bronchodilators may have played a part in these epidemics. Other risk factors for asthma mortality include cigarette smoking, previous hospital admission, and in particular admission to an intensive care unit, old age, social disturbance, ethnicity, dampness in the home environment, and the keeping of pets such as cats or dogs. Some of these factors may interact with and amplify other risk factors. There is one intriguing suggestion that lack of exposure to childhood viral infections (sometimes seen in single-child families) may actually predispose to the subsequent development of asthma.

Pathogenesis Cellular and molecular mechanisms in asthma About 70% of asthmatics have demonstrable atopy, with the production of specific IgE to environmental allergens, and positive skinprick tests with elevated serum IgE levels. Those who do not have atopy are labelled 'intrinsic'; they tend to have asthma that comes on later in life, are more commonly women, and may have aspirin sensitivity and nasal polyps. It is still a subject of speculation whether lowgrade allergy plays a part in intrinsic asthma. Inhaled allergens are critical to the initiation and perpetuation of airway inflammation in atopic individuals. Antigen-presenting cells are located in the bronchial epithelium and will present allergen to resident T cells, resulting in their activation and Th2 differentiation. Th2type cytokines (IL-4, IL-5, IL-10, IL-13 and GM-CSF) are derived from these activated T cells. The T cells also play an important part in the switching of B cells to IgE synthesis and mast cell survival, thus facilitating the inflammatory process by the activation and recruitment and stimulating survival of eosinophils in the airway. The hyperresponsiveness demonstrated in asthmatic airways correlates with severity, and these two factors in turn have been related to the numbers of activated eosinophils and T cells expressing IL-5 in asthmatic airways. Airway damage is one of the characteristics of asthma, and inflammatory mediators released by activated eosinophils such as MBP (major basic protein), ECP (eosinophil cationic protein) and LTC4 (leukotriene C4) play an important part in this. The role of the neutrophil is still being explored, but one important observation has been that in asthmatics dying suddenly from their disease, eosinophils are scarce in the airways but there is a marked neutrophil inflammation; in asthmatics whose death has been caused by a more indolent and slow downhill progression the eosinophil is the predominant inflammatory cell found in the airway at postmortem.

Bronchial 'hyperreactivity' appears to be the key to the asthmatic reaction. It is the unique property of asthmatic airways to react, by bronchoconstriction, inflammation and mucus production, to stimuli that would not elicit this reaction in normal airways. Possible mechanisms for bronchial hyperreactivity are listed in Table 13.25. There can be little doubt that immunological mechanisms (Ch. 4) play an important part in the asthmatic reaction (Fig. 13.28). Both immediate and delayed responses

TABLE 13.25 Mechanisms of bronchial hyperreactivity Inflammation, immunological liberation of bronchoconstrictor mediators Alterations in autonomic nervous control Abnormal ities of Ca 2+ flux Damage to epithelial 'tight junction'

are involved, and these can be demonstrated by the technique of bronchial provocation challenge testing. This is a method of provoking an asthmatic reaction in susceptible subjects by inhalation of antigens or chemicals. The immediate hypersensitivity reaction is mediated by degranulation of mast cells (Fig. 13.29). The mast cell degranulation is produced by the union of antigen with IgE antibody which is bound to the mast cell surface. Degranulation liberates a variety of substances, including preformed mediators and newly formed metabolites of arachidonic acid from the cell membrane (prostaglandins, leukotrienes etc.; see below). When asthmatic subjects are challenged with appropriate inhaled antigen and lung function is measured, they may show an immediate bronchoconstriction response within a few minutes. Lung function then returns towards baseline, but a further apparently unprovoked fall may occur 4-6 hours later.

FIG. 13.28 Interaction of inflammatory cells, inflammatory mediators and reflex mechanisms in the airway wall in asthma.

13

653

The early response is thought, in the main, to be due to the acute effect of histamine release from mast cells. The late response is more complex, mediated through the actions of inflammatory mediators such as leukotrienes and platelet-activating factor and the recruitment and activation of inflammatory cells. This highly complex pathogenic mechanism is the focus of much research, as it may represent a model for the ongoing inflammatory reaction in asthma and therefore be the target for antiinflammatory prophylactic drugs. Bronchial smooth muscle and secretions are regulated by neural mechanisms (Fig. 13.30). The balance of autonomic nervous control is important. The cholinergic vagal impulses have bronchoconstrictor and secretory actions. These effects are opposed by pYadrenergic receptors, which are bronchodilator and antisecretory. The (3adrenergic mechanism is supported by circulating cate-

cholamines. In patients with asthma, vagal reflexes may play an important part in bronchial hyperreactivity, particularly to stimuli such as cold air and exercise. Other mechanisms may play a part in bronchial hyperreactivity. Calcium flux across cell membranes is important for smooth muscle contraction and for mast cell degranulation, and agents that block this flux may prove to be useful in treatment. Finally, there is some evidence that the 'tight junction' between epithelial cells on the surface of the bronchial epithelium may be abnormally 'leaky', allowing antigen readier access to the mast cells underneath. It is therefore clear that no single mediator or mechanism is responsible for the asthmatic reaction, and research continues to elucidate this highly complex process. Pathology and pathophysiology The pathogenic mechanisms involved in asthma result in the histological changes shown in Figure 13.31. These changes can be acute or chronic. The sputum of asthmatics is often yellow and tenacious and contains many eosinophils, epithelial cells and bronchiolar casts ('plugs'). At postmortem in patients who die of acute asthma, many plugs are found in large and small airways.

Clinical features

FIG. 13.29 Mast cell mediators The release of preformed and newly formed mediators from mast cells involves calcium flux across the cell membrane, and in the case of the newly formed mediators, phospholipase A2 activation.

654

Asthma symptoms are episodic. Between attacks, patients, especially younger ones, may be asymptomatic and have no abnormal physical signs. The common symptoms of an asthma attack are cough, wheeze and breathlessness. Cough may be the only symptom, especially in children. Almost all asthmatics are worse at night and most will feel 'tight' in the morning. The fall in peak flow rate in the early hours of the morning is referred to as 'morning dipping' (Fig. 13.32). Acute asthma and sudden asthma deaths occur

FIG. 13.30 The control of airway calibre The balance of bronchodilatation from sympathomimetic p-receptor stimulation [A], and bronchoconstriction from parasympathetic vagal activity [B], is an important determinant of airway calibre.

TABLE 13.26 Clinical syndromes associated with asthma Eosinophilic syndromes Allergic bronchopulmonary aspergillosis Idiopathic pulmonary eosinophilia Tropical eosinophilia Churg-Strauss syndrome

13

Aspirin sensitivity and nasal polyposis Pericarditis Carcinoid syndrome

FIG. 13.31 The pathology of asthma Intense inflammation, with oedema and intraluminal mucous plugs, is the striking pathology of asthma.

contracting the accessory muscles of respiration. When the condition progresses despite active treatment, the patient is then said to have 'acute severe asthma' (p. 658). A number of clinical conditions are associated with asthma (Table 13.26). Allergic bronchopulmonary aspergillosis and other eosinophilic syndromes are discussed on page 660.

Investigation X-ray findings Most asthmatics have normal or hyperinflated lungs on chest X-ray. Mucus plugs can cause focal or segmental atelectasis, and sometimes collapse of a lobe or lung. Transient infiltrates are seen in association with eosinophilia. Spontaneous pneumothorax can complicate asthma in the acute phase.

FIG. 13.32 Peak expiratory flow rate (PEFR) in asthma The measurement of peak flow is essential for the management of asthma. PEFR is variable in asthmatics; it is lower in the early hours of the morning and on wakening ('morning dipping') and increases with p2 agonists (e.g. salbutamol). The chart is typical of a patient recovering from an episode of moderately severe asthma.

most frequently in the early hours of the morning. There is a personal or family history of allergy (rhinitis, conjunctivitis, eczema or urticaria) in up to 50% of patients. In many asthmatics, particularly in adult life, the airflow obstruction becomes persistent and somewhat resistant to treatment. This may be because 'remodelling' occurs in the asthmatic airway by the laying down of connective tissue inside and outside the airway wall and the proliferation of smooth muscle in the airways. Some of this may be related to epithelial damage activation and repair, but whatever the mechanism it does appear to be an important feature in the long-term pathogenesis of airflow obstruction. It is intriguing to speculate whether early treatment might prevent such morphological changes. In chronic asthma, particularly that which has persisted from childhood, there may be chest wall deformity. When asthma is severe the patient is distressed and vigorously

Other investigations Blood eosinophilia is common; it is usually greater in atopic asthma. Skinprick tests to common allergens may occasionally reveal an unsuspected source of allergy. Lung function tests may be normal between attacks. During episodes of asthma the PEFR, FEV1 and FEV1/VC ratio will be reduced and airways resistance increased. In addition to airflow obstruction the patient with acute or chronic asthma will have laboratory evidence of hyperinflation with increased total lung capacity (TLC), residual volume (RV), and ratio of RV to TLC. The transfer factor (diffusing capacity) for carbon monoxide is normal or increased in uncomplicated asthma. Hypoxaemia is common.

Diagnosis and management Recognition of asthma is not always straightforward. It is often missed in children who present with episodes of cough, wheeze and dyspnoea. Such 'wheezy bronchitis' in childhood is almost always asthma. The differential diagnosis of chronic asthma includes 'asthmatic' bronchitis, left heart failure, centrally obstructing tumour or foreign body, and recurrent pulmonary emboli. Asthma which repeatedly fails to respond symptomatically, or in terms of lung function improvement, to inhaled (3-agonists should

655

TABLE E 13.27

1. 2. 3. 4. 5. 6.

Six-part asthma management programme

Educate patients to develop a partnership in asthma management. Assess and monitor asthma severity with objective measures of lung function. Avoid or control asthma triggers. Establish medication plans for chronic management. Establish plans for managing exacerbations. Provide regular follow-up care.

always cause the diagnosis to be reconsidered. It may be difficult or impossible to distinguish late-onset asthma from chronic obstructive bronchitis in the elderly smoker. Many of them respond well to bronchodilators and have a better prognosis than patients with fixed airways obstruction. Supervision of chronic asthma is as important as in any serious chronic condition. Some hospitals run a very effective 'open-door' policy for asthmatics, allowing them to obtain immediate specialist advice whenever their asthma is severe. Education about the nature of the disease, the use of drugs and delivery systems, the avoidance of allergens and the recognition of deteriorating asthma are all important. Asthma can seldom be 'cured'. Long-term therapy is usually required, for prophylaxis or control, and it is critically important that this simple truth is appreciated by both the patient and the physician.

TABLE 13.28 Treatment of asthma Proven benefit

Dubious benefit

Potential benefit

Anticholinergics Disodium cromoglycate

Cough suppressants Potassium channel Mucolytics openers Phosphodiesterase Corticosteroids Vaccination inhibitors Calcium antagonists Sympathomimetics Antihistamines Ciclosporin A Prostaglandin analogues/ Tneophylline, aminopriylline Expectorants antagonists Leukotriene antagonists

Management

656

An international consensus report on the diagnosis and management of asthma recommended a six-part management programme (Table 13.27). Education includes the joint development of a treatment plan by the clinician and the patient, with guided self-management strategies. For the assessment and monitoring of severity, spirometry is recommended for the initial consultation and a period of home peak flow monitoring may also be useful. In advising the patient to avoid or control trigger factors, these must first be identified. Accurate identification and consequent avoidance may reduce symptoms, the need for medication and levels of airway hyperresponsiveness. Medication plans need to accommodate variability among patients, and a stepwise approach to drug therapy is now accepted worldwide (Fig. 13.33). It is important that patients are able to identify the symptoms of the onset of acute exacerbations and are given appropriate advice about what course to follow; following exacerbations, regular supervision and support by a clinician is essential. Treatment regimens (Table 13.28) can be broadly classified as long-term prophylactic treatment and treatment of exacerbations and acute severe asthma. Hyposensitization therapy (desensitization immunotherapy) is discouraged in the UK but widely practised throughout the world, despite little evidence of efficacy.

FIG. 13.33 The treatment of asthma

Drags of proven benefit Sympathomimetics (3-Agonists exert their action by stimulating the enzyme adenyl cyclase, which catalyses the formation of cyclicAMP within the bronchial smooth muscle cell (Fig. 13.30). All available pVagonists can be given by inhalation, which is the preferred route, and some by oral, i.v. and subcutaneous administration. Salbutamol and terbutaline are the two most commonly prescribed. Salbutamol is available in a metered dose inhaler (100 jig per dose), as a dry powder for inhalation ('Rotacap') for patients whose coordination is poor or for children, and as tablets, slow-release tablets, i.v. injection, a nebulizer solution and a syrup. Terbutaline is available in similar forms. Much research effort has been directed to the development of inhalers, such as 'turbohalers' that do not use fluorocarbon carriers, to comply with the proposed ban on the use of these chemicals.

Inhaler spacer devices can improve drug inhalation efficiency in patients unable to manage a conventional inhaler. With agonists, muscle tremor and tachycardia commonly occur, but tend to settle with prolonged usage. Longer-acting f^-agonists (e.g. salmeterol, formoterol) have been introduced, and may help control nocturnal symptoms. Leukotriene antagonists Leukotrienes are formed by the action of 5-lipoxygenase and 5-lipoxygenase activating protein (FLAP) on arachidonic acid to form leukotriene A4 (LTA4) which is in turn metabolized to leukotriene B4 or the cysteinyl leukotriene C4, leukotriene D4 and leukotriene E4 (LTC4, LTD4, LTE4). Leukotriene B4 is a powerful chemotactic agent, whereas the cysteinyl leukotrienes LTC4, LTD4 and LTE4 cause bronchoconstriction, increase mucous secretion, airway wall oedema, eosinophilia and increase bronchial hyperresponsiveness. These actions are of interest in the pathogenesis of chronic bronchitis and asthma; in particular, leukotriene B4 inhibition might reduce airway inflammation in chronic bronchitis and these studies are being pursued. However it was the role of the cysteinyl leukotrienes that generated most pharmacological interest and there are now specific cysteinyl leukotriene receptor antagonists such as montelukast, zafirlukast and pranlukast available for clinical use in asthma. There are data to show that they give good control of symptoms, inhibit exercise induced asthma, and are safe when given by the oral route. Additive effects to inhaled corticosteroids have also been demonstrated. Methyl xanthine derivatives (aminophylline and theophylline) Oral use is preferred in routine treatment, and intravenous use for acute asthma. Divided doses of slow-release theophyllines are effective and safe bronchodilators and they have prophylactic anti-inflammatory activity. These sustained-release formulations are useful when given at night for the prevention of nocturnal asthma and 'morning tightness'. Although 'therapeutic' serum theophylline levels of 10-20 mg/L have been targeted in the past, it is likely that levels as low as 3-5 mg/L have useful antiinflammatory actions. If excessive doses are given, or occasionally even when blood levels are in the therapeutic range, side-effects may occur. These include nausea, abdominal pain, headache, tremor, insomnia and palpitations, convulsions and cardiac arrhythmias. The dose administered should be reduced in elderly patients, those with liver disease, and those taking other drugs that increase theophylline blood levels (e.g. erythromycin). Anticholinergics Anticholinergic agents have proved to be of some value, particularly in older patients with late-onset asthma or 'asthmatic bronchitis'. Atropine can be delivered by inhalation in its methonitrate form from a nebulizer, although it is more convenient to administer the anticholinergics, ipratropium and oxitropium, as a metered-

dose aerosol. For patients optimally treated with inhaled pYagonists the additional benefit of anticholinergic drugs is seldom important.

13

Disodium cromoglycate It has been suggested that cromoglycate acts by 'stabilizing' the mast cell and preventing mediator release. However, modification of neural mechanisms may be equally important. Therapy is only effective when used on a regular prophylactic basis, and it has no intrinsic bronchodilator activity. Younger asthmatics tend to benefit most from its regular use, and it can also be used to 'block' exercise-induced asthma when administered about 20 minutes before planned exertion. Useful nasal and ophthalmic preparations are available for the treatment of hay fever. The drug has no serious side-effects. Corticosteroids Corticosteroids, probably by their anti-inflammatory action, are the most powerful drugs available for the treatment of asthma. Attitudes towards, and use of, Corticosteroids for the long-term management of asthma have been dramatically changed by the introduction of powerful, topically active fluorinated Corticosteroids for use by the inhaled route. Beclomethasone diproprionate, budesonide and fluticasone are most often prescribed. They enable asthmatics to be treated with low-dose inhaled corticosteroids, which are not absorbed in sufficient quantities to cause adrenal suppression or iatrogenic Cushing's syndrome, and are highly effective when used on a regular basis in the prophylaxis of asthma. Inhaled corticosteroid can also be delivered via a nebulizer. However, high doses for prolonged periods can cause side-effects, including osteoporosis. In some patients inhaled Corticosteroids predispose to oral Candida infection, but this can usually be easily controlled. If long-term oral Corticosteroids are essential for the control of asthma, patients should also be on maximum doses of inhaled Corticosteroids, thereby permitting the oral prednisolone dosage to be reduced by as much as 10 mg daily.

Inhaler therapy Many different inhaler devices are available. They include metered-dose inhalers (MDI), which traditionally used CFCs as a propellant, but which now, for environmental reasons, are being switched to CFC-free inhalers, using HFA (hydrofluoroalkane) carriers. MDI inhalers may be used with 'spacer' devices to facilitate use and improve airway deposition of the drugs. Breath-activated devices are useful aids to coordination. Many patients now use dry powder inhalers, which have the twin advantages of ease of use and being environmentally friendly. Nebulizer therapy Nebulizers can deliver pYagonists, anticholinergic drugs and, occasionally, Corticosteroids to the airways. They are increasingly used at home by patients with chronic airways obstruction, but in view of the large doses of drug used

657

(particularly pVagonists), such treatment requires careful assessment and supervision. It is important to document that nebulized bronchodilators confer a genuine advantage over maximal treatment with metered-dose inhalers. Summary of routine treatment of asthma Figure 13.33 illustrates the treatment strategy for asthma. Oral steroids are prescribed only when maximum therapy with all other drugs fails to control the disease. Selected patients with severe chronic asthma may be helped by immunosuppressive therapy (e.g. methotrexate). Treatment of acute severe asthma 1 The patient with acute severe asthma is distressed, dyspnoeic, and using accessory respiratory muscles; there are widespread inspiratory and expiratory wheezes and the chest is hyperinflated. When airways obstruction and bronchial plugging is severe and widespread, the flow of air through bronchi is greatly decreased and wheezing becomes less obvious, eventually resulting in the ominous 'silent chest' of advanced acute severe asthma. As the condition progresses, cyanosis, dehydration, exhaustion and ventilatory failure ensue. Indices of the severity of acute asthma are listed in Table 13.29. In the accident and emergency department, PEFR, heart rate and pulsus paradoxus (p. 620) are all related to asthma severity and should be documented with particular care. Arterial blood gas tensions should be measured in all patients whose asthma is sufficiently severe to require hospital assessment. Acute asthma requires immediate evaluation and treatment. Oxygen is necessary, and in almost all circumstances can be liberally administered without fear of CO2 narcosis. Salbutamol solution in a dose of 2.5-5.0mg can be given via a nebulizer and is usually very effective. This can be repeated 3-6-hourly as required. In addition, in very severe episodes, aminophylline is given in a dose of 5 mg/kg body weight intravenously over 20 minutes, to be followed by an infusion of 0.5mg/kg/h. In patients already taking theophyllines an infusion of aminophylline (0.5 mg/kg/h) can be started and the blood theophylline level checked. The dose of aminophylline is reduced if the patient is more than 55 years old, or if there is evidence of liver disease or cardiac failure. A theophylline blood level should be measured 18 hours after starting the infusion and the dose adjusted to remain within the therapeutic range of 10-20mg/L. If there are problems with venous access, subcutaneous terbutaline (0.25-0.5 mg) is helpful. Systemic corticosteroids must be given to all patients whose asthma is sufficiently severe to warrant hospital admission, and also to many less severe cases satisfactorily controlled in the emergency department. It is probably immaterial how corticosteroids are given; it may be as an i.v. bolus of hydrocortisone 3-6 mg/kg body weight fol-

TABLE 13.29 Indications of severe asthma Low PEFR (<150L/min, poor bronchodilator response) Associated with (one or more of): Vigorous accessory muscle activity Heart rate greater than 130 Pulsus paradoxus Arterial Po2 less than 8kPa (60mmHg) Arterial Pco2 greater than 5.5 kPa (41 mmHg) Gross overinflation (X-ray) Central cyanosis Disturbances of consciousness ECG abnormalities Pneumothorax/pneumomediastinum

lowed by an infusion of 3-6 mg/kg 6-hourly, or alternatively 60 mg of oral prednisolone for the average-sized adult. Supportive measures are also important. Patients with acute asthma become dehydrated. Physiotherapy is not helpful. Reassurance helps, but the best reassurance is improvement of the asthma. Sedatives should never be used unless intubation and mechanical ventilation are being contemplated. Antibiotics are prescribed if there is evidence of bacterial infection. Some patients fail to respond to treatment and become progressively more hypoxic. At the onset of an asthma attack patients hyperventilate and therefore have an acute respiratory alkalosis. As airways obstruction increases, exhaustion supervenes, ventilatory pump failure occurs, alveolar ventilation diminishes, CO2 rises and pH falls, leading to an acute respiratory acidosis. At this point the patient may require intubation and mechanical ventilation. Small tidal volumes should be used, and most asthmatics who need mechanical ventilation will need at least 48 hours on the ventilator to allow their asthma to respond to therapy. Bronchodilator and steroid therapy should be continued while the patient is being ventilated. Prognosis About 50% of children 'grow out of their asthma in early adult life. Late-onset asthma is almost always chronic, tends to be more severe, and usually needs more aggressive treatment. About 30% of asthma deaths occur within 2 hours of the onset of the acute attack, and these may be difficult to prevent. However, many of the remaining 70% of the 2000 deaths that occur annually in the UK should be preventable, by appropriate long-term therapy and the early recognition of deteriorating asthma.

ALLERGIC RHINITIS 1

658

MCQ 13.12

Allergic rhinitis (commonly known as 'hay fever', although it is not related to hay and there is no fever) is a condition

characterized by rhinorrhoea, nasal obstruction, sneezing, conjunctivitis, lacrimation, and nasal and pharyngeal itching. It is usually seasonal, with tree pollens being the allergen in the spring months and grass pollen in the summer. A perennial form occurs in those patients sensitive to allergens such as house dust, which are in the air all year round. Food allergy is an often quoted but clinically rare cause of rhinitis. Pathologically the mucosa of the nose is oedematous and inflamed, and the secretions are rich in eosinophils. Nasal polyps and sinus infection may be present, especially in the perennial type. The conjunctivae are congested and oedematous. The diagnosis is made by taking a careful history. Skinprick testing for a battery of common allergens is helpful. Serum IgE is usually elevated, and occasionally a radioallergoimmunosorbent test will help to define the allergen. The major differential diagnosis is from vasomotor rhinitis, which is a similar syndrome but with no documented allergic basis. Other conditions that may present in a similar fashion include exposure to irritants and repeated upper respiratory tract infections.

Management Although the symptoms of allergic rhinitis tend to improve as the subject gets older, they can, when active, be severe. Most of the relevant allergens are ubiquitous and therefore unavoidable. The first line of treatment is topical: nasal sprays or drops containing corticosteroids, such as beclometasone or budesonide, or disodium cromoglycate, are useful when used on a regular basis. Disodium cromoglycate eyedrops are also helpful for the conjunctivitis. These drugs can be supplemented by a non-sedating antihistamine by mouth, such as loratadine and fexofenadine or astemizole. The place of hyposensitization is debatable, and is usually reserved for isolated grass pollen allergy. Hyposensitization can be complicated by anaphylactic shock and should only be undertaken where resuscitation facilities are available. Occasionally a small dose of oral corticosteroids, e.g. prednisolone (10-15mg/day), or one or two injections of a depot corticosteroid during the season may be necessary.

FURTHER READING ON ASTHMA British Thoracic Society 1993 Guidelines on the management of asthma. Thorax 48 (2: suppl). The British Guidelines on Asthma Management 1995 Review and Position Statement. 1997 Thorax 52 (Suppl 1). Hall J B, Corbridge T C 1995 The assessment and management of adults with status asthmaticus. Am J Respir Crit Care Med 151:1296-1316. Holgate S 1993 Mediator and cytokine mechanisms in asthma. Thorax 48:103-109. Monson J P 1993 Systemic effects of inhaled corticosteroids. Thorax 48:955-956.

TABLE 13.30 Classification of pulmonary eosinophilia Known causes Fungal allergy Aspergillus fumigatus Candida albicans Stemphylium canugenosum Dresclera lawaiiensis Curvularia lunata Helminosporium Helminthic infections Ascaris lumbricoides (Lb'ffler's syndrome) Strongyloides Filaria ('tropical pulmonary eosinophilia') Ancylostoma

13

Drugs and toxins Sulphonamides (including sulphasalazine) Nitrofurantoin Sodium aminosalicylate Penicillin Tetracycline Chlorpropamide 'Spanish toxic oil' syndrome Inorganic chemicals (e.g. nickel) Lymphangiography Blood transfusion

Unknown causes Simple pulmonary eosinophilia Chronic eosinophilic pneumonia Acute eosinophilic pneumonia Churg-Strauss syndrome Idiopathic hypereosinophilia syndrome

PULMONARY EOSINOPHILIA Pulmonary eosinophilia is a transient pulmonary infiltrate associated with an elevated blood eosinophil count. A cause can only be identified in some cases (Table 13.30). The blood eosinophil count is elevated above its normal upper level of 400/mm3 and may in some circumstances be as high as 50000/mm3. The serum IgE level is usually elevated in those cases with an identifiable cause, especially the allergic mycoses. Asthma is almost always present in cases of allergic bronchopulmonary aspergillosis, occurs in about half the patients with cryptogenic pulmonary eosinophilia, and is unusual with helminth infections apart from filariasis.

Pathogenesis The eosinophil has cytotoxic and anti-inflammatory properties. The cytotoxic properties act against parasites but may also damage host tissues. The pathological response to the various identifiable causes of pulmonary eosinophilia depends on the route by which they reach the lungs; if inhaled (e.g. allergic mycoses) the response is in the airways (bronchocentric), whereas if delivered to the lung via the pulmonary circulation the response is in the blood vessels (angiocentric).

659

Allergic bronchopulmonary mycoses By far the commonest cause in developed countries is allergic bronchopulmonary aspergillosis (ABPA). This is due to sensitivity to the ubiquitous fungus Aspergillus fumigatus. This fungus thrives in warm, wet conditions, and symptoms therefore tend to occur in the autumn. The bronchial tree provides an ideal environment for colonization. When the fungus is inhaled there is an immediate hypersensitivity reaction, with eosinophils being drawn into the area. There may be several responses thereafter: • A simple asthmatic reaction with an eosinophilia but no radiological change; • Pulmonary eosinophilic infiltrates with consolidation and chest X-ray shadowing; • Mucus impaction with distal collapse and progressive airway damage due to the release of tissue-damaging factors; this in time may lead to a characteristic proximal bronchiectasis and upper zone fibrosis (Fig. 13.34), sometimes making it difficult to differentiate ABPA from old tuberculosis. Clinically, ABPA presents, often in early adult life, as worsening asthma. The patient may complain of coughing up rubbery brown or green plugs, and exacerbations of asthma are often accompanied by fever and fleeting pulmonary infiltrates. Almost all patients have a positive immediate-type hypersensitivity to Aspergillus fumigatus on skinprick testing, and about 90% have precipitating antibody. Treatment with oral prednisolone (30-40 mg/day) leads to rapid symptomatic and radiological improvement; whether or not it affects the long-term outcome is uncertain. Physiotherapy and occasionally bronchoscopy help to remove troublesome plugs, and the asthma responds to conventional treatment, usually including inhaled steroid. Many patients require a small maintenance dose of oral prednisolone to prevent relapse. Drugs and toxins There are a wide variety of drugs and toxins which can cause a pulmonary eosinophilic syndrome (Table 13.30). The drug reaches the lung via the pulmonary circulation and sets up an allergic reaction in the vessel wall. There is frequently an associated skin reaction. Patients may present with a pneumonia-like illness, with dyspnoea, cough, fever and pleuritic pain, or as a pulmonary vasculitis. Rarely, if the subject is exposed over a prolonged period, a type of fibrosing alveolitis develops. Whatever the progression, fleeting chest X-ray shadows and blood eosinophilia will be present. Almost all cases improve if the offending agent is stopped, and oral corticosteroids help speed recovery.

1

660

MCQ 13.13

2

MCQ 13.14

FIG. 13.34 [A] Chest X-ray of a patient with allergic bronchopulmonary aspergillosis and asthma, showing acute right upper zone shadowing. This shadowing resolved with steroid therapy. [B] A subsequent CT scan demonstrates right upper lobe bronchiectasis - dilated, thickened bronchi (arrow).

Helminthic infections Some helminths (Ascaris, Strongyloides, Toxocard) are borne to the lungs via the circulation, where they are attacked by eosinophils. During this phase the patient may be febrile and have pulmonary infiltrates, and the dying parasite may initiate an allergic response and thus an asthmatic attack. Filarial infection (e.g. with Wuchereria bancrofti) may cause asthma, and patients can also develop diffuse fibrosis and pulmonary hypertension. The chest X-ray in helminthic infections shows widespread nodular infiltrates which are at first transient but may become persistent. The diagnosis is confirmed by appropriate complement fixation tests and by finding ova in the stools (see also Ch. 9). 1

Unknown causes There is a spectrum of conditions of unknown aetiology that cause fleeting eosinophilic infiltrates in the lung and a blood eosinophilia. Pathologically, these range from a localized or generalized pulmonary eosinophilia to eosinophilic syndromes with vasculitis and granuloma formation. Classification is difficult; the most important distinction is between the clinical syndrome of cryptogenic pulmonary eosinophilia and those syndromes associated with vasculitis and granuloma formation. Cryptogenic pulmonary eosinophilia (eosinophilic pneumonia) Cryptogenic pulmonary eosinophilia is a syndrome of fever, weight loss, cough and breathlessness, with widespread, mainly peripheral, pulmonary infiltrates (Fig. 13.35) and a marked blood eosinophilia. The patient is usually non-atopic, but 10% have previous asthma. No cause can be identified. Some patients progress to a vasculitic/granulomatous type of disease. There is a rapid response to oral corticosteroids; treatment should be carefully tailed off, as recurrences are common, and usually corticosteroid treatment for up to 2 years is required. Allergic angiitis and granulomatosis This is a rare group of conditions in which allergic asthmatic patients present with symptoms similar to those of cryptogenic pulmonary eosinophilia. Radiologically they may show extensive transient pulmonary infiltrates, nodules, cavities, and occasionally pleural and pericardial

FIG. 13.35 Pulmonary eosinophilia Extensive, ill-defined peripheral consolidation in the left upper zone and less marked consolidation in the left lower zone in a patient with cryptogenic pulmonary eosinophilia. Characteristically, new shadows appear and previous ones resolve spontaneously.

effusions. There is a blood eosinophilia and a raised erythrocyte sedimentation rate and raised serum IgE, and about 50% of patients will have a positive ANCA. Lung biopsy shows a vasculitis, and if there is an allergic granulomatous reaction this is known as the Churg-Strauss syndrome, although there is probably a broad spectrum of pathological overlap in these syndromes. In many ways the syndrome is similar to polyarteritis nodosa, although lung involvement in typical polyarteritis nodosa is rare. There have been reports of the development of the ChurgStrauss syndrome in patients receiving treatment for asthma with leukothriene receptor antogonists; it is likely, however, that this is an 'unmasking' phenomenon caused by withdrawal of oral corticosteroid therapy. Other organs, such as the liver, spleen, heart, skin and central nervous system, may be affected by the vasculitis but, unlike in polyarteritis nodosa, renal involvement is uncommon. Treatment with prednisolone in high doses, with or without azathioprine or cyclophosphamide, significantly reduces mortality, although infective or cardiac complications may still be fatal. About half the patients with this condition will die within 3 months without corticosteroid therapy. Adequate treatment increases mean survival time to 2 years.

13

CHRONIC BRONCHITIS AND EMPHYSEMA 2 Chronic bronchitis Chronic bronchitis is a clinical diagnosis in which there is cough productive of sputum on most days for 3 months of the year for 2 or more years, which is not due to a specific respiratory disease such as bronchiectasis. The disorder is characterized by excess mucus secretion. Pathologically there is hyperplasia and hypertrophy of the tracheal and bronchial mucus glands and an increase in the glandular elements of the bronchial wall. There is also inflammation, and eventually fibrosis, of small airways. The increase in intraluminal mucus and thickening of the bronchial wall produce airways narrowing and increased resistance, the fall in FEV1 correlating with the increase in mucous membrane thickness. Emphysema Emphysema is denned by its pathology and is characterized by the destruction of respiratory tissue and permanent enlargement of the unit of the lung distal to the terminal bronchiole (the acinus). The injury to alveolar septa is caused by proteolysis. Lung tissue is normally protected by a shield of proteinase inhibitors, derived from the blood but also synthesized locally. When the proteinase-antiproteinase balance is disturbed, favouring proteolytic activity, elastin destruction and septal digestion occur. The consequences of this are loss of the 'tethering' support of airways, leading to collapse on expiration and

661

reduction of lung elastic recoil and pulmonary capillary bed. In the past much importance has been placed on the distinction between chronic bronchitis and emphysema. In the majority of patients both conditions coexist, usually in heavy cigarette smokers, and the physician therefore makes a clinical diagnosis of chronic bronchitis and emphysema.

Aetiology and prevalence Chronic bronchitis and emphysema are responsible for the personal disability and misery of tens of thousands of patients and impose a huge social and economic burden on society. Respiratory disorders are an important cause of death in the UK and, of these, chronic bronchitis and emphysema constitute a large proportion. In the UK 10% of absences from work are caused by chronic bronchitis and emphysema, and approximately 10% of occupancy of acute general medical hospital beds is the result of these diseases. Atmospheric pollution and occupational dust exposure are minor aetiological factors in chronic bronchitis; the dominant causal agent is cigarette smoke. For symptoms and physiological changes to be demonstrated it probably needs a smoking history of 20 pack years. Smoking also causes emphysema, probably damaging the lung by the release of proteolytic enzymes. Smoke-affected pulmonary alveolar macrophages, present in greater numbers than usual, release neutrophil chemotactic factor and the attracted neutrophils are damaged by smoke and release proteolytic enzymes, especially elastase, capable of lysing elastin, collagen and basement membranes. The effectiveness of oci-antitrypsin is impaired by smoking and the unchecked proteolysis results in centrilobular emphysema. This process is particularly rapid in patients who are deficient in aj-antitrypsin. Overall severity of airways obstruction is related to the number of cigarettes smoked.

662

a-Antitrypsin deficiency This genetic disorder affects one in 5000 people, is equally common in men and women, and predisposes to the early development of emphysema, often before the age of 40. The emphysema is panlobular and basal (Fig. 13.14) and lung destruction is accelerated by cigarette smoke. Nonsmoking heterozygotes do not have emphysema, but it is possible that smoking heterozygotes develop emphysema more readily than smoking normals. Heterozygotes have about 60% of normal ocrantitrypsin values and homozygotes have values of < 10%. oci-Antitrypsin appears to inhibit neutrophil elastase, but more recent studies have suggested that several enzymes and inhibitors are involved in the proteinase-antiproteinase balance in the lung. The feasibility and clinical value of arantitrypsin replacement therapy is under investigation. Some patients with advanced disease have been managed by lung transplantation.

Mechanism of airflow obstruction In chronic bronchitis and emphysema the fundamental cause of reduced ventilatory capacity and breathlessness is the limitation of expiratory airflow (p. 608). Secondary to expiratory limitation, residual volume is increased, the thorax is hyperinflated and inspiratory capacity is markedly impaired, in part because of the compromised function of the inspiratory muscles. In chronic bronchitis airways resistance is increased and expiratory airflow reduced by the thickening of the bronchial wall and excessive intraluminal mucus. In emphysema a more important mechanism is the narrowing and collapse of airways during expiration, as a consequence of loss of the lung elastic recoil which normally keeps them open.

Clinical features Chronic bronchitis and emphysema develop over many years and patients are rarely symptomatic before middle age. Symptoms are initially minor, perhaps a morning cough productive of a little sputum. In some patients chronic bronchitis can remain a trivial problem, but in many smoking-related airways obstruction coexists and the patient then develops breathlessness. Initially breathlessness is on exertion, but exercise capacity progressively and slowly deteriorates and eventually patients become respiratory cripples, distressed by dyspnoea even at rest. Patients with predominant bronchitis are prone to periodic infections. Often the infective process remains confined to the bronchial tree, but sometimes infection involves the surrounding lung, with consequent pneumonia; thus, although some patients may have radiological changes of pneumonia in association with an infective exacerbation, the majority show no change on chest X-ray. Eventually patients with chronic bronchitis develop severe hypoxaemia, hypercapnia and peripheral oedema-cor pulmonale (the so-called 'blue bloater'). Patients with substantial emphysema tend to be very breathless, ventilating sufficiently to maintain normal arterial carbon dioxide and near-normal oxygen tensions. The development of cor pulmonale is a late and terminal event in these 'pink puffers'. These two clinical extremes overlap; many patients with a normal arterial carbon dioxide tension eventually become cyanosed, but the broad clinical distinction between the 'pink puffer' and the 'blue bloater' is useful. Oxygen therapy is much more difficult in the hypercapnic blue and bloated patient, sedation is more hazardous, and general anaesthesia poses greater risks because of postoperative sputum retention and difficulty in weaning such patients from assisted ventilation. Some patients have bronchial hyperreactivity, and exposure to cold air, acute temperature changes, dusts and cigarette smoke causes increasing breathlessness, respiratory distress and wheeze. The development of cor pulmonale indicates a poor prognosis, with a 30% 5-year survival. In association with severe hypoxia and hypercapnia, pulmonary arterial hypertension develops and some patients also have an ele-

vated jugular venous pressure. However, in many patients cor pulmonale is not simple cardiac failure, as cardiac output is frequently normal and there can be peripheral oedema without elevation of the jugular venous pressure. The mechanism of cor pulmonale remains obscure, but in part it reflects an increase in blood volume, perhaps due to hypercapnia and hypoxia causing sodium retention by the kidney. Cor pulmonale is unusual in the absence of hypercapnia. There is evidence that optimum therapy and careful follow-up of patients with cor pulmonale can produce good results. Physical signs In predominantly emphysematous patients, inspiratory airways resistance is not increased and inspiration is therefore quiet, whereas patients with predominantly chronic bronchitis have noisy breathing. To control airways collapse on expiration, patients with emphysema apply a positive pressure to the bronchial tree by the technique of purse-lipped breathing. As the diseases progress the physical signs may eventually include those of severe airways obstruction, hypoxia, hypercapnia and cor pulmonale. Severe hyperinflation is indicated by a reduction in the distance between the suprasternal notch and the cricoid cartilage, a barrel-shaped chest, and reduced abdominal outward movement during inspiration (reflecting limited

FIG. 13.36 Bullous emphysema The large bullae (right upper zone, right lower zone peripherally) display no vascular markings. Elsewhere the lung markings appear obvious, partly because of increased blood flow through relatively normal lung and partly because bullae compress adjacent lung tissue.

descent of the flattened diaphragm) coupled with pronounced movement of the upper chest. Patients with chronic airflow limitation often have early inspiratory crackles which can be heard at the mouth by the examining physician. Weight loss is common with advanced emphysema. Patients with emphysema have translucent lungs on chest X-ray, with few vascular markings, and bullae may also be present (Fig. 13.36). CT scanning is a sensitive technique for detecting emphysema (Fig. 13.37). Patients with predominantly chronic bronchitis may show bronchial wall thickening. Hyperinflation is more severe in emphysema. Commonly, radiological abnormality is most striking in the upper zones, but in cti-antitrypsin deficiency disease is predominantly basal.

13

Respiratory function Many respiratory function tests designed to detect early COPD in the lung have been investigated. These include tests of 'small airways disease' which, although sensitive, have not been found to be useful in clinical practice. The forced expiratory volume in one second (FEV)) remains the single most useful indicator of airflow obstruction and correlates well with long-term prognosis. Furthermore, decline in FEV1 is significantly reduced by stopping smoking, although it never returns to normal predicted levels (Fig 13.40). In patients with chronic bronchitis and emphysema the FEV1 is reduced and the FEV1/VC% is low, indicating expiratory flow limitation (p. 608). VC for a relaxed manoeuvre is usually greater than for a forced effort. The PEFR, although low, is a less sensitive measure of the

FIG. 13.37 CT scan demonstrating diffuse emphysema, with multiple translucent areas (arrowed).

663

severity of airways obstruction. Hyperinflation is confirmed by the increase in RV and TLC. Gas transfer is more dramatically reduced in the 'pink puffer' with a substantial element of emphysema. In the 'blue bloater' there is severe hypoxaemia, hypercapnia and sometimes secondary polycythaemia.

Management Restoration of normal function is not possible in chronic bronchitis and emphysema. The aim of therapy must therefore be to reduce disability by tackling the interrelated problems of airways obstruction, recurrent infections, breathlessness, hypoxia and poor exercise tolerance. Factors aggravating chronic bronchitis, particularly cigarette smoking, must be withdrawn.

664

Airways obstruction Conventionally the airways obstruction of chronic bronchitis and emphysema is regarded as being irreversible. However, the majority of patients show some improvement in lung function with therapy directed at relaxing bronchial smooth muscle and, although small, this improvement can have an important impact on the disability of these patients. Prognosis in patients who respond well to bronchodilators is considerably better than for those with completely fixed obstruction. The most important bronchodilator agents are the selective pYadrenergic agonists (e.g. salbutamol and terbutaline), which are best administered by inhalation. For some patients maximal bronchodilatation requires a large drug dose and may be best administered by nebulizer (e.g. salbutamol 2.5-5 mg). Inhaled atropine analogues (ipratropium and oxitropium) can be helpful but, provided optimum doses of (32 agonists are administered, confer little additional benefit. Oral theophyllines available in slow-release formulation are of marginal use in chronic bronchitis and emphysema. It had been claimed that theophylline improved respiratory muscle contractility, but subsequent studies did not confirm this effect. Patients with severe airways obstruction should have a therapeutic trial of steroids - for example oral prednisolone (30 mg daily) for a period of 2-3 weeks - provided there are no contraindications. For patients in whom oral steroids pose problems (e.g. diabetes) a period of highdose inhaled steroid is appropriate. Long-term oral steroids will not be indicated in most patients but, when prescribed, require regular assessment; dosage should seldom exceed 7.5-10 mg daily, thereby minimizing drug complications. Steroid-responsive patients should be treated with inhaled steroid. Although inhaled corticosteroids are extensively prescribed for COPD, there are no data to suggest they affect the long term decline in lung function. When given systemically, they are valuable in acute exacerbations. There is some recent evidence to suggest that mucolytic agents such as acetyl cysteine may be beneficial in reducing the numbers of exacerbations of

chronic bronchitis. Early studies of leukotriene receptor antagonists and specific phosphodiesterase inhibitors are encouraging. Infection In acute exacerbations of chronic bronchitis an infective viral or bacterial pathogen is isolated in less than 50% of cases. However, viral infections are frequently complicated by bacterial overgrowth and the majority of patients develop purulent sputum. Severe exacerbations have a mortality of up to 25%, and prompt antibacterial therapy is of the greatest importance. Common infective organisms are Strep, pneumoniae and H. influenzae, and suitable antibiotics are amoxycillin, erythromycin, clarithromycin, azithromycin, macrolide antibiotics, ciprofloxacin and the cephalosporins. For most patients long-term chemoprophylaxis is not helpful, but if intermittent treatment fails a trial of continuous rotating antibiotic therapy is appropriate. Regular immunization against influenza is sensible, and pneumococcal vaccine should be administered every 3 years. Oxygen therapy During an acute exacerbation of chronic bronchitis and emphysema, oxygen therapy is necessary to avoid death from hypoxia. In patients with hypercapnia oxygen must be given at low and controlled concentrations (usually 24% or 28% O2). A more contentious question is the value of long-term domiciliary oxygen therapy. Studies suggest that long-term controlled oxygen therapy can benefit patients with airways obstruction who have severe hypoxia and who refrain from smoking cigarettes. It is necessary to administer oxygen virtually continuously, including during sleep. Long-term follow-up of domiciliary oxygen therapy in severe COPD demonstrates that the treatment can prolong mean survival by 5 years. The administration of continuous oxygen presents considerable practical and financial difficulties. Oxygen may be supplied in cylinders, as a liquid, or, preferably, generated by an oxygen concentrator. This therapy should be reserved for patients with severe disease who are well motivated. Oxygen therapy requires careful pretreatment assessment and long-term supervision. In practice, much oxygen used by patients in their homes is for a few minutes only and the main purpose is to relieve breathlessness. Oxygen is also available in small portable cylinders, and some patients find this helpful in reducing breathlessness on exercise, improving exercise capacity, and permitting excursions from the home. Many different techniques of delivering oxygen, including transtracheal catheters, are available. Patients with severe chronic bronchitis and associated hypoxia, hypercapnia, pulmonary hypertension and cor pulmonale have profound nocturnal hypoxaemia (Fig. 13.38). This hypoxaemia is much more a feature of the 'blue bloater' than of the 'pink puffer'. Much of the hypoxia is caused by hypoventilation and irregular breathing patterns during rapid eye movement (REM) sleep, and

6 weeks will also reduce the haematocrit. Whenever possible the first-line therapy for secondary polycythaemia is to improve lung function.

FIG. 13.38 Oxygen saturation during sleep in severe chronic bronchitis Patients who are already hypoxic can markedly desaturate during sleep, particularly REM sleep (vertical bars).

is not due to obstructive sleep apnoea (p. 668). Oxygen therapy during the night (2L/min by nasal cannulae) can alleviate hypoxaemia and make hypoxic periods less severe. The return of normal breathing patterns following REM sleep is impaired by sedatives, and these should therefore be avoided in patients with chronic bronchitis. The non-sedative tricyclic drug protriptyline reduces time spent in REM sleep, and can improve nocturnal hypoxaemia in some patients. Drug therapy for breathlessness In patients with airways obstruction it is the 'pink puffers' with normal CO2 values and mild or moderate hypoxia who are most breathless. In some patients reducing ventilation with diazepam, promethazine and dihydrocodeine can reduce breathlessness, and in patients without carbon dioxide retention a careful trial of such therapy is justified when symptoms are severe. For the devastating dyspnoea which is frequently a feature of terminal respiratory failure, diamorphine is helpful whatever the cause. Secondary polycythaemia Polycythaemia produces symptoms attributable to hyperviscosity, including somnolence, lethargy, poor concentration and headache. Venesection reduces these symptoms but does not affect lung function or blood gases. Venesection also improves cerebral blood flow, pulmonary hypertension and exercise tolerance. Patients with a haematocrit of 0.60 or more should benefit, and the haematocrit should be reduced to 0.50. This can be achieved either by simple venesection, exchange transfusion with dextran or plasma, or erythropheresis. Continuous oxygen therapy for

13

Chest physiotherapy There have been few scientific studies of the long-term effect of chest physiotherapy techniques aimed at improving breathing patterns in patients with chronic bronchitis and emphysema; no benefit has been proven. Postural drainage is of undoubted benefit in patients with excessive bronchial secretions. Its effectiveness is increased by coughing or forced expiratory manoeuvres ('huffs'). Postural drainage has little place if patients do not have sputum. Vigorous physiotherapy, especially forced expiratory manoeuvres, can increase bronchoconstriction in asthma. During acute exacerbations of bronchitis and pneumonia several studies have, rather surprisingly, shown no benefit from physiotherapy. In the absence of excessive secretions physiotherapy is not justified in these patients. Mucociliary clearance is decreased and basal collapse is more common following general anaesthesia and surgery, particularly of the upper abdomen. The case for routine preoperative physiotherapy in chronic bronchitis patients is, however, not proven except for very high-risk patients. Pulmonary rehabilitation The evidence base for pulmonary rehabilitation in patients with chronic lung disease, especially COPD, is now strong. Rehabilitation can reduce breathlessness, improve quality of life and probably reduce overall requirements for health care (see Recent Advances). Assisted ventilation in chronic bronchitis Patients with severe progressive chronic bronchitis and emphysema eventually die of intractable ventilatory failure, having been breathless at rest and confined to their homes in the months before death. Their poor quality of life precludes such patients from being treated with mechanical ventilation to delay the inevitable outcome. Difficult decisions are posed by the less severely disabled, who require mechanical ventilation if they are not to die during an acute exacerbation of bronchitis. Many of these patients are difficult to wean from ventilation, particularly those with pre-existing hypercapnia and blunted ventilatory response to CO2. Before embarking on mechanical ventilation, every effort must be made to establish the background severity of respiratory disability, exercise tolerance and quality of life. For selected patients, non-invasive nasal positive-pressure ventilation is useful (p. 723). Cessation of cigarette smoking Tobacco smoke damages the bronchial tree and produces airflow limitation by a number of different actions. Smoke impairs mucociliary clearance and causes bronchial smooth muscle to contract by stimulating receptors and provoking the release of inflammatory mediators. In

665

RECENT ADVANCES IN PULMONARY REHABILITATION Patients with chronic lung disease, most commonly COPD, are greatly limited by breathlessness and become locked into a downward spiral of deteriorating performance. As a consequence of their disease, any physical exertion causes breathlessness and patients therefore reduce their activities. This immobility leads to severe cardiorespiratory deconditioning and atrophy of skeletal muscle. As a consequence of these physiological changes, secondary to reduced mobility, any exercise invokes exaggerated production of lactate and CO2, which in turn increases ventilatory requirements and breathlessness. The excessive breathlessness further reduces mobility, and so on (Fig. 13.39). The abnormalities that develop in skeletal muscle, particularly the muscles of the legs, are striking and may reflect hypoxia, hypercapnia, poor nutrition and systemic factors, as well as disuse. A number of controlled clinical trials and metaanalyses have demonstrated that pulmonary rehabilitation can reduce breathlessness, increase exercise capacity and enhance quality of life. Rehabilitation programmes typically include education about the relevant disease and its therapy, and exercises (walking, cycling, leg and arm exercises). A usual programme would consist of 1-2 hours of exercise, two to three times a week for 8-12 weeks, delivered on an outpatient basis. The mechanisms underlying the undoubted benefits that most patients experience with pulmonary rehabilitation are not clear. Patients are unable to exercise with sufficient intensity to achieve large physiological improvements in respiratory or cardiac parameters. It is likely that the benefits are due in part to physiological improvements and in part to increased confidence and control of symptoms by patients. Ongoing clinical studies are seeking to identify those patients most likely to benefit, to define the most effective methods of rehabilitation, to determine the most cost-effective way of delivering programmes, and to determine the best way to maintain benefit. However, at present pulmonary rehabilitation is arguably the most effective intervention available for many patients with moderate/ severe COPD (and who have previously stopped smoking).

666

addition, smoke increases mucus production and causes mucous gland hypertrophy. Smokers are predisposed to bronchial infection and consequent inflammation; the increase in elastase and the impairment of arantitrypsin, as well as impaired surfactant production, contribute to emphysema. It is therefore not surprising that chronic bronchitis and emphysema are found in 15% of middleaged males who smoke moderately or heavily but are rare

in non-smokers, and that deaths from bronchitis increase with the amount smoked. If patients with chronic bronchitis and emphysema stop smoking, the rate of decline in pulmonary function is reduced to that of non-smokers (Fig. 13.40). Indeed, if patients stop smoking early in their disease there is an improvement in pulmonary function. However severe the disease, stopping smoking will reduce cough. About half of regular long-term smokers die from smoking-related diseases. In middle age, smoking is responsible for 30% of all deaths in western countries: a tragic and avoidable epidemic. Deaths are from many

FIG. 13.39 The downward spiral of progressive disability in patients with chronic lung disease

FIG. 13.40 Airflow limitation and smoking FEV, declines with age. In smokers this decline is more marked, particularly in 'susceptible' smokers. Such patients, therefore, become breathless in middle or old age. When smokers stop the habit, the rate of decline in FEV! is markedly slowed. (From classic study of Fletcher C M, Peto R 1997 Br Med J; 1:1645.)

RECENT ADVANCES IN NON-INVASIVE POSITIVE-PRESSURE VENTILATION (NIPPV) Many patients with chronic ventilatory failure due to stable neuromuscular disease, for example kyphoscoliosis or weakness from past poliomyelitis, can be effectively treated by intermittent long-term mechanical ventilation, usually within their own homes. In the past such ventilatory assistance was provided by negativepressure devices, including tank ventilators (the 'iron lung') and cuirasses, or by positive-pressure ventilators via a tracheostomy. Now most of these patients are managed with nocturnal non-invasive positive-pressure ventilation, via a nasal mask. Nocturnal NIPPV is often highly successful. The technique is well accepted by patients, enhances their quality of life and improves ventilatory failure. Often patients lead normal active lives by day while receiving NIPPV by night. The key to success is that NIPPV controls the severe hypoventilation and hypercapnic acidosis that would otherwise occur during sleep, and this restores normal ventilatory control by day. The end result is that for many patients chronic respiratory failure is substantially or completely reversed, and this improvement can be sustained for many years. The success of NIPPV in the management of selected patients with chronic ventilatory failure has led to its use in other clinical settings. Increasingly, NIPPV is being used in the management of acute exacerbations of chronic obstructive pulmonary disease (COPD). Controlled clinical trials - performed in the UK, France and North America - have demonstrated that NIPPV can improve blood gases, reduce the need for intubation, and reduce mortality in patients admitted with severe acute COPD. NIPPV is a flexible treatment: hours spent on the ventilator can be varied from day to day and the therapy can be easily stopped and just as easily restarted. In those patients with COPD in whom it is felt that intubation and mechanical ventilation on the intensive care unit is not appropriate, because of age or severity of underlying disease, NIPPV should be considered: it may be life-saving. In those patients in whom intubation would be appropriate, NIPPV should also be considered as the first treatment option because it may avoid the need for intubation and ICU care. The success of NIPPV has resulted in increasing numbers of patients being treated by the technique, and this is a trend that is likely to continue.

TABLE 13.31 Smoking-related diseases • Cancer (lung, mouth, pharynx, larynx, oesophagus, bladder, renal pelvis, pancreas) • Chronic bronchitis and emphysema • Coronary heart disease • Peripheral vascular disease • Cerebrovascular disease • Gastric and duodenal ulceration • Crohn's disease • Osteoporosis • Fetal and neonatal death • Childhood disorders from passive smoking (respiratory infections, 'glue' ear)

13

cian (and other medical staff) is effective and is most likely to work when the patient is unwell. In patients who are determined to stop smoking, nicotine replacement therapy by gum, nasal spray or transdermal patch can be helpful. Nicotine replacement therapy approximately doubles the chance of a patient stopping smoking. Clinical trials have also demonstrated the efficacy of bupropion, although its mechanism of action is not clear. Increased appreciation that the key problem is nicotine addiction will hopefully lead to better nicotine replacement therapies becoming available. In the UK nicotine replacement therapy is available on prescription, and services to support those seeking to stop are being strengthened. If patients are able to stop smoking, the health risks, particularly death from lung cancer, are greatly reduced (see Fig. 13.59, p. 701). Cigarette smokers who change to smoking a pipe or cigars inhale the smoke and this change is therefore not helpful. If patients are unable to stop smoking completely, they should smoke as few cigarettes as possible, change to lowtar brands, reduce the number of puffs, and smoke less of each cigarette. Lung surgery Surgical removal of bullae can help selected patients with severe bullous emphysema, particularly if they are compressing relatively normal adjacent lung. Selected patients with COPD can also be treated by lung transplantation. Lung volume reduction surgery is increasingly being considered in patients with severe emphysema (see Recent Advances).

FURTHER READING ON CHRONIC BRONCHITIS AND EMPHYSEMA

causes (Table 13.31), particularly coronary artery disease and lung cancer. Worldwide, four million people die each year from smoking. Patients with chronic bronchitis and emphysema have smoked for many years and have great difficulty in abstaining. Stopping smoking is particularly difficult if a spouse or cohabitee smokes. The firm advice of the patient's physi-

American Thoracic Society 1995 Standards for the diagnosis and care of patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 152:577-5120. American Thoracic Society 1996 Cigarette smoking and health. Am J Respir Crit Care Med 153:861-865. Barnes P J 2000 Chronic obstructive pulmonary disease. N Engl J Med 343 (4): 269-280.

667

RECENT ADVANCES IN LUNG VOLUME

REDUCTION SURGERY (LVRS) Some patients with severe emphysema can benefit from surgery that reduces the volume of the lungs. These patients are in addition to those who routinely benefit from removal of large bullae by bullectomy. Volume reduction surgery, whereby 20-30% of each lung is resected, has been carried out in many countries around the world over recent years. There have been few controlled clinical trials, but the data suggest that LVRS may well help carefully selected patients with disabling emphysema. If they are not suitable for lung transplantation these patients have few therapeutic strategies available. Furthermore, if patients who might otherwise be transplanted are treated with volume reduction surgery, scarce donor lungs are available for those with other fatal conditions. Lung transplantation for advanced COPD is, in most instances, not practicable because many of the patients are too old and infirm to be considered for the operation, particularly with the scarcity of donor organs. Surgical resection of parts of the lungs increases lung elastic recoil, which reduces the dynamic expiratory collapse of airways and therefore reduces expiratory airflow limitation. The reduction of lung volume lengthens the diaphragm, improves its mechanical advantage, and enables a given muscle tension to generate more transdiaphragmatic pressure; all of these changes improve the capacity of the diaphragm to achieve inspiration. The resection of badly diseased lung also decompresses relatively good-quality lung, allowing its function to improve. Volume reduction surgery therefore enhances expiratory flow, increases inspiratory capacity and improves the ventilation of relatively normal lung; emphysematous patients thereby become less breathless. To be considered for volume reduction surgery patients must have severe COPD with emphysema, be severely breathless, hyperinflated, with an FEV! of usually less than 0.7 L, and often requiring oxygen therapy. In major centres undertaking this surgery approximately 20% of referred patients are suitable, and in expert hands the results are encouraging with a significant improvement in FEV l5 arterial blood gases, breathlessness and quality of life. Improvements are often sustained for many months; long-term follow-up data show that benefits persist for about 3 years. The operative and postoperative mortality can be low. The results of further controlled clinical trials on mortality are awaited with interest, but it is likely that volume reduction surgery will be increasingly used to help symptoms in severely breathless patients with emphysema.

1

668

MCQ13.15

British Thoracic Society Statement 2001 Pulmonary rehabilitation. Thorax 56: 827-834. Bronchard L 2000 Non-invasive ventilation for acute exacerbations of COPD: a new standard of care. Thorax 55: 817-818. BTS guidelines for the management of chronic obstructive pulmonary disease 1997. Thorax, 52 (Suppl 5). Cooper J D, Lefrak S S 1996 Is volume reduction surgery appropriate in the treatment of emphysema? Yes. Am J Respir Crit Care Med 153:1201-1204. Geddes et al. 2000 Effect of lung-volume reduction surgery in patients with severe emphysema. N Engl J Med 343:239-245.

RESPIRATION DURING SLEEP SLEEP APNOEA SYNDROMES O Apnoea is defined as no gas flow at the nose or mouth for 10 seconds, and a sleep apnoea syndrome is conventionally, and rather arbitrarily, diagnosed if more than 30 such episodes occur during the night. Apnoea may be central, when there is no airflow and no chest wall movement; obstructive, when there is no airflow despite chest wall movement because of upper airways obstruction; or may result from both mechanisms. Obstructive sleep apnoeo Obstructive sleep apnoea is much more common than central apnoea, occurring perhaps to some degree in up to 1 % of the adult male population. The typical features are hypersomnolence by day and upper airways obstruction during sleep. Many (but not all) patients are obese and have thick necks; hypertension is common. In severe cases there is chronic alveolar hypoventilation at night, daytime hypoxia, secondary polycythaemia and cor pulmonale. Patients often have loud, persistent snoring, and the frequent arousals from apnoeic episodes greatly disturb sleep (although the patient may be unaware of this). Patients fall asleep by day and may have road and other accidents. There is difficulty in concentrating, particularly in the morning, intellectual impairment, personality change and irritability. Abnormal limb and body movements during arousals following apnoea (thrashing around in bed), sexual dysfunction and morning headaches are important features. During apnoea there are cardiovascular disturbances, particularly bradycardia. In some patients the upper airway is abnormally narrow (e.g. hypertrophy of tonsils and adenoids), but in most no

SUMMARY 8 Features of obstructive sleep apnoea Snoring Disturbed sleep Witnessed apnoeas Cardiac arrhythmias Excessive daytime sleepiness Morning headache

Intellectual deterioration Systemic hypertension Pulmonary hypertension Right heart failure Obesity/thick neck Small oropharynx

anatomical abnormality is found. In normals the upper airway narrows during sleep and this is more pronounced in individuals with thick necks. During obstruction the posterior and lateral walls of the pharynx collapse, occluding the oropharynx. The collapse of the upper airway is probably due to a failure to activate the upper airway musculature sufficiently during inspiration, and the negative pressures generated in the upper airway by the contraction of the respiratory muscles suck in the soft tissues of the oropharynx (Fig. 13.41). This negative pressure will be greater if the upper airway is narrowed, and nasal obstruction can be a contributory factor. Central sleep apnoea In central apnoea the patients snore less, are seldom obese, and complain of night-time wakening. The central drive to breathing is abnormal and inspiration may fail to be initiated. The syndrome may be primary or secondary to organic lesions of the brainstem. To document sleep apnoea, to determine whether it is central or obstructive and to assess the severity of nocturnal hypoxaemia, a formal sleep study in a specialist laboratory is required (Fig. 13.42). In patients with obvious obstructive sleep apnoea the assessment of nocturnal

Pathogenetic factors

Primary events

hypoxaemia by oximetry may be all that is necessary. However, it is possible to have severe symptoms due to disturbed sleep (arousal is due to the great efforts of breathing during obstruction) without significant hypoxaemia.

13

Sleep in other respiratory disorders Patients with airways obstruction and hypercapnia may become profoundly hypoxic at night as a result of hypoventilation (Fig. 13.39, p. 666). Nocturnal hypoxaemia is also a clinical problem in patients with advanced kyphoscoliosis or severe respiratory muscle weakness (p. 718). Sleepdisordered breathing in these conditions may cause daytime sleepiness and headache on waking.

Treatment of apnoea syndromes In the management of obstructive sleep apnoea (OSA) weight reduction is helpful but difficult to achieve. Nasal

Pathophysiological and clinical

consequences

FIG. 13.41 The mechanism of obstructive sleep apnoea (OSA) In severe cases, because sleep stops ventilation, sleep is grossly disturbed and daytime somnolence is very striking. (From Strading J R, Philipson E A Breathing disorders during sleep. Q J Med 1986; (new series) 58: 3-18).

FIG. 13.42 Sleep study in obstructive sleep apnoea The record shows a period when ventilation is obstructed and there is no airflow (V). At this time the diaphragm continues to contract (EMGdi); oesophageal (Pes) and gastric (Pga) pressure swings occur, and there is ribcage and abdominal movement (RC and AB). As obstruction continues arterial oxygen saturation (So2) falls. Eventually the activity of the upper airway muscles, including genioglossus (EMGge), increases, the upper airway becomes patent, and airflow is restored. Patients can have hundreds of similar episodes during the night. (From Onal et al Am Rev Respir Dis 1982; 125:167-174.)

669

continuous positive airway pressure (CPAP) is the standard therapy for symptomatic obstructive sleep apnoea, is very effective and remarkably well tolerated. A therapeutic trial of CPAP is often a useful diagnostic strategy: patients with obstructive sleep apnoea usually report dramatic improvement of their symptoms. CPAP has been shown conclusively to improve sleepiness in patients with OSA. Palatopharyngoplasty (PPP) - surgical reconstruction of the pharynx - can also be effective, but the site of obstruction may not be relieved in some patients and the long-term effects of surgery are not yet known. Anterior mandibular repositioning devices which can be fixed or titratable, may also be of use in some patients who are unable to tolerate CPAP. These are generally well tolerated and can significantly improve sleep quality in patients with mild-moderate OSA, probably by directly increasing the size of the upper airway. Drugs may sometimes be helpful, particularly protriptyline (a non-sedative tricyclic antidepressant); this reduces REM sleep, during which, as a consequence of reduced activity of upper airways musculature, obstruction is most likely. For many patients weight reduction and perhaps protriptyline are worthy of trial before considering CPAP or PPP. Troublesome snoring, without desaturation, can be treated by minor palatal surgery in selected cases, but it is important to document that the snoring really is a problem.

FURTHER READING ON RESPIRATION DURING SLEEP Engleman H M, Martin S E, Kingshott R N, MacKay T W, Deary I J, Douglas N J 1998 Randomized placebo controlled trial of daytime function after continuous positive airway pressure (CPAP) therapy for the sleep apnoea/hypopnoea syndrome. Thorax 53:341-345. McNamara S G, Grunstein R R, Sullivan C E 1993 Obstructive sleep apnoea. Thorax 48:754-764.

RESPIRATORY FAILURE 1

Definition There is no precise clinical definition of respiratory failure; the diagnosis rests on the interpretation of arterial blood gas measurements. A patient can be said to be in respiratory failure if the arterial oxygen tension (-Pao2) falls below 8.0 kPa/60 mmHg (normal range 11.3-13.3kPa or 85-100mmHg) or if the arterial carbon dioxide tension (Paco2) rises above 6.6kPa/50mmHg (normal range 4.6-6.0kPa or 35-45 mmHg), when the subject is at sea level, awake and breathing air. The condition may be acute or chronic; if the patient has arterial hypoxaemia with a

1 MCQ 13.16 670

TABLE 13.32 Causes of types I and II respiratory failure Type 1 (hypoxaemic) Pneumonia Emphysema Acute asthma Pulmonary oedema Pulmonary fibrosis Chronic bronchitis Pulmonary vascular disease Miscellaneous: lymphangitis, radiation pneumonitis etc. Type II (ventilatory) CNS

Trauma Cerebral tumour Raised intracranial pressure Drugs Hypoventilation syndromes Neuromuscular Cervical cord lesion Guillain-Barre syndrome Motor neuron disease Poliomyelitis Muscular dystrophies and myopathies Botulism Myasthenia gravis Muscle relaxant drugs Organophosphorus poisoning Status epilepticus

Thoracic cage and pleura Crushed chest Kyphoscoliosis Extensive thoracic surgery (thoracopiasty) Ankylosing spondylitis (rare) Lung and airways Severe acute asthma Pneumonia Upper airway obstruction, including obstructive sleep apnoea Chronic airflow limitation: chronic bronchitis, emphysema, bronchiectasis

normal or low Paco2 then he or she is said to have type I respiratory failure; and if the Paco2 is elevated, type II respiratory failure. Arterial hypoxaemia does not always imply respiratory failure: a low inspired O2 at altitude or an anatomical right-to-left shunt, as in congenital heart disease or arteriovenous malformation, can cause hypoxaemia despite normal lung function. The causes of types I and II respiratory failure are shown in Table 13.32. In general, the main disturbance in type I failure is ventilation-perfusion mismatch, and in type II it is inadequate ventilation. Chronic respiratory failure in the UK is most commonly due to chronic bronchitis and emphysema (p. 661). Measurement of the alveolar-arterial (A-a) gradient is useful in hypoxic patients, and may help to decide, for instance, whether the patient's hypoxaemia is secondary to central hypoventilation or to intrinsic lung disease. By calculating the A-a gradient the day-by-day progress of patients can be assessed even when they are being treated with varying concentrations of inspired oxygen. The A-a gradient indicates the contribution of venous admixture (i.e. true shunt or V/Qabnormality) to hypoxaemia. It can be calculated by using the concept of 'ideal alveolar air'

PAO2 = FiO2 -

Paco2 0.8

where FiO2 = inspired Po2 (20kPa in room air) and 0.8 = respiratory exchange ratio. Thus the normal alveolar oxygen tension is approximately 14 kPa, the normal arterial oxygen tension approximately 12.5 kPa, and the normal A-a gradient is therefore 1-2 kPa (see Ch. 14, p. 736). In the management of the hypoxic patient it should be remembered that tissue O2 delivery depends not just on Pao2 but also on the haemoglobin concentration and cardiac output. The most useful index of tissue oxygenation is the mixed venous (pulmonary arterial)/arterial O2 content difference. The normal oxygen combining capacity is 1.4mL O2/g haemoglobin (approximately 0.06mmol/g). The oxygen content of blood is the oxygen combining capacity x haemoglobin concentration x saturation (plus a small amount of dissolved oxygen in plasma, which is negligible at normal atmospheric pressure). Assuming the haemoglobin concentration is 15g/100mL, and the oxygen content of arterial blood is 1.4 x 15 x 0.95 = 20.58 mL/ 100 mL, and the oxygen content of mixed venous blood is 1.4 x 15 x 0.75 = 15.75, the arterial/mixed venous oxygen content difference is 4.83mL/100mL. If oxygen delivery is reduced (e.g. reduced cardiac output) the arterial/venous content difference widens. Type II hypercapnic ventilatory failure represents inadequacy of the respiratory muscle 'pump'. The concept of 'pump' failure leading to hypercapnia is a useful one, focusing attention on the importance of CNS output, neuromuscular function and chest wall movement in the maintenance of adequate ventilation (p. 718), although severe impairment of gas exchange can contribute to an elevated CO2.

Clinical features Acute hypoxaemia or hypercapnia is less well tolerated than chronic gradual alteration in blood gases. Virtually all patients with respiratory failure, with the exception of some with CNS dysfunction (e.g. drug overdose), are breathless. Acute hypoxaemia causes restlessness, confusion and sweating, with a tachycardia, poor peripheral perfusion and central cyanosis. Patients with acute severe hypercapnic ventilatory failure are breathless and cyanosed, and in addition may have confusion, flapping tremor of the hands, warm peripheries, bounding pulse and occasionally papilloedema. The mental state is probably the best clinical index to follow, as increasing confusion and restlessness often parallel a rising Paco2. When patients (for instance those with advanced chronic bronchitis) progress to severe chronic respiratory failure, with a compensated respiratory acidosis, symptoms may also include sleep disturbance with early morning headache and personality change. Such patients develop polycythaemia, pulmonary hypertension and cor pulmonale. Chronic hypercapnic respiratory failure can occur

despite normal lung function. Sleep apnoea syndromes, for example (p. 668), may be severe enough to produce pulmonary hypertension and cor pulmonale. Patients with neuromuscular disorders (e.g. myasthenia gravis or the Guillain-Barre syndrome) may also have life-threatening ventilatory failure despite normal lungs (p. 718).

13

Management Acufe respiratory failure The general principles for the management of acute respiratory failure are those of any other respiratory emergency. The airway should be kept clear of sputum and protected. Oxygen should be given in sufficient concentration to ensure a Pao2 of 8kPa (60mmHg). This can be delivered via a Venturi mask, which will give a fixed inspired oxygen concentration of 24, 28 or 35%, or a Hudson or MC mask which will deliver variable, but higher, inspired oxygen concentrations. In practice it is difficult to administer more than 40% O2 by conventional face masks. Higher concentrations can be given using a tight-fitting face mask; this also permits the application of continuous positive airway pressure (CPAP), which further increases Pao2. If an adequate Pao2 can only be achieved by a high FiO2 or CPAP, the patient is oxygen dependent and requires careful observation. In all patients with hypoxia the non-invasive monitoring of SaO2 is useful; however, in patients with ventilatory failure it is also crucial to monitor arterial pH and Pco2. If identifiable, the underlying cause should be treated: for example, antibiotics for acute bacterial pneumonias, or bronchodilators and corticosteroids for acute asthma. If the patient's condition is worsening, with deepening cyanosis, disturbance of consciousness, a rising Paco2, and progressive acidosis, then a decision whether to intubate and employ assisted mechanical ventilation must be made. This decision is based on a number of considerations, including the previous exercise capacity, the possibility of treating the underlying cause, the overall clinical state of the patient and serial arterial blood gas measurements. Chronic respiratory failure Patients with chronic respiratory disease who develop additional acute problems require a rather different pattern of management. Patients with long-standing hypoxaemia and hypercapnia are dependent on hypoxaemia to maintain ventilatory drive. High concentrations of inspired oxygen reduce respiratory drive and ventilation, resulting in a rise in Paco2 and a worsening respiratory acidosis. Oxygen must therefore be given in a controlled fashion, starting with an inspired concentration of 24% and with repeated monitoring of arterial blood gases. It should be emphasized that it is not necessary to raise the Pao2 into the physiological range: a Pao2 of more than 6 kPa (45 mmHg) is usually sufficient to maintain the patient in a reasonable clinical state. Although it is correct to exercise caution when administering oxygen to

671

CASE STUDY 13.5 SEVERE VENTILATORY FAILURE IN A 71-YEAR-OLD WOMAN A 71-year-old woman, originally from the Caribbean, was referred to the Accident & Emergency department by her family doctor because of severe leg oedema, abdominal swelling and breathlessness on exertion. The patient had been unwell for 2 months, but previously she had been able to walk to the local shops and climb the stairs in her home. Hypertension had been diagnosed 6 years previously, and she was taking lisinopril 5 mg daily; more recently she had been started on frusemide 40 mg daily for her oedema. She had never smoked cigarettes and gave no history of respiratory problems. A married woman, she lived with her husband who was well. One week prior to admission, her GP had arranged an ultrasound scan which had shown substantial ascites. On examination, the patient was comfortable at rest. There was no fever. The blood pressure was 125/70, pulse 104, JVP elevated 8cm with prominent systolic venous waves present, the heart apex beat was displaced laterally and was forceful. The respiratory rate was 16, percussion and breath sounds were reduced at both bases, and there were a few basal crackles. There was severe pitting oedema of both legs, extending up to the anterior abdominal wall. Obvious ascites was present, but it was not possible to feel the liver. The patient was 5 ft tall and weighed 82 kg. The oxygen saturation measured by oximetry breathing air was 78%. The chest X-ray showed a large heart and clear lung fields (Case Fig. 13.5.1). The ECG demonstrated sinus rhythm, tall P waves and right axis deviation. The Hb was 13.6, WBC 6.2, urea and electrolytes normal. An echocardiogram showed grossly dilated right heart chambers, severe tricuspid regurgitation, and pulmonary artery systolic pressure of

672

CASE FIG. 13.5.1 The cardiac silhouette is enlarged and the pulmonary arteries are prominent. The lung fields are clear, although there are small bilateral pleural effusions.

40 mmHg. There was pulmonary regurgitation, a dilated IVC, LVH and good left ventricular systolic function.

Questions

1. What is the differential diagnosis? 2. What further investigations are required? Discussion The most striking finding in this patient is the severe right heart failure. This could be due to primary cardiac disease, but the echocardiogram demonstrated no left heart problems: indeed, left ventricular function was good. Pericardial disease causing constriction can lead to high rightsided pressures, ascites and oedema, but the ECG showed no evidence of

a pericardial effusion or thickening. The right heart failure could be secondary to lung disease, but there was no history of respiratory problems, few abnormal lung signs and a clear chest X-ray. That the patient has never smoked is clearly important. One possibility could be that the pulmonary hypertension and . right heart failure were secondary to pulmonary vascular disease or pulmonary emboli. However, when pulmonary emboli cause such severe hypoxia (the oxygen saturation was 78%) and right heart failure, more severe breathlessness and circulatory disturbance might be expected. Urgent next investigations would include arterial blood gas analysis, lung function tests, and a ventilation/perfusion lung scan. The latter showed no substantial perfusion defects and no ventilation/perfusion mismatch, and was considered to indicate a low

probability of pulmonary emboli. The arterial blood gases, breathing air, showed a pH of 7.38, pco2 8.86, po2 5.85 and HCO3 39.1. On breathing 2 L of oxygen, the Po2 increased to 15.5 and the Pco2 rose to 10.4. The blood gas results confirmed severe hypoxia; it is striking that cyanosis was not noticed at the bedside, and illustrates how difficult this physical sign can be in dark-skinned patients. The results demonstrate type 2 respiratory failure (ventilatory failure) with a fully compensated respiratory acidosis. Despite the low Po2, the gas exchange function of the lung is clearly good, as evidenced by the high Po2 with 2 L of added oxygen. Lung function tests showed a predominantly restrictive ventilatory defect, with an FEVi l.OL (predicted 1.4), vital capacity 1.4 (1.8), gas transfer (TCO) 5.1 (5.8), transfer coefficient (KCO) 2.1 (2.0), TLC 2.8 (3.9), RV 1.7 (1.8). The flow-volume loop did not suggest airways obstruction. The investigations therefore demonstrated that the patient had severe hypercapnic ventilatory failure, pulmonary hypertension and right heart failure, with relatively normal lungs (on chest X-ray and in terms of gas transfer). Question 3. What is the most likely diagnosis?

Discussion There are many causes of chronic hypercapnic ventilatory failure (see Table 13.58). Most commonly there are severe abnormalities of lung function (e.g. COPD) or chest X-ray (e.g. kyphoscoliosis). There are relatively few causes of chronic ventilatory failure in the presence of a clear chest X-ray and relatively normal gas transfer. One such cause is respiratory muscle weakness (as in motor neuron disease), but the absence of symptoms and signs of neuromuscular disease makes this diagnosis unlikely in this patient. The most likely diagnosis is sleep apnoea. Such apnoea can be central, but in the majority of patients it is obstructive in nature (OSA, see p. 668). The patient had normal inspiratory muscle strength (as measured by maximum static inspiratory pressures - PImax and maximum sniff nasal pressure (SNIP)). A sleep study confirmed obstructive sleep apnoea (see p. 668). Subsequent questioning of the patient revealed that she had been sleepy during the day for many months, and that her family had thought her to be less mentally alert in recent years; she was also a loud snorer. The patient's ventilatory failure was initially treated with non-invasive ventilation and the blood gases rapidly improved. After a few weeks of nocturnal ventilation her obstructive sleep apnoea was treated with night-time nasal CPAP of

hypercapnic patients, it should be remembered that severe hypoxaemia - Pao2 less than 5kPa (38mmHg) - is lifethreatening and must be relieved, and hypoxaemic patients without CO2 retention (a common situation in fibrotic lung disease, for example) can receive high inspired oxygen concentrations without developing hypercapnia. If, with the proper use of oxygen supplements, the Paco2 continues to rise, and particularly if the respiratory acidosis worsens, a respiratory stimulant such as doxapram may, in the short term, occasionally be helpful. This may allow the use of a higher inspired oxygen concentration without precipitating a rise in Paco2. However, some patients will

15 cmH2O. Symptomatically she was much improved. With CPAP therapy and treatment with diuretics, her oedema resolved. Two weeks after starting CPAP therapy, daytime blood gas analysis, breathing air, showed a Po2 of 9.4 and a Pco2 of 6.1. The patient was discharged from hospital feeling well. She and her relatives considered that she was more alert, with much improved exercise tolerance and less breathlessness. With CPAP she was able to sleep well. The patient has remained well 3 years after admission, and when checked recently her daytime blood gases were Po211.4, Pco2 5.22 and HCO3 23.6. She has no signs of right-sided heart failure, the heart has reduced in size, and the ECG demonstrates a reduction in her pulmonary hypertension. Obstructive sleep apnoea (OSA) is discussed on page 668. Severe OSA is a recognized cause of ventilatory failure. In some patients there are additional contributory factors (e.g. severe obesity, COPD). To restore normal ventilatory control a period of noninvasive ventilation is often helpful and, once blood gases are normalized, ongoing treatment with CPAP is highly effective and known to improve quality of life. Hypertension is commonly associated with OSA (as in this patient) and may be more easily controlled when patients no longer have sleepdisordered breathing.

continue to deteriorate, and the point may come when a decision will have to be made about whether to offer assisted mechanical ventilation. In this group of patients, above all, their background in terms of respiratory function, exercise tolerance, and when they were last reasonably well, must be taken into consideration. Discussions should involve the patient's family and, very often, the patient. For selected patients with ventilatory failure, including those with chronic bronchitis, the technique of noninvasive intermittent nasal positive pressure ventilation may obviate the need for intubation and reduce mortality.

13

673

LUNG TRANSPLANTATION For patients with end stage pulmonary disease, lung transplantation is now a realistic therapeutic option. The indications are listed in Table 13.33, and not all patients are suitable (Table 13.34). About 8000 procedures have been carried out to date. There is approximately a 70% one year survival rate, 54% 3 year survival rate and 42% 5 year survival rate with a median survival of 3.7 years; the procedure therefore offers longer survival and an improved quality of life. It is limited by the availability of donor organs and complications are common: these include primary graft failure, infection, airway complications and acute and chronic rejection; the chronic rejection is usually manifest as bronchiolitis obliterans. The technique of lung transplantation can be either single lung (e.g. for COPD or pulmonary fibrosis), sequential single lung, heart lung transplantation or lobar donation. The lobar donation technique can be either from living related donors or can be from a cadaveric source. As with any transplants the operation must be followed by treatment with irnmunosuppressive agents to avoid rejection, and prompt treatment of opportunistic infections. The commonly used drugs for immunosuppression include

TABLE 13.33 Clinical indications for lung transplantation End stage chronic obstructive pulmonary disease Idiopathic pulmonary fibrosis Cystic fibrosis Primary pulmonary hypertension Eisen monger's syndrome Sarcoidosis Collagen vascular lung disease Alveolar cell carcinoma

ciclosporin, tacrolimus, azathioprine, mycophenolate and corticosteroids.

FURTHER READING ON LUNG TRANSPLANTATION Arcasoy S M, Kotloff R M 1999 Lung transplantation. N Engl J Med 340:1081-1091. Maurer J, Frost A, Estenne M et al. 1999 International Guidelines for Selection of Lung Transplant Candidates. J Heart Lung Trans 17:70-709.

PULMONARY HYPERTENSION The normal resting pulmonary artery pressure is approximately 25/8 mmHg; values greater than 30/15mmHg indicate pulmonary hypertension (Table 13.35). More than one factor may operate in any particular patient: for example, patients with ventricular septal defects have increased pulmonary blood flow but eventually develop secondary obstructive and obliterative changes.

Clinical features Pulmonary hypertension leads to right ventricular hypertrophy and failure, low cardiac output, syncope, fatigue and breathlessness. Physical signs include a prominent 'a' wave in the jugular venous pulse, right ventricular hypertrophy and a loud pulmonary sound. The ECG confirms right atrial and right ventricular hypertrophy. The chest X-ray shows large central pulmonary arteries. Patients may also have signs of the particular underlying cause of pulmonary hypertension (e.g. airways obstruction, Fig. 13.43). The most important causes of pulmonary hypertension are listed in Table 13.36.

TABLE 13,35 Classification of pulmonary hypertension TABLE 13.34 Suitability for lung transplantation • Severe and progressive disease (but not moribund) with a poor prognosis (less than 2 years) • Age normally less than 50 years for heart-lung transplantation, less than 60 years for single lung transplantation • No significant renal, hepatic or other progressive systemic disease (no significant coronary artery disease in patients undergoing lung transplantation) • Psychologically stable

1 MCQ 13.17

674

Passive pulmonary hypertension Left heart failure Mitral stenosis LVF Hyperkinetic pulmonary hypertension Increased pulmonary blood flow ASD VSD Patent ductus

Increased pulmonary vascular resistance Vasoconstrictive e.g. response to hypoxia Obstructive Pulmonary emboli Veno-occlusive disease Obliterative Polyarteritis nodosa Systemic lupus erythematosus Systemic sclerosis Schistosomiasis Chronic bronchitis and emphysema Primary pulmonary hypertension

FIG. 13.43 Severe chronic bronchitis with marked hypoxaemia and pulmonary hypertension In keeping with airways obstruction the lungs are of large volume. The main pulmonary artery (see below the aortic knuckle) is massively enlarged, as are the branches of the left and right pulmonary arteries.

TABLE 13.36 important causes of pulmonary hypertension Diffuse lung disease Recurrent pulmonary embolism (p. 682) Chest wall abnormalities (p. 722) Sleep apnoea syndromes (p. 668) Mitral stenosis and congenital heart disease Primary pulmonary hypertension (Ch. 12)

monary veno-occlusive disease. Both conditions are rare. In the differential diagnosis it is important to consider silent recurrent thromboembolic disease, systemic lupus erythematosus and lupus-like syndromes (including the antiphospholipid syndrome). Primary pulmonary hypertension is of unknown cause and is seen most often in young women. Pulmonary function shows hypoxia due to ventilation-perfusion mismatch, and reduced gas transfer. Patients become increasingly breathless, have syncopal episodes, and develop right ventricular hypertrophy and eventually right ventricular failure. There is peripheral cyanosis, with a small-volume peripheral pulse and cold, blue hands. Central cyanosis is a late development. The jugular venous pulse has a giant 'a' wave and on auscultation there is a right atrial gallop, a tricuspid regurgitant murmur, a pulmonary systolic ejection click, a closely split second sound with a loud pulmonary component and a murmur (Graham Steell) of pulmonary regurgitation. The chest X-ray shows a dilated pulmonary artery. The signs are similar in pulmonary venoocclusive disease, with the important difference that there is pulmonary oedema and therefore more severe hypoxia as well as shadowing on the chest X-ray. The diagnosis of primary pulmonary hypertension is by exclusion of other causes of pulmonary hypertension, and the diagnosis of veno-occlusive disease requires a lung biopsy. The prognosis is poor, with a mean survival from diagnosis of about 3 years. In primary pulmonary hypertension pulmonary vasodilators (e.g. calcium antagonists, prostacycline) may be of value; some authorities also favour anticoagulation. In veno-occlusive disease anticoagulants and possibly azathioprine may be helpful. Patients with primary pulmonary hypertension can be successfully treated by heart-lung transplantation.

13

PULMONARY THROMBOEMBOLISM 1 Pulmonary hypertension due to diffuse lung disease Chronic bronchitis is the commonest cause of pulmonary hypertension, but it also occurs in severe tuberculosis, bronchiectasis, cystic fibrosis, cryptogenic fibrosing alveolitis and other respiratory diseases causing widespread lung destruction and hypoxaemia. Pulmonary hypertension is due to destruction of the capillary bed, a secondary arteritis, and also hypoxic vasoconstriction. In patients with airways obstruction continuous oxygen therapy can reduce pulmonary arterial hypertension and improve prognosis. Primary pulmonary hypertension This diagnosis includes classic primary pulmonary hypertension, in which the pulmonary arteries are diseased, and also a disorder localized to the pulmonary veins, pul-

Tumour, fat, amniotic fluid, parasites, air and injected material can all embolize to the lung, but most emboli are from venous thrombosis, particularly in the lower limbs and pelvis (p. 604). Although massive pulmonary embolism is an important cause of death, few patients with such large emboli survive to reach hospital. Most commonly the clinical picture is of less serious pulmonary emboli and pulmonary infarction. There is often a recent history of illness, trauma, anaesthesia or other factors known to predispose to deep venous thrombosis.

PULMONARY EMBOLISM AND INFARCTION Pulmonary infarction follows embolization to the peripheral branches of the pulmonary arteries. Emboli are

675

CASE STUDY 13.6 BREATHLESSNESS AND COUGH IN A 64-YEAR-OLD MAN WITH CHRONIC LYMPHATIC LEUKAEMIA A 64-year-old male retired civil servant was seen in the outpatients' department having been referred by his GP because of cough and breathlessness. The breathlessness had started 3 months previously and had been progressive, such that he was now short of breath on minimal exertion and his exercise tolerance was reduced to approximately 100 yards on the flat at a slow pace. During the period over which the breathlessness had developed, there had been several occasions when there had been sharp deteriorations. The cough had started 2 months prior to the consultation, and was dry and particularly troublesome on exercise and at night. There had been occasional episodes of wheezing. On three occasions, after exercise, he had felt faint and on one occasion he had blacked out. Over a 3-month period the patient had lost about one and a half stones in weight. Shortly before the onset of breathlessness he had been investigated for a lymphocytosis and, following haematological investigations, including a bone marrow examination, he had been diagnosed as having early asymptomatic B-cell chronic lymphatic leukaemia. The total white cell count was 21000, with normal haemaglobin and platelets. The patient was a non-smoker, drank little alcohol, had no relevant past medical history and was taking no medication. Both his father and his only brother had died in their 60s of myocardial infarction. On examination the patient was comfortable at rest; blood pressure 140/80, heart rate 108, and the pulmonary second sound was loud. The chest was clinically clear, with no wheeze. The chest X-ray was normal, except for some prominence of the proximal pulmonary arteries.

676

Questions 1. What could explain Ms symptoms? 2. What investigations would be helpful?

Discussion Possibilities to be considered as explanations for the patient's symptoms would include his lymphoma, a complicating opportunistic infection, asthma, left heart failure as a consequence of cardiac disease, and pulmonary embolic disease. Lymphoma can involve the lung, but usually in patients with extensive and aggressive disease. Lymphomatous infiltration can cause a restrictive ventilatory defect and severe breathlessness. Although the shortness of breath may be greater than expected from the chest X-ray (as is often the case in diseases that cause lymphatic infiltration), the X-ray is nevertheless invariably obviously abnormal. Opportunistic pulmonary infection should be considered, but the long history, the clear chest X-ray and the absence of fever make this unlikely. Breathlessness in a patient with a normal chest X-ray could be due to asthma, but little else supports this diagnosis. Patients with pulmonary emboli can develop bronchoconstriction immediately after embolization, and therefore have wheeze. Left heart failure requires consideration, although this is unlikely without cardiac enlargement or other abnormalities on the chest X-ray. Nevertheless, a left atrial myxoma, mitral stenosis and aortic stenosis need to be excluded. Useful investigations would be lung function tests, a ventilation/perfusion lung scan, an echocardiogram and

Doppler ultrasound examination of the lower limbs. Lung function tests demonstrated an FEV1 of 4.0 L, vital capacity of 5.4 and a peak flow of 555 L/min. These results were better than predicted for his age and sex, and demonstrate no evidence of airways obstruction (e.g. asthma) or of a significant restrictive defect (e.g. pulmonary infiltration). The echocardiogram demonstrated dilatation of both the right ventricle and right atrium, with minor tricuspid regurgitation. There was significant pulmonary hypertension. The ventilation/perfusion scan showed multiple abnormalities in both lungs, with reduced perfusion and relatively well preserved ventilation (ventilation/perfusion mismatch) (Case Fig. 13.6.1). The appearances were considered to indicate a high probability of pulmonary emboli. Doppler ultrasound demonstrated thrombus in the external iliac, common femoral, deep femoral, superficial femoral and popliteal veins on the left side. Following the confirmation of deep vein thrombosis and pulmonary embolic disease, the patient was treated with heparin by infusion and subsequently started on oral anticoagulation with warfarin. Within days there was marked improvement in symptoms, and after 10 days' treatment he was up and about, walking freely without breathlessness. At outpatient follow-up 2 months after his hospitalization the patient reported that he was well, not short of breath, with no cough, and able to play a full round of golf.

Question 3, What would be your anticoagulation strategy for this patient?

13

CASE FIG. 13.6.1 Ventilation/perfusion lung scan. The lung perfusion scans (top row) show multiple areas of reduced/absent perfusion. The ventilation scans (bottom row) are much less abnormal, and in many instances there is normal ventilation of poorly perfused areas of the lung. The scans indicate a high probability of pulmonary emboli. Seven months after his initial hospital presentation, and 10 days after having stopped his warfarin therapy, the patient developed breathlessness and chest pain. The breathlessness started acutely and deteriorated rapidly over a 4-day period. By the time of admission to hospital he was breathless at rest, and on the day of admission had an episode of collapse with loss of consciousness. He described his chest pain as a feeling of severe tightness. There had been no haemoptysis or pleuritic chest pain. On examination the patient looked ill, with cool pale peripheries. Blood pressure was 120/75, heart rate 119, and the oxygen saturation was 86% breathing air. The jugular venous pressure was not raised and there were no murmurs. Auscultation of the chest demonstrated no abnormalities. Questions 4. What is the explanation of the chest pain? 5. What would be your immediate management?

In severe pulmonary embolic disease the myocardium becomes ischaemic. Blood flow through the lung is substantially reduced because of widespread pulmonary emboli, and therefore cardiac output falls. The drop in cardiac output and arterial pressure reduces coronary artery blood flow. The patient's tachycardia reduces coronary perfusion during diastole. At the same time, the work of the heart is greatly increased because of the obstruction to the pulmonary vasculature. The combination of increased oxygen requirement of heart muscle combined with reduced oxygen delivery to the muscle results in ischaemic cardiac chest pain. Most patients with pulmonary emboli can be satisfactorily treated with heparin. However, if there is substantial cardiovascular impairment (as evidenced by hypotension, tachycardia, poor peripheral perfusion, reduced urine output or confusion) patients require treatment by thrombolysis. The most likely diagnosis was considered to be a recurrence of pulmonary emboli, and the patient had an urgent CT

CASE FIG. 13.6.2 CT scans. (A) Note the filling defect (due to thrombus) in the proximal left pulmonary artery (arrowed). (B) The distal branches of the pulmonary arteries are occluded by thrombus (arrows).

scan. This showed widespread pulmonary emboli affecting the pulmonary arteries to both lungs (Case Fig. 13.6.2). The patient was treated with a heparin infusion and subsequently with warfarin. Again, he made a rapid and complete recovery. Within 10 days he was discharged from hospital, and at outpatient follow-up he remained well. One year later his exercise tolerance was normal, he

677

CASE STUDY 13.6 CONTINUED continued to play golf and had no symptoms. The white cell count had slowly but progressively risen and was 40 000. He continues on warfarin therapy. Question

Deep vein thrombosis is more likely in patients with acquired or inherited risk factors for thrombosis. Acquired risk factors include advanced age, immobilization, surgery, malignancy, oral contraceptive and hormone replacement therapy, antiphospholipid syndrome and

myeloproliferative diseases. Inherited predispositions include factor V Leiden, antithrombin deficiency, and protein C and S deficiency. Therefore, before deciding on long-term management, patients need to be screened for thrombotic factors.

6. What other investigations should have been performed?

frequently multiple, involving the lower zones more frequently than the upper part of the lungs, and only some cause infarction. In many instances of pulmonary embolism without infarction the distal lung remains viable because of adequate oxygenation via the bronchial arterial blood supply and the airways. Pulmonary infarction is much more likely to occur when there is coexisting disease affecting the airways or bronchial circulation. Pathologically, the lesions of pulmonary infarction are markedly haemorrhagic and involve the visceral pleural surface.

Clinical features The characteristic symptom of pulmonary embolism is sudden breathlessness. Indeed, relatively few processes cause such sudden dyspnoea (p. 615). Lateral, usually basal, pleuritic chest pain and haemoptysis develop some time after the onset of breathlessness, and are only clinical features if infarction has occurred. The haemoptysis consists of frank red blood without sputum. In addition to respiratory symptoms there may be pain or swelling of a leg, suggesting deep vein thrombosis, or a history indicating an increased risk of thrombosis. On examination there may be signs of deep vein thrombosis. The respiratory rate is usually raised, and if infarction has taken place there may be a pleural rub and a small pleural effusion. If embolization has been extensive there will be cyanosis and signs of cardiovascular stress. The most important cardiovascular sign is a tachycardia, and some patients will have an elevated jugular venous pressure and a fourth heart sound. With extensive embolism patients may have signs of pulmonary hypertension, and occasion-

1

678

Fig. 13.30

2

Fig. 13.31

3 Fig. 13.32

ally a systolic murmur can be heard over the lung fields, as a consequence of turbulent pulmonary blood flow past partial pulmonary arterial occlusion. Within a few hours of pulmonary infarction fever is the rule. Arterial blood gas analysis usually demonstrates hypoxaemia. However, not all patients are hypoxaemic and hypoxaemia is in itself a very non-specific abnormality. As a consequence of hyperventilation, there is hypocapnia. Patients with pulmonary embolic disease are frequently anxious as well as breathless, and their hypocapnia is not uncommonly taken to reflect anxiety. However, anxiety hyperventilation syndromes produce hyperoxaemia. Immediately following embolization there is often bronchoconstriction, there may even be wheeze, and later, a reduction in surfactant in the affected lung is a contributory factor to atelectasis. Symptomatic pulmonary embolism occurs in about 30% of patients with deep vein thrombosis in the leg or pelvis; if asymptomatic events are added to this figure about 50-60% of patients with deep venous thrombosis will have a pulmonary embolism at some stage. However, in patients with suspected deep venous thrombosis only 25% will actually prove to have the condition when fully investigated.

Diagnosis The clinical diagnosis (Fig. 13.44) of pulmonary embolism is difficult and clinical criteria alone are seldom sufficient. Accurate diagnosis is important, as without specific therapy there may be further, fatal, embolization. Furthermore, the institution of anticoagulation therapy has important consequences for the patient and is not justifiable without sound evidence. Important clinical factors in the diagnosis are a history of predisposing factors, a clinically obvious deep vein thrombosis, haemoptysis and a pleural rub. The chest X-ray in pulmonary embolic disease is frequently normal. When pulmonary infarction occurs, radio-

13

FIG. 13.44 Strategy for the diagnosis of pulmonary emboli (PE) The indication for venography to investigate patients with a matched ventilation-perfusion detect will depend on the degree of clinical suspicion of embolic disease. The same argument applies to a subsequent pulmonary arteriogram when venography is normal.

logical changes develop over the following 24 hours and may include peripheral shadows adjacent to the pleura, most often basal 1 and frequently multiple, an elevated hemidiaphragm, some segmental or lobar volume loss, and small pleural effusions (Fig. 13.45A). If pleural fluid is aspirated it is found to be an exudate and is frequently haemorrhagic. Secondary infection of infarcted lung causes cavitation. With resolution, the linear shadows of pulmonary infarction are characteristic and usually have a pleural component (Fig. 13.45B). The ECG is rarely helpful, except in the diagnosis of myocardial infarction or pericarditis, which may require consideration in the differential diagnosis of pulmonary embolic disease. Occasionally, the ECG changes are suggestive of pulmonary embolism in that they demonstrate right axis deviation, right bundle branch block or an S wave in lead I, a Q wave and an inverted T wave in lead III. However, non-specific abnormalities, particularly T-wave changes, are much more common. Severe or recurrent pulmonary emboli may be associated with pulmonary hypertension and the ECG changes of right ventricular and right atrial hypertrophy. Isotope lung scans are a useful investigation in many patients with pulmonary embolic disease. A normal perfusion lung scan largely excludes pulmonary embolic disease, but an abnormal scan is non-specific. Although an abnormal scan in conjunction with a normal chest X-ray may be taken to suggest pulmonary embolism, perfusion scans may be markedly abnormal in some

pulmonary disorders associated with a clear radiograph, as, for example, in asthma. Furthermore, pulmonary angiography demonstrates emboli in only 50% of patients with an abnormal perfusion lung scan and a normal chest X-ray. Thus, for many patients the confirmation of pulmonary emboli, rather than the exclusion of this diagnosis, necessitates a ventilation as well as a perfusion lung scan. Characteristically, the perfusion scan demonstrates a filling defect due to vascular obstruction, whereas ventilation to the affected area is relatively normal (Fig. 13.46). 2 In doubtful cases serial scans, during which the perfusion defects of pulmonary emboli resolve, are diagnostically helpful. Large or multiple ventilation-perfusion mismatched areas are highly likely to be due to pulmonary emboli (80-90% of cases). In a minority of patients - perhaps up to 20% of cases - matched defects can be due to pulmonary embolism confirmed angiographically. For patients with a matched defect and a normal chest X-ray the diagnosis of pulmonary embolism can only be confidently excluded by pulmonary arteriography. The importance of achieving a definite diagnosis in equivocal cases rests on the fact that, untreated, about 50% will have recurrent, sometimes fatal, thromboembolism. Spiral CT scan of the thorax is a rapidly evolving technique that can demonstrate intravascular clot in the pulmonary arterial circulation. 3 It has become one of the diagnostic methods of choice, often in combination with isotope lung scans and measurement of D-dimers in the

679

FIG. 13.45 Pulmonary infarction: chest X-rays [A] Early changes. There is consolidation at both bases, particularly the left. The left hemidiaphragm is elevated. The patient has made a poor inspiratory effort due to pleuritic pain. [B] Resolving infarction. Linear basal shadows are characteristic.

blood. Normal levels of D-dimer make thromboembolisim unlikely (Fig. 13.44). In patients documented angiographically to have pulmonary emboli, 70% will have deep vein thrombosis, and therefore the demonstration or exclusion of deep vein thrombosis is helpful in determining the likelihood of pulmonary emboli. Venography has been helpful in the past in the management of patients with suspected deep venous thrombosis. However, more sophisticated duplex Doppler techniques, which are easier to perform and do not require the use of contrast media, are now and are reliable and reproducible methods of demonstrating venous thrombosis in leg and pelvic veins.

Management

680

The purpose of therapy is to halt the propagation of peripheral thrombus and avoid further pulmonary embolism, thereby allowing time for natural thrombolysis. The treatment of acute pulmonary embolism is intravenous heparin. Massive pulmonary embolism may require additional therapy (p. 681). An effective heparin treatment regimen is a continuous infusion, such that the partial thromboplastin time is prolonged to 1.5-2.0 times normal, commonly requiring 30 000-40 000 units of heparin daily. Heparin requirements tend to fall after 2 or 3 days. It is important to emphasize that the most critical period for aggressive treatment of deep venous thrombosis and

pulmonary embolism is in the first 48 hours after diagnosis; adequate heparinization during this period will greatly reduce the extension of clot and episodes of recurrence. Low molecular weight heparin treatment, with enoxaparin or dalteparin, which can be given subcutaneously once daily, has been studied for the treatment of deep venous thrombosis and has been found to be effective. It is likely that these drugs, as they are studied further, will replace intravenous heparin in the management of pulmonary embolism. If heparin treatment is complicated by bleeding, the action of heparin is rapidly reversed by the administration of protamine sulphate. Heparin therapy should be undertaken for 7-10 days. Thrombolytic therapy, most commonly used with massive pulmonary embolism (see below), is occasionally necessary when heparin treatment is inadequate, including cases with severe or extensive venous thrombosis, particularly when the inferior vena cava is involved. Long-term anticoagulant therapy with warfarin is initiated 3-4 days before heparin therapy is stopped, and is continued for 3-6 months unless a chronic predisposing factor for deep venous thrombosis (e.g. chronic venous insufficiency of the legs) indicates the need for prolonged therapy. If bleeding complicates warfarin therapy, treatment is with vitamin K1. In pregnancy, and in patients in whom intravenous heparin therapy is particularly hazardous, subcutaneous heparin is the treatment of choice. The resolution of pulmonary emboli is rapid, usually within 2-4 weeks. Heparin therapy is highly effective in preventing further thromboembolism.

13

FIG. 13.46 Pulmonary embolic disease: ventilation and perfusion lung scans Posterior scans demonstrate normal ventilation [Al. and numerous, mainly peripheral, defects of perfusion [B] characteristic of multiple pulmonary emboli. Such changes usually resolve over a period of several weeks.

MASSIVE PULMONARY EMBOLISM Massive pulmonary embolism is a relatively rare hospital emergency as most patients with fatal pulmonary emboli die within 2 hours, frequently before reaching medical help. For those who survive beyond 2 hours the outlook is good in most cases, provided they receive appropriate medical treatment. Surgical intervention is only rarely required.

Clinical features The clinical features of massive pulmonary embolism reflect the haemodynamic consequences of obstruction to a large proportion of the pulmonary arterial tree; more than 60% obstruction is necessary to produce serious consequences in a previously healthy individual. Less extensive embolism will be life-threatening in patients with severe pre-existing cardiorespiratory disease. Following embolism, some of the increase in pulmonary vascular resistance may be due to vasoconstriction, secondary to the release of agents such as serotonin from thrombus. The dominant symptom of massive embolism is severe breath lessness. In some patients there is central chest pain caused by myocardial ischaemia, reflecting the combination of a massive increase in right ventricular oxygen requirements and a reduction in coronary artery blood flow. Hypotension may cause syncope. Examination shows tachypnoea, tachycardia, hypotension and poor peripheral perfusion. Signs of pulmonary hypertension and right ventricular strain are not always present. The ECG is frequently unhelpful, with the most common abnormality, other than a tachycardia, being T-wave changes. The main importance of the ECG is in the differential diagnosis from massive myocardial infarction. Echocardiography is a very useful additional tool in the diagnosis of massive pulmonary embolism. During the acute phase it may show right ventricular dilatation and a reduction in right ventricular contractile function, in addition to elevated right heart

pressures. These will resolve when the embolic episode has passed. The chest X-ray is frequently normal or nonspecifically abnormal. Oligaemia of the lung fields is helpful when present, but can seldom be appreciated on emergency supine chest X-ray films.

Diagnosis and treatment Therapy is urgent and potentially life-saving. If the diagnosis of massive pulmonary embolism is probable and the ECG does not suggest myocardial infarction, urgent investigations are required, but immediate treatment with intravenous heparin (15000 units) is reasonable. When hypotension is severe, inotropic support to the right ventricle is needed. In patients who have collapsed with massive pulmonary embolism, external cardiac massage can move clot peripherally and restore part of the pulmonary circulation. Patients have and require high right heart filling pressures, and treatment with diuretics and vasodilators, including morphine, is deleterious. If the patient is capable of cooperation and facilities are immediately available, a ventilation-perfusion scan is performed. Indeed, a perfusion scan alone will be grossly abnormal and strongly support the diagnosis, provided the chest X-ray is clear. Patients who, for whatever reason, are unable to have isotope lung scans, require CT scanning or a pulmonary arteriogram, as do all patients for whom surgery is being considered. Pulmonary arteriograms and CT scans normally demonstrate large central filling defects (Figs 13.47,13.48). Most patients who survive long enough to reach hospital, and who are then treated with intravenous heparin and inotropic support, rapidly improve. Following confirmation of the diagnosis some are best treated with thrombolytic agents, which achieve more rapid and possibly more complete resolution than does intravenous heparin. Bleeding complications are, however, twice as likely as with heparin, and thrombolytics should not be given if patients have active haemorrhage, a history of a recent cerebrovascular accident or recent surgery, organ biopsy or childbirth.

681

FIG. 13.47 Massive pulmonary embolism The pulmonary arteriogram shows a large filling delect in the right pulmonary artery, with almost no perfusion of the right lung and little perfusion of the left lower lobe peripherally.

Streptokinase is the thrombolytic agent most commonly used, but is antigenic. Urokinase is less antigenic but more costly. Thrombolytic agents are usually infused directly into the pulmonary artery. A small number of patients with severe haemodynamic disturbance fail to improve on medical therapy and require embolectomy. Surgery is performed with the patient on cardiac bypass. It is critically important that the definitive diagnosis of massive pulmonary embolism has been substantiated prior to surgery and, if so, the operative results are good (75% survival).

RECURRENT PULMONARY EMBOLISM With adequate anticoagulation therapy recurrent pulmonary emboli are uncommon. Rarely, anticoagulants are absolutely contraindicated, and in such patients, and those with recurrent emboli despite full anticoagulant therapy, inferior vena caval sieves can be used. Such surgical procedures are not always effective: the inserted devices may become dislodged, blocked or infected, and indeed may become a source of future emboli.

Repeated, often silent, pulmonary emboli can gradually obstruct a large proportion of the pulmonary vasculature and lead to pulmonary hypertension. Such patients may have a chronic low cardiac output state and intractable right heart failure. They have severe breathlessness, the cause of which is frequently not appreciated at first presentation. Any possibility of recurrent pulmonary emboli warrants comprehensive investigations including angiography or CT scanning. Anticoagulant therapy can prevent progression of the disease. When angiography or CT demonstrates central filling defects, some cardiothoracic centres have produced benefit to patients by surgical removal of organized clot. There is debate about the pathogenesis of pulmonary hypertension in relation to recurrent pulmonary emboli. It has been hypothesized that the raised pulmonary arterial pressures are not simply due to mechanical blockage of the pulmonary arterial tree, but may be caused by an intrinsic pulmonary arteriopathy associated with in situ thrombosis. New surgical techniques for pulmonary thromboendarterctomy often discover widespread hypertensive lesions throughout the pulmonary vasculature, without any visible thrombosis.

Prevention of venous thromboembolism The prevention of pulmonary embolism is possible if patients at high risk for the development of deep vein thrombosis and therefore pulmonary emboli are recognized and suitable prophylaxis undertaken (Ch. 12, p. 604).

FURTHER READING ON PULMONARY THROMBOEMBOLISM AND HYPERTENSION Hyers T M 1999 Venous thromboembolism. Am J Respir Crit Care Med 159:1-14. Recommendations on the management of pulmonary hypertension in clinical practice. Heart 2001; 86(suppl 1) il-i!3.

SARCOIDOSIS Sarcoidosis is a multisystem disorder of unknown aetiology. It can involve the liver, spleen, lymph nodes, heart, nervous system, salivary glands, muscles, bones and other organs. It usually affects young and middle-aged adults and the commonest presentation is with bilateral hilar gland enlargement. The diagnosis is made on the basis of the clinical and radiological findings, in combination with histological evidence of non-caseating granulomas.

Prevalence and distribution 1

682

MCQ 13.18

Sarcoidosis may present to all specialties and is relatively common. Worldwide, the prevalence is approximately 20 in 100000. It is more common in temperate climates, black

13

FIG. 13.48 Pathogenesis of sarcoidosis

Americans, and West Indian and Irish immigrants to the UK, and is less common in Chinese and peoples of the Indian subcontinent. However, differences in incidence may reflect, in part, a difference in the awareness of the disease and the availability of diagnostic facilities, particularly radiology, and in many parts of the world the disorder has been frequently misdiagnosed as tuberculosis. Sarcoidosis is unusual in children and in the elderly and is most common between the ages of 20 and 40, with a slight female preponderance. In general, the disease is more florid in Afro-Caribbeans than in Caucasians. There are occasional reports of familial and occupational 'clusters'.

Aetiology and pathogenesis (Fig. 13.48) The cause of sarcoidosis is unknown, but it is postulated that it results from the exposure of genetically susceptible individuals to environmental agents. The evidence for host reaction to environmental exposure includes epidemiological studies, the nature of the inflammatory response in sarcoidosis with the involvement of activated macrophages and CD4 T lymphocytes, and T-cell receptor studies in affected individuals. Inhalation is the most likely route for the causal agent and the sequence of events is probably one of alveolar injury and an influx of immune effector cells, followed by an inflammatory alveolitis. The alveolitis of sarcoidosis is predominantly lymphocytic (as shown by examination of lung tissue and bronchoalveolar lavage fluid), mostly acti-

vated T lymphocytes, with a reduction in circulating blood T lymphocytes. The increased lymphocytes in the lung are mostly T-helper cells, which are relatively depleted in peripheral blood. These changes may be responsible for the alveolitis, and partly explain the peripheral anergy of delayed-type hypersensitivity of sarcoidosis. Activated T lymphocytes secrete monocyte chemotactic factor, recruiting circulating monocytes and macrophages into the alveolitis, leading to granuloma formation. Lymphocytes also activate macrophages to form the epithelioid and giant cells of granulomas. The metabolic activity of macrophages is associated with increased levels of angiotensin-converting enzyme (ACE) in lung tissue, bronchoalveolar lavage fluid and serum. Whether the granulomas resolve or progress to fibrosis may be determined by the intensity of the alveolitis. The increase in T-cell activity stimulates B lymphocytes, which results in raised serum immunoglobulins, circulating antibodies and immune complexes. Heredity may play a part in the predisposition to sarcoidosis, and the condition is more common with some HLA groups. The characteristic pathological finding of sarcoidosis is therefore granulomas that consist of collections of large histiocytes (epithelioid cells) with occasional multinucleated giant cells and, more peripherally, lymphocytes (CDS-suppressor cells). There is no central necrosis, a point of contrast to many cases of tuberculosis, and the reticulin between the histiocytes is therefore intact. Non-caseating granulomas are not specific for sarcoidosis and can occur in

683

TABLE 13.37 Causes of non-caseating granulomas Sarcoidosis Tuberculosis Primary biliary cirrhosis Leprosy Tertiary syphilis Brucellosis HypogammagIobuIinaemia Fungal infection

Berylliosis Foreign body reactions Cat-scratch disease Hypersensitivity pneumonitis Granulomatous arteritides Lymphomas Carcinoma (regional lymph node) Crohn's disease

other disorders (Table 13.37). Sarcoid granulomas are dynamic entities that develop, age and resolve. In chronic cases they are replaced by hyalinization and fibrosis.

Clinical features Because sarcoidosis can affect virtually all organs and tissues of the body, with the probable exception of the adrenal glands, the clinical features of this disorder are varied (Table 13.38). Non-specific constitutional symptoms are the presenting feature in about a third of patients; these include fatigue, fever and weight loss. Sarcoidosis is a cause of pyrexia of unknown origin. The lungs, liver, eyes and skin are most frequently involved, with more than 85% having some degree of pulmonary involvement, and many cases are diagnosed following a routine chest X-ray. Extensive sarcoidosis causes fever. Eye involvement can cause blindness. Respiratory, cardiac and renal failure are occasionally fatal.

TABLE 13.38 Clinical spectrum of sarcoidosis in the UK Clinical feature

% of cases

Abnormal chest X-ray (lung involvement ± hilar or mediastinal adenopathy) Erythema nodosum Lymphadenopathy Palpable splenomegaly Eye involvement Skin sarcoid Enlargement of parotid/lachrymal/other salivary glands Symptomatic nervous system involvement Bone cysts Symptomatic heart disease Symptomatic myopathy Liver biopsy positive*

>85 20-30 10-15 6 15-25 5-20 5-20 5 3-8 <5 <1 50-80

* Most patients with sarcoidosis can be diagnosed without the need for a liver biopsy.

Thoracic sarcoidosis The chest X-ray is abnormal in more than 85% of cases and the radiological appearances form the basis for the classification of thoracic sarcoid (Table 13.39). Stage I Hilar adenopathy is usually bilateral and symmetrical, 1 but can be unilateral. Adenopathy is the commonest manifestation of thoracic sarcoidosis, and more than half of patients have no symptoms. In addition to hilar gland enlargement the right paratracheal gland is frequently enlarged (Fig. 13.49). Bilateral hilar adenopathy can also occur in other disorders (Table 13.40). In up to 40% of stage I cases there is also erythema nodosum, and in the UK 80% of cases of erythema nodosum in young adults are due to sarcoidosis. 2 In association with bilateral hilar adenopathy and erythema nodosum there is commonly low-grade fever, polyarthralgia and a raised ESR.

684

1

Fig. 13.33 2 Fig. 13.34 3 Fig. 13.35

4

Fig. 13.36

FIG. 13.49 Stage I sarcoidosis There is bilateral hilar lymphadenopathy (BHL) and also enlargement of the right paratracheal nodes (arrow), characteristic of sarcoidosis.

Although few patients have severe symptoms with stage I sarcoidosis a minority have cough and chest pain due to adenopathy. Bilateral hilar adenopathy is usually benign, with 80% of cases resolving in less than 12 months and 90% by 2 years. In 10% of cases glandular enlargement persists and eggshell calcification may develop, or patients may develop pulmonary opacities. Patients with bilateral hilar lymphadenopathy without erythema nodosum probably do rather less well, with up to 20% developing a pulmonary infiltrate.

TABLE 13.39 Classification of thoracic sarcoid Stage

Description of chest X-ray

0 I II

Normal Hilar adenopathy Hilar adenopathy plus parenchymal infiltrate Parenchymal infiltrate ± fibrosis

III

Stage at presentation (%)

Elevated ACE levels (%)

Positive transbronchial lung biopsy (%)

Positive Kveim test (%)

Positive liver biopsy (%)

8 51

-

50

-

65-80

29

30 56 71

75-100

70 80

85 65

12

56

75-85

40

60

TABLE 13.40 Differential diagnosis of bilateral hilar adenopathy

TABLE 13.41 Differential diagnosis of stage II and stage III sarcoidosis

Sarcoidosis* Tuberculosis Lymphoma Leukaemia Metastatic malignant disease

Cryptogenic fibrosing alveolitis Pulmonary tuberculosis Extrinsic allergic alveolitis Carcinoma Pneumoconiosis

Beryllium disease Hypogammaglobulinaemia Histoplasmosis Coccidioidomycosis Enlargement of pulmonary arteries

13

Bronchopulmonary aspergillosis Ankylosing spondylitis Berylliosis Histiocytosis X Pneumocystis carinii pneumonia

* Compared with sarcoidosis, all other causes of bilateral hilar adenopathy are uncommon.

prognosis for stage II sarcoidosis is less good than for stage I. Although 50% resolve spontaneously within 2 years, some take longer, 30-40% require steroid therapy, and in 10-15% long-term steroid therapy is required, despite which a small number develop progressive pulmonary fibrosis. The differential diagnosis of stage II and stage III sarcoidosis is shown in Table 13.41. Stage III With the development of pulmonary fibrosis, patients become breathless, hypoxic, and ultimately develop cor pulmonale. Pulmonary function tests demonstrate a severe restrictive ventilatory defect. In addition to fibrosis, which is characteristically most marked in the mid and upper zones, the chest X-ray may also show cavitation and bulla formation. 4 As would be expected, the prognosis for stage III sarcoidosis is less good, with about 30% showing significant improvement with steroid therapy. Extrathorocic sarcoidosis FIG. 13.50 Stage II sarcoidosis The X-ray shows florid stage II sarcoidosis, with bilateral hilar lymphadenopathy (BHL) and extensive pulmonary infiltrates. Pleural abnormalities in sarcoidosis are rare.

Stage II Pulmonary infiltrates coexist with adenopathy in about 30% of patients (Fig. 13.50). More severely affected patients have a restrictive ventilatory defect, with exertional breathlessness and a reduction in gas transfer. The chest X-ray appearances correlate poorly with physiological disturbances, the patient being frequently asymptomatic despite widespread pulmonary shadowing. The

Skin Erythema nodosum (see p. 407) is the most common skin abnormality, and is associated with acute, usually stage I, sarcoidosis. When there is erythema nodosum and BHL in a young adult a clinical diagnosis of sarcoidosis can be firmly made without resort to tissue biopsy. Erythema nodosum is an immunological response associated with sarcoidosis, and skin biopsy demonstrates a characteristic histology which is the same as for other causes of erythema nodosum. Sarcoid tissue itself can also infiltrate the skin: biopsy shows typical noncaseating granulomas. Small papules, plaques and subcutaneous nodules are common, and the skin of the nose may be involved (lupus pernio), as

685

are scars and keloid tissue (Fig. 10.29, p. 408). When there is lupus pernio the upper respiratory tract is frequently involved by the disease and nasal mucosal biopsy is positive. Skin sarcoid more commonly affects females, and is more common, extensive and florid in Afro-Caribbeans.

between 2 and 10% of patients with sarcoidosis, and hypercalciuria in about three times that number. If undetected or untreated it may cause nephrocalcinosis, renal stones and eventually, renal failure.

Eyes Eye involvement is both common (up to 25% of cases) and potentially serious. Early eye involvement may be asymptomatic and many patients should be examined by an ophthalmologist. Uveitis, conjunctivitis and retinal involvement can all occur. Dry eyes (keratoconjunctivitis sicca) occurs in a Sjogren-like syndrome when the salivary glands are involved. The lachrymal, salivary and parotid glands may be enlarged. Eye involvement is an indication for steroid therapy, usually systemic.

Diagnosis

Hypercalciuria Hypercalciuria and hypercalcaemia can occur, but these disorders of calcium metabolism (p. 965) are not common. Lymphadenopathy Enlarged lymph nodes are present in 10-15% of patients and splenomegaly occurs in 6%, both abnormalities being more common in Afro-Caribbeans. Occasionally the spleen is very large and associated with anaemia, neutropenia and thrombocytopenia, which may improve following splenectomy. Cardiac sarcoid Involvement of the myocardium by granulomatous infiltration may be relatively common and is discussed on page 573. Nervous system Sarcoid affecting the nervous system is discussed on page 1431. Sarcoid involvement of muscle is usually asymptomatic, but occasionally patients have myopathic symptoms. Bone sarcoidosis Arthritis, acute and transient, is common with BHL and erythema nodosum, but chronic skeletal problems are unusual (3%), characteristically affecting the hands and feet in patients with skin sarcoidosis. Bone cysts, especially of the terminal phalanges, are most typical. Steroids have little effect and bone lesions are usually confined to cases of chronic sarcoidosis. Parotid glands Heerfordt's syndrome is fever, parotid gland enlargement, facial palsy and anterior uveitits, all due to sarcoidosis of the involved organs. Endocrine manifestations Abnormal production of calcitriol by activated macrophages and granulomas leads to hypercalcaemia in 1

686

Fig. 13.37

Chest X-ray, CT scans and transbronchial biopsy give the diagnosis in the majority of patients. High-resolution CT scans can be particularly helpful, showing features characteristic of sarcoidosis. 1 It is noteworthy that even when the chest X-ray shows extensive parenchymal disease, auscultation is remarkably normal. Biopsy of any clinically involved tissue will confirm the diagnosis, and a liver biopsy will be positive in 50-80% of patients. Frequently, the most important differential diagnosis is between sarcoidosis and tuberculosis; tissue should therefore be stained for acid-fast bacilli and cultured for M. tuberculosis. Depression of delayed-type hypersensitivity in sarcoidosis reduces tuberculin reactivity. Two-thirds of patients do not react to 100 tuberculin units, and more than 90% fail to react to 10 tuberculin units. A depressed tuberculin reaction persists for many years, long after clinical resolution of sarcoidosis. In the differential diagnosis of sarcoidosis and tuberculosis, the tuberculin test is of considerable value. In cases where biopsy is relatively difficult (e.g. hilar adenopathy) or negative, a diagnosis of sarcoidosis can be substantiated by the Kveim test (Table 13.38). A suspension of particulate human sarcoid tissue is injected intradermally; in patients with active sarcoidosis epithelioid cell granulomas gradually develop and can be demonstrated following skin biopsy of the purplish-red nodule at 4 weeks. A positive Kveim test strongly suggests a diagnosis of sarcoidosis (false positives 1-2%). However, the Kveim test is not always positive in sarcoidosis and there is a relatively high false negative rate in chronic fibrotic disease. As a result of advances in radiology and biopsy techniques, Kveim tests are now seldom performed.

Management Assessment of the severity and activity of disease Although symptoms and signs are the main guide to management, further investigations are necessary and helpful in decision-making. Radiology Chest X-rays provide a crude index of the extent of thoracic sarcoidosis, and the correlation between radiological findings and functional impairment is poor. However, serial chest X-rays over months and years remain indispensable for documenting qualitative changes that occur either spontaneously or in response to therapy. Serum angiotensin-converting enzyme Serum angiotensin-converting enzyme (SACE) levels are high with acute granulomatous disease and correlate with

CASE STUDY 13.7 FEVER AND CHEST X-RAY ABNORMALITIES IN A 45-YEAR-OLD ASIAN WOMAN A 45-year-old Asian woman presented to her GP with a fever which she had noticed for 6 weeks. She had felt generally weak for 2 months and had experienced bilateral arthralgia in the lower limbs, particularly the knees. There had been 1 stone weight loss. The patient had not experienced shortness of breath and had no cough or sputum. Originally from India, she had been in the UK for 20 years and worked as a teacher. Soon after arriving in London she had developed pulmonary tuberculosis and had been treated with conventional therapy for 9 months. On examination the temperature was 37.6°C; there were no abnormal physical signs and, in particular, auscultation of the chest was normal. Routine blood tests were normal, except that the ESR was 38 mm/h. The chest X-ray showed bilateral pulmonary infiltrates, most severe in the midzones, and hilar lymphadenopathy (Case Fig. 13.7.1).

1,1 What is the difftirenflal .diagnosis?' 2, What should be the next investigations?

CASE FIG. 13.7.1 Chest X-ray. There is bilateral hilar lymphadenopathy and enlargement of the paratracheal nodes on the right. The bilateral pulmonary infiltrates are most marked in the mid and upper zones.

Discussion The two most likely diagnoses are tuberculosis and sarcoidosis. Tuberculosis is favoured by the past medical history, the ethnic origin and, to an extent, the fever. However, the radiology is more suggestive of sarcoidosis; the bilateral hilar lymphadenopathy and the bilateral mid and upper zone infiltrates being characteristic, as is the enlargement of the right paratracheal nodes. In sarcoidosis there may be no abnormal signs in the chest (e.g. crackles, bronchial breathing) despite substantial chest X-ray changes. Fever is not uncommon. Other infections should be considered, but the history is rather long. Lymphadenopathy occurs in lymphoma, as does fever, but the symmetrical pulmonary infiltrates would be unusual. Three helpful tests would be a tuberculin skin test, serum ACE and bronchoscopy with lavage (particularly seeking acid-fast bacilli), and transbronchial biopsy. The tuberculin skin test (Mantoux 1 in 1000) was negative. In a patient with active tuberculosis or past tuberculosis it would be expected that the skin test would be obviously positive. Skin anergy is characteristic of sarcoidosis, and the negative test in a patient known to have had tuberculosis in the past supports the diagnosis. Tuberculin skin testing is therefore a very useful, simple investigation when considering the differential diagnosis between tuberculosis and sarcoidosis. The ACE was raised at 118. A number of conditions cause a raised ACE, but sarcoidosis is the most common. Although some patients with tuberculosis can have a raised ACE, it is relatively unusual and therefore in this patient, the increase in the ACE favours the diagnosis of sarcoid. Bronchial lavage demonstrated no AFB and subsequent culture was negative. Transbronchial biopsy

13

showed non-caseating granulomas consistent with a diagnosis of sarcoidosis. Because the histology of tuberculosis and sarcoidosis is granulomatous, histopathologists are often reluctant to make a definite diagnosis unless acid-fast bacilli are also seen. The granulomata of sarcoidosis do not caseate and, particularly given the other features supporting the diagnosis, the histology makes a diagnosis of sarcoidosis in this case certain.

Question 3. What treatment would you initiate?

In essence, the question is whether or not to treat with oral steroids. Steroids markedly reduce granuloma formation, reduce the serum ACE, reduce the extent of sarcoid tissue (e.g. lymph nodes or pulmonary infiltrates) and reduce systemic symptoms. It is less clear whether steroids alter the long-term outlook, but in patients with abnormalities in the lung that cause reduced lung function the use of steroids results in better long-term respiratory function. In this patient, with fever and bilateral pulmonary infiltrates, it is on balance reasonable for her to be treated with prednisolone.

Question 4. What are common extra-, if pulnjonary manifestations of sarcoidosis? Sarcoidosis can affect virtually any organ in the body (see p. 1431); after lung and lymph nodes, the tissues and organs most commonly involved are the liver, skin and eyes. Soon after the diagnosis of sarcoidosis was made, this patient developed a left lower motor neuron

687

CASE STUDY 13.7 CONTINUED facial weakness and nodular changes in the skin on her face (Case Fig. 13.7.2). Although neurological involvement in sarcoidosis is relatively uncommon, the most frequent neurological manifestation is a facial nerve palsy. This complication should be treated with steroids, and if therapy is given early the outcome is usually good. However, in some patients the response is only partial and in a minority no improvement occurs. Skin lesions are relatively common in sarcoidosis. It is important to distinguish erythema nodosum, which is commonly associated with sarcoidosis, from infiltration with granulomatous tissue. Erythema nodosum is an immunological response to sarcoidosis rather than involvement by the disease. Sarcoidosis of the skin responds to steroid therapy, may also be helped by chloroquine, and particularly troublesome lesions can benefit from local injections of steroid. This patient was treated with prednisolone 40 mg daily. On this therapy there was rapid improvement of her facial palsy, the skin lesions disappeared, the chest X-ray improved with

CASE FIG. 13.7.2 The patient has been asked to close her eyes and complete lower motor neuron facial paralysis is apparent. Note also the multiple sarcoid skin nodules. substantial resolution of the pulmonary infiltrates, her fever subsided and she felt much stronger. The patient's progress was monitored in the outpatient clinic, with regular review of symptoms, signs, lung function, serum ACE and chest X-ray. Oral steroids were gradually reduced and eventually stopped altogether after 18 months. Soon after stopping the steroids the

chest X-ray abnormalities. However, not all patients with sarcoidosis have raised SACE levels and, conversely, some patients without sarcoid (perhaps up to 20%) can have high levels (Table 13.42). Tuberculosis and lymphoma seldom increase serum ACE. SACE is most useful in monitoring granulomatous activity and the response to steroid therapy. Bronchoalveolar lavage In normal subjects BAL usually shows less than 10% lymphocytes, whereas in sarcoidosis average figures for lymphocyte counts are 30%, with up to 50-60% in florid,

MCQ 13.19

688

patient developed erythema nodosum, the ACE was much elevated and her facial weakness became more pronounced. Her steroids were reintroduced and these problems responded quickly. She has now been followed up for 5 years and remains well. The ACE level fluctuates and she continues on 5 mg of prednisolone daily. There is minor residual facial weakness.

TABLE 13.42 Causes of a raised ACE Sarcoidosis Tuberculosis (occasionally) Liver disease Gaucher's disease Asbestos is Berylliosis

Silicosis Diabetes Hyperthyroidism Leprosy Histoplasmosis

acute disease. However, there is an overlap of the cell count in sarcoidosis with those in normal subjects, particularly smokers, as well as in patients with extrinsic allergic alveolitis, lymphomas and cryptogenic fibrosing alveolitis. BAL is not necessary for the clinical management of the disease.

Clinical Lung function tests

The correlation between chest X-ray appearances and pulmonary function is poor. Pulmonary sarcoidosis leads to a restrictive ventilatory defect with a reduction in lung volumes, pulmonary compliance and gas transfer. The most sensitive index of impaired pulmonary function is abnormal gas transfer, which may be reduced even in stage I disease. Some patients with sarcoidosis have an obstructive defect due to endobronchial disease (confirmed by bronchoscopy and bronchial biopsy and improved by steroid therapy) or to airway distortion from fibrosis. With moderate or severe disease patients may be hypoxic at rest, and in mild disease there is hypoxia on exertion. Pulmonary function testing is essential in the assessment of sarcoidosis, and repeated measurements are of great value in longterm management.

Treatment The known effects of steroids on immunological mechanisms suggest that these drugs should be helpful in suppressing the immunologically mediated alveolitis of sarcoidosis and, indeed, patients with acute florid disease appear to respond rapidly. In most patients the short-term effect of steroids is to improve symptoms, suppress inflammation and granuloma production and lower SACK levels. The chest X-ray can improve quickly, within 2 or 3 weeks. Prednisolone at a dosage of 30-40 mg daily, with a gradual reduction towards the minimum dose that controls disease activity, frequently 7.5-10mg daily, is normally prescribed. Treatment with steroids is indicated whenever vital organs are severely involved, including the lung, or when there is substantial systemic disturbance (Table 13.43). Treatment with systemic corticosteroids usually results in remission of granulomas in biopsies and substantial clearing of chest X-ray infiltrates and improvement in symptoms. Although data are limited, in patients with persistent pulmonary disease steroids do improve long-term lung function. Alternative immunosuppressive agents (i.e. methotrexate, azathioprine or ciclosporin) and chloraquine may be

TABLE 13.43 Indications for steroid treatment In sarcoidosis Eye involvement Severe chest X-ray changes (stage II and stage III disease) associated with a high serum angiotensin converting enzyme Intense alveolitis on bronchoalveolar lavage Breathlessness Hypercalcaemia and hypercalciuria Severe skin infiltration Severe involvement of the upper respiratory tract Cardiac involvement Nervous system involvement Salivary gland involvement

used for chronic disease, particularly chronic skin infiltration, but there is no evidence that these agents are superior to oral steroids. In severe intractable disease, and particularly CNS sarcoidosis, high-dose steroids (typically pulses of methylprednisolone) may be useful.

13

Prognosis Approximately 60% of all patients with thoracic sarcoidosis resolve spontaneously and have a normal chest X-ray, mostly within 2 years. A further 20% will resolve with steroid therapy, which can then be discontinued. These patients have an acute or subacute type of the disease, clearly associated with a good prognosis. For the minority 10-20% with chronic disease, resolution, even with steroid therapy, is less likely, with radiographic clearing in about 50% who have stage II and 20% who have stage III disease. Patients with chronic pulmonary sarcoidosis may eventually develop pulmonary fibrosis and respiratory failure. With cavitary and bullous disease haemoptysis can be troublesome and mycetomas can occur. Although only a small percentage (2%) of patients with sarcoidosis die of the condition, most of these deaths are from pulmonary fibrosis and respiratory failure. Chronic extrathoracic sarcoidosis often responds poorly to treatment. Central nervous system involvement is particularly chronic and poorly responsive to steroid therapy.

FURTHER READING ON SARCOIDOSIS Gibson G J, Prescott R J, Muers M F et al. 1996 British Thoracic Society Sarcoidosis Study: effects of long term corticosteroid treatment. Thorax 51:238-247. American Thoracic Society 1999 Statement on sarcoidosis. Am J Respir Crit Care Med 160:238-247.

CONNECTIVE TISSUE AND RELATED DISEASES AND THE LUNG 1 Systemic lupus erythematosus (SLE) Pleurisy is common in SLE, affecting 50% of patients, with effusions in approximately 30%. An acute pulmonary vasculitis can cause widespread pneumonitis in about 10% of cases, sometimes leaving a residual basal atelectasis. Interstitial fibrosis is unusual. Patients may develop smallvolume lungs, the cause of which is unknown and which, in most cases, is not due to diaphragm weakness. In patients with the SLE-like syndrome characterized by the presence of anticardiolipin antibodies, thrombotic obliteration of the pulmonary vasculature can lead to pulmonary hypertension and severe breathlessness.

Mixed connective tissue disease Mixed connective tissue disease (MCTD) is a syndrome composed of features of SLE, polymyositis, rheumatoid

689

arthritis and scleroderma; it is not infrequently complicated by pulmonary involvement, including pleurisy, pulmonary fibrosis and pulmonary hypertension secondary to vascular involvement.

Rheumatoid arthritis The pulmonary manifestations of rheumatoid arthritis are described in Chapter 22, page 1136.

Systemic sclerosis The most common pulmonary manifestation of systemic sclerosis is pulmonary fibrosis, and pleurisy is unusual. Occasionally, pulmonary vascular disease occurs in the absence of parenchymal abnormalities, leading to pulmonary hypertension. Aspiration, with consequent pneumonitis, infection and bronchiectasis, is not uncommon in systemic sclerosis. Pneumonitis and fibrosing alveolitis can occur in Sjogren's syndrome and dermatomyositis. In dermatomyositis muscle weakness can be a major factor contributing to breathlessness and ventilatory failure.

Goodpasture's syndrome Goodpasture's syndrome consists of crescentic glomerulonephritis and pulmonary haemorrhage, associated with circulating antiglomerular basement membrane antibodies (see p. 1076). The onset of the syndrome frequently follows an upper respiratory tract infection, and is most common in young or middle-aged males. Haemoptysis is a key symptom, along with breathlessness. The chest X-ray shows diffuse alveolar shadowing (Fig. 13.51). The pulmonary haemorrhage causes an elevation in gas transfer, particularly KCO. Pulmonary bleeding is the most common cause of death. Prompt therapy with plasma exchange and immunosuppressive drugs has greatly improved the prognosis.

VASCULITIDES AND THE LUNG 1

Wegener's granulomatosis In Wegener's granulomatosis, described by Wegener in 1936, the characteristic pathology is a necrotizing granulomatous arteritis of the upper and lower respiratory tract and also of the kidney, where it causes a focal and segmental glomerulonephritis (p. 1076). Upper respiratory tract manifestations - particularly sinusitis, ulceration of the nasal mucosa and bleeding from the nose - are a

1 690

MCQ 13.19

FIG. 13.51 Chest X-ray showing extensive pulmonary haemorrhage in a patient with Goodpasture's syndrome.

common presentation, but unlike midline granuloniatous disease there is no facial ulceration. Middle-aged males are most frequently affected. Increasing breathlessness, haemoptysis and pleurisy occur with pulmonary involvement, which is present in 95% of cases. In most cases there is a low-grade pyrexia. In contrast to polyarteritis nodosa, hypertension is unusual. Wegener's granulomatosis is a multisystem disease and the widespread vasculitis may affect the eyes, skin, joints, heart and nervous system. In most instances it is the renal involvement, present in 85% of cases, that is critical. The chest X-ray shows nodules or masses which are commonly bilateral (Fig. 13.52), and which may be large and cavitate. Mediastinal or hilar adenopathy is rare. Endobronchial disease may be visible at bronchoscopy. In Wegener's, the ESR and CRP are usually raised and the platelet count may also be elevated. Eosinophilia is not a feature. The diagnosis is a clinical one (suggested by disease of the upper as well as the lower respiratory tract, and by renal involvement), supported by tissue biopsy demonstrating characteristic histology. Open lung biopsy gives the highest diagnostic yield, but diagnosis is often possible by biopsy of more accessible structures, such as the skin or upper airway mucosa. Renal biopsy confirms kidney involvement but the tissue obtained is commonly not diagnostic of Wegener's. A positive antineutrophil cytoplasmic antibody test (ANCA) strongly supports the diagnosis. Without treatment 80% of patients die within 1 year, the advancing renal disease being the major determinant of survival. Treatment with cyclophosphamide (l-2mg/kg/day), initially combined with steroids (prednisolone, 60mg daily), is often remarkably effective. Longterm remission is achieved by such therapy in up to 90% of patients, and in some cases with severe chronic renal impairment successful renal transplantation has been undertaken.

treatment of choice is local radiotherapy, and steroids are not helpful.

13

Polyarteritis nodosa Pulmonary involvement in classic polyarteritis nodosa is unusual, but is a feature of the variant of polyarteritis known as allergic granulomatosis or the Churg-Strauss syndrome (p. 661), characterized by a necrotizing vasculitis, granuloma formation, eosinophilia and asthma. Treatment is with highdose steroids, and cyclophosphamide if necessary. Pulmonary involvement occurs in many other systemic vasculitides, including giant cell arteritis, Behcet's syndrome, Henoch-Schonlein purpura, hypersensitivity vasculitis (usually drug-induced) and lymphomatoid granulomatosis.

PULMONARY ARTERIOVENOUS MALFORMATIONS AND SHUNTS PULMONARY ARTERIOVENOUS MALFORMATIONS

FIG. 13.52 Wegener's granulomatosis A Chest X-ray showing typical appearance, with multiple irregular large opacities and cavitation. B CT scan demonstrating cavitation,

Arteriovenous malformations (AVMs) can be single or multiple and are usually in the lower lobes. Most are congenital and gradually increase in size as the individual becomes older. AVMs are fed by a branch of the pulmonary artery and drain into a pulmonary vein. A relatively common cause is hereditary haemorrhagic telangiectasia, an autosomal dominant condition associated with epistaxis and gastrointestinal bleeding. AVMs cause hypoxia through shunting. Severe shunting causes breathlessness and patients are clubbed; many patients, however, are asymptomatic. Haemoptysis is common. Auscultation may demonstrate a bruit. Emboli, sometimes cerebral, are an important complication. Because of the anatomical shunt, the hypoxia does not correct with 100% oxygen. Hypoxia also intensifies when the patient stands. In most patients the chest X-ray is abnormal, usually showing a peripheral nodule; AVMs can be well demonstrated by angiography, CT and MRI. Screening of relatives is advisable, to detect asymptomatic AVMs and, by appropriate intervention, to avoid the complications of paradoxical emboli. Treatment is by obstructing the AVMs, usually by coils, thereby relieving the hypoxia and preventing emboli. Lung resection is occasionally necessary.

Midline granuloma This is a rare disease, now thought to be a T-cell lymphoma, in which there is destruction of the tissues of the nose and upper respiratory tract. The characteristic pathology is of granulomas; patients are most often middle-aged females. Nasal involvement is the most striking feature, with the development of septal perforation and saddle nose deformity. The disease is localized (unlike Wegener's) and there is no arteritis. Without treatment the disease is fatal. The

HEPATOPULMONARY SYNDROME Many patients with cirrhosis are hypoxic, some severely so. Hypoxaemia intensifies with exercise and when the patient is upright. In most cases the hypoxia is due to abnormal pulmonary arterial vasodilatation, causing ventilationperfusion mismatch and intrapulmonary shunting. Shunting through the lungs can be demonstrated by lung

691

scanning and echocardiography. The diagnosis is also supported by the worsening of hypoxia in the upright posture (orthodeoxia). Patients have a high resting cardiac output, low pulmonary vascular resistance and a mixed venous oxygen saturation. Some cases are greatly improved following liver transplantation.

PULMONARY FIBROSIS AND DIFFUSE INTERSTITIAL LUNG DISEASE

Clinical features of pulmonary fibrosis The conditions that cause diffuse pulmonary fibrosis (Table 13.44) share many clinical, radiological and pathophysiological features. The common presenting symptom is dyspnoea, which is mild at first but can progress to total incapacity. Often a persistent dry cough, particularly at night, is a feature. There may be symptoms of an associated systemic syndrome, including fever and weight loss during the acute phase. On examination the patient may have few physical signs outside the chest. However, when fibrosis is widespread there may be tachypnoea, cyanosis and, at a late stage, cor pulmonale. There may be finger clubbing, especially in patients with cryptogenic fibrosing alveolitis and asbestosis. Auscultation of the chest usually reveals predominantly basal inspiratory crackles, but there are exceptions (see below). The chest X-ray in the early stages shows a hazy, reticular and nodular infiltrate. As the lungs become more fibrotic, cystic spaces become evident and a 'honeycomb' pattern, sometimes with large bullae, develops. Lung function tests show a restrictive ventilatory defect (p. 608) with a reduced transfer factor and hypoxaemia.

OCCUPATIONAL LUNG DISEASE 1 The lung can be injured by the inhalation of dusts, fumes or other noxious substances at work in certain specific occupations. There are strict criteria for the diagnosis of these conditions and, in the UK, for determining the level of compensation. Pulmonary damage can occur in several ways: • • • • • •

Mechanical effects of dust retention Fibrogenesis (e.g. coal dust, silica) Granulomatous reactions (e.g. berylliosis) Toxicity (e.g. mercury vapour) Irritation of air passages (e.g. chlorine) Promotion of bronchial hyperreactivity (e.g. occupational asthma in laboratory workers who become hypersensitive to animal fur or urine). 1

692

MCQ 13.20

TABLE 13.44 Causes of pulmonary fibrosis Inhalation of dust or chemicals Organic dusts (e.g. avian proteins, moulds) Mineral dusts (e.g. asbestos, silica, coal dust) Chemicals (e.g. nitrogen dioxide, chlorine) Unknown causes Cryptogenic fibrosing alveolitis (syn. diffuse pulmonary fibrosis, idiopathic pulmonary fibrosis, usual or organizing interstitial pneumonia) Sarcoidosis Idiopathic haemosiderosis Histiocytosis X (Letterer-Siwe disease, Hand-Schuller-Christian disease, eosinophilic granuloma) latrogenic and poisoning Drugs (e.g. bleomycin, hydralazine, busulphan, amiodarone) Paraquat Radiation pneumonitis Oxygen toxicity Pulmonary involvement in systemic disease Rheumatoid arthritis Ankylosing spondylitis Systemic sclerosis Sjogren's syndrome Dermatomyositis Congenital Niemann-Pick disease Gaucher's disease Tuberous sclerosis Neurofibromatosis

Important occupational lung diseases include those due to inorganic dusts (e.g. silica, talc, asbestos, tin, iron), those due to organic dusts (e.g. cotton dust, maple bark dust) and those due to gases and fumes (e.g. chlorine, ammonia). Occupational exposure can also cause asthma, and lung and pleural malignancy.

Pneumoconioses due to inorganic dust inhalation The common pneumoconioses are listed in Table 13.45.

Coal miners' pneumoconiosis Inhalation of coal dust over a prolonged period damages the lung. The incidence varies with the type of coal being mined, occurring in about 12% of all miners but up to 50% in those who have mined anthracite for more than 20 years. For certification for compensation in the UK the diagnosis is radiological, not clinical. In the disease's early form these radiological changes consist of small nodules less than 1.5mm in diameter; this is termed simple pneumoconiosis.

13

TABLE 13.45 The common pneumoconioses Disease

Occupation

Cause

Pathology

Coalminers' pneumoconiosis Silicosis

Coal mining Mining, quarrying, metal grinding, stone dressing, etc.

Coal dust Silica

Asbestosis

Demolition workers, shipbreaking, lagging and brake lining manufacture etc. Electronics, atomic energy reactors, aero engines Rubber industry Smelting Arc welding

Asbestos

Nodular and progressive massive fibrosis Focal fibrosis, centrilobular emphysema, progressive massive fibrosis, associated tuberculosis Focal fibrosis, pleural calcification, bronchial carcinoma, pleural and peritoneal mesothelioma Granulomas and interstitial fibrosis Focal fibrosis Mineral deposition Mineral deposition

Berylliosis Talcosis Stannosis Siderosis

Beryllium Magnesium silicate Tin oxide Iron oxide

In a proportion of cases the nodules grow up to 10mm in size, are irregular, and coalesce to form large masses which may cavitate, commonly in the upper lobes: this is referred to as complicated pneumoconiosis or progressive massive fibrosis.

About 15% of patients have positive antinuclear antibodies and, in a small number, active rheumatoid arthritis is present. This is Caplan 's syndrome, and the chest X-ray shows rounded peripheral nodules in the lung fields 0.55.0cm in diameter. First described in coal miners, the same appearances have since been seen in other forms of pneumoconiosis.

Silicosis Although now less common as a result of protective measures, silicosis is still a major occupational hazard because of the ubiquitous distribution of free silica. Silicosis results from the inhalation of fine particles of silicon dioxide crystals or quartz particles. Those involved in mining, quarrying, metal casting, sandblasting and the pottery industry are at risk. Rarely, inhalation can cause an acute febrile illness with rapidly progressive dyspnoea and cyanosis, which can occur after a short exposure and lead to death in a few weeks. The more common chronic form of the disease is due to the highly fibrogenic potential of silica and presents as slowly progressive dyspnoea and cough, occasionally with haemoptysis. Coexisting tuberculosis is not uncommon. The disease may progress even after exposure ceases. The radiological appearances are of upper zone fibrosis with characteristic hilar 'eggshell' calcification.

Asbestos-related disease Asbestos is the generic name for several silicates, including chrysolite (90% of all world asbestos), crocidolite (blue asbestos), osmosite and anthopyllate. Fibres persist in the lung long after inhalation and cause characteristic pathological changes. Those at high risk include demolition workers, pipe laggers, brake-lining industry workers and boiler makers. The respiratory effects of asbestos are a progressive pulmonary fibrosis (asbestosis), pleural plaques, and malignancy of the lung and pleura.

FIG. 13.53 Asbestosis Bilateral pulmonary fibrosis of asbestosis. There is also pleural thickening, best seen adjacent to the lateral chest wall on the left and diaphragmatic pleural calcification on the right.

Asbestosis presents as progressive dyspnoea and cough, with finger clubbing and basal crackles. Like the other pneumoconioses it may progress to respiratory failure. Asbestos fibres are inhaled and phagocytosed, causing damage to cell membranes and the release of lysosomal enzymes. Radiologically there may be basal linear and irregular shadows which in the later stages progress to 'honeycombing' throughout the lung fields, with the cardiac silhouette becoming indistinct (Fig. 13.53). Lung function tests show a restrictive defect with reduced transfer factor. There are usually 'asbestos bodies' in the sputum or bronchial washings. Pleural plaques are an indication of previous exposure to asbestos. They do not indicate lung disease or any increased risk over similarly exposed subjects of developing a tumour. They are usually basal or diaphragmatic and frequently calcify. CT scanning of the thorax is very useful in picking up early lesions, demonstrating the extent of pleural involvement, and distinguishing pleural from parenchymal disease.

693

TABLE 13,46 Some causes of extrinsic allergic alveolitis Type

Disease

Source

Antigen

From fungi

Farmer's lung

Mouldy hay

Micmpolyspora faeni Thermoactinomycetes vulgaris Thermoactinomycetes sacherii

Mushroom worker's lung Bagassosis Ventilator pneumonitis Malt worker's lung Maple bark stripper's lung

Mushroom compost Mouldy bagasse Air conditioning and hot air systems Mouldy barley Mouldy maple bark

Aspergillus clavatus Cryptostroma (coniosporium corticale)

From birds

Bird fancier's lung

Pigeon/budgerigar/hen/parrot droppings

Avian serum proteins in droppings

From mammals

Pituitary snuff taker's lung Rat handler's lung

Porcine pituitary powder Rat droppings

Serum protein/pituitary antigens Rat serum protein

Bronchogenic carcinoma can be caused by asbestos exposure; cigarette smoking in asbestos workers greatly increases the risk, exerting a multiplicative rather than an additive effect. The tumours are usually adeno- or squamous cell carcinomas. Mesotheliomas of the pleura and peritoneum are associated with asbestos exposure (p. 717).

in some cases, by improvement. Patients with relatively acute disease can benefit from corticosteroid therapy, but such treatment is ineffective in the majority of cases. Lung transplantation is appropriate in selected patients. EXTRINSIC ALLERGIC ALVEOLITIS

Berylliosis Berylliosis differs from other pneumoconioses in that the initial pathological response to inhaling beryllium is a granulomatous one, which progresses to fibrosis. Histologically it may be difficult to distinguish from sarcoidosis. It can occasionally cause an acute pneumonic type of illness. Those at risk include workers in the atomic power and aero-engine industry.

Other inorganic dusts Other inorganic dusts vary in their ability to generate fibrosis; iron and iron oxides in arc welding and silver finishing (siderosis) and tin oxide in smelting (stannosis) are relatively non-fibrogenic, whereas magnesium silicate in the rubber industry (talcosis) can be a potent cause of fibrosis. Other dusts cause mucus hypersecretion and airway irritation with or without pneumonitis, e.g. cement, graphite, cadmium and many others.

Management In the management of lung disease caused by inorganic dusts, stopping further exposure is crucial and is followed,

1

694

Case 13.2

2 Fig. 13.38

3

Fig. 13.39

1

This group of conditions is caused by the inhalation of organic dusts (Table 13.46). Chronic exposure to dusts that contain antigen sensitizes the individual, a precipitating antibody is produced, and granulomas appear in the lung which heal by fibrosis. Clinically, the patient may present with an acute illness 6-8 hours after heavy antigen exposure, with fever, 'flulike' symptoms, cough, dyspnoea and crackles. These resolve spontaneously in 24-48 hours if there is no further antigen exposure. The clinical features reflect the pathogenic mechanism, which is primarily an IgG-mediated 'late' type III hypersensitivity or Arthus' reaction to the antigen, although there is probably also a cell-mediated response. The result of repeated exposure is granuloma formation and eventually extensive fibrosis. More commonly the presentation is with established fibrosis, with dyspnoea and dry cough. Finger clubbing is uncommon, in marked contrast to cryptogenic fibrosing alveolitis. Inspiratory crackles and squeaks are characteristic. Radiologically, the changes are predominantly upper zone, with nodular and linear shadows and volume loss. 0 Lung function tests show reduction in lung volumes and transfer factor. Precipitating antibodies are present in the blood, and intrar dermal skin tests, read at 6 hours, are available for some of the causes and are especially useful in the diagnosis of bird fancier's lung. Bronchial challenge tests, with measurement of temperature, lung volumes and gas transfer, are occasionally valuable. The treatment of these conditions is first to stop exposure to the antigen. For the acute type of allergic

alveolitis this is usually sufficient. However, when chronic exposure has led to granulomatous change in the lung, treatment with corticosteroids is required. In advanced fibrotic disease, as with all end-stage fibrotic lung diseases, there is relatively little that can be done apart from considering lung transplantation.

13

CRYPTOGENIC FIBROSING ALVEOLITIS Cryptogenic fibrosing alveolitis (CFA), a disorder of unknown aetiology, is characterized pathologically by migration of large mononuclear cells into the alveolar spaces with later thickening of alveolar walls, with fibrosis. The mononuclear cellular response is associated with acute symptoms, whereas the alveolar thickening and fibrosis are responsible for the more chronic clinical picture. CFA is an unusual disease but its incidence is increasing. Overall, deaths have increased threefold in the UK over a 20-year period. The pathogenesis of the condition is poorly understood. It is considered to be an altered state of host responsiveness to an as yet unknown stimulus. CFA is about twice as common in smokers, and is associated with exposure to metal and wood dust. An alteration of cell-mediated immunity has been proposed, as stimulated T lymphocytes increase fibroblastic activity. Activation of complement with consequent leukocyte aggregation may play a part. The pathological picture that emerges is, at first, an intraalveolar inflammatory response, with many mononuclear cells and an increase in reticulin, collagen and elastin. As the disease progresses there is fibrosis and destruction of lung architecture, with the formation of cystic spaces. Characteristically, CFA is a disease of late middle age and affects males twice as frequently as females. The acute variant may present with rapidly progressive dyspnoea and cough, widespread crackles and progressive respiratory failure within a few months. The chronic (and much more common) form presents with dyspnoea and dry cough, and in approximately 20% of cases polyarthralgia. About twothirds of the patients have finger clubbing and almost all have basal crackles. The condition may be associated with a variety of systemic diseases: rheumatoid arthritis, Sjogren's syndrome and systemic sclerosis. Other known causes of pulmonary fibrosis, such as the pneumoconioses or sarcoidosis, should be excluded, but with end-stage pulmonary fibrosis it is often difficult, even at postmortem, to determine the aetiology. The chest X-ray shows fine nodular and irregular shadowing (Fig. 13.54) and eventually 'honeycombing'. Lung function tests show a restrictive pattern with reduced transfer factor. Antinuclear antibodies are present in 45% of patients and rheumatoid factor in 20-25%. BAL usually yields a predominantly polymorphonuclear leukocyte pattern. To make a definitive diagnosis a lung biopsy may be necessary, by thoracoscopy or thoracotomy. However, the appearance of CFA on high-resolution CT scans is characteristic, 0 reducing the need for surgical biopsy.

FIG. 13.54 Cryptogenic fibrosing alveolitis (CFA) Diffuse pulmonary fibrosis, wifh linear and nodular opacities throughout both lungs. In less severe CFA the radiological abnormalities are commonly maximal at the bases, with differential volume loss of the lower lobes.

Patients who have a very active cellular histology respond better to treatment than those with established fibrosis.

Treatment The usual treatment is with oral corticosteroids and the response is variable, with only 10% improving substantially. Treatment may need to be continued over several years. Some patients respond to cyclophosphamide, which can also be used, as can azathioprine, as a steroid-sparing agent. It is unlikely that this treatment substantially affects the long-term prognosis. There is recent evidence that the subcutaneous administration of interferon-y1b over a period of a year may improve lung function and gas exchange. The progression of CFA is variable: some patients with very acute, severe disease (the 'Hamman-Rich' syndrome) may die within months. Overall, approximately 50% survive for 5 years and 25% for 10 years. There is a tenfold increase in the incidence of bronchogenic carcinoma. In appropriate patients fibrosing alveolitis can be successfully treated by lung transplantation.

Idiopathic interstitial pneumonia The term idiopathic interstitial pneumonia (IIP) embraces a group of diffuse parenchymal lung diseases which can be distinguished on the basis of clinical history, physical signs, pathology and chest radiology, particularly the recent use of high-resolution computerized tomography (HRCT). They can be described separately from other diffuse interstitial lung diseases, which include environmental and occupational lung disease and lung pathology associated with coronary and vascular disease, and those with dis-

695

tinctive pathology such as sarcoidosis. The recent development of minimally invasive surgery and thoracoscopic lung biopsy and its clinical correlations, along with the rapid development and sophistication of HRCT, has helped to further define and classify these conditions. Bronchiolitis

The bronchioles are the 'silent area' of the lung. Abnormalities in these airways may be primary or as a result of extension of diseases of the bronchi (such as chronic bronchitis or bronchiectasis) or of the lung parenchyma (such as cryptogenic organizing pneumonia). Diagnosis of these diseases is difficult because of the relative inaccessibility of the small airways and the nonspecificity of clinical and radiological features. Recent advances in HRCT and specific physiological tests of small airways disease have greatly added to our knowledge. As with idiopathic interstitial pneumonias, classification of diseases of the small airways has been confusing, with differences of emphasis from clinicians, pathologists and radiologists. The terms bronchiolitis/obliterative bronchiolitis/bronchiolitis obliterans have been used indiscriminately for what we now know are a wide range of different conditions. For simplicity bronchiolitis can be subdivided into 'proliferative' and 'constrictive' types, which can be accurately described in terms of the clinical presentation, radiology and pathology. Bronchiolitis obliterans, organizing pneumonia (BOOP)/cryptoqenic organizing pneumonitis (COP} Initially this condition was thought to be a disease of the small airways but is now recognized to be primarily a condition of the alveolar spaces. It is separate from obliterative bronchiolitis and has a characteristic histological pattern.

IATROGENIC LUNG DISEASE

There is a wide variety of drugs and physical agents, including oxygen and ionizing radiation, that can cause temporary or permanent lung damage. The drugs most commonly responsible are cytotoxic agents (Table 13.47; see also p. 161). Bleomycin toxicity is commonest, occurring in 3-6% of patients treated with the drug; it is dose dependent and there is increased toxicity in the elderly and in those who have received thoracic radiotherapy or combination chemotherapy. Likewise, the pulmonary side-effects of the nitrosoureas are more common in the elderly. The changes

Honeycomb lung

TABLE 13.47 Cytotoxic drugs causing pulmonary toxicity

This is a radiological description of a group of conditions that cause widespread cystic or 'honeycomb' appearances on chest X-ray. A 'honeycomb' appearance can be seen as the end stage of pulmonary fibrosis of any cause. Rare causes include Letterer-Siwe disease, tuberous sclerosis, Hand-Schuller-Christian disease, neurofibromatosis and eosinophilic granuloma. Patients with honeycomb lung tend to develop pneumothoraces. The long-term prognosis is poor, with respiratory and cardiac failure supervening.

Drug

Toxic effects

Bleomycin

Interstitial pneumonitis Pulmonary fibrosis

Nitrosoureas

Interstitial fibrosis

Mitomycin

Pneumonitis and fibrosis

Busulfan

Diffuse pulmonary fibrosis

Melhotrexate

Pneumonitis Blood eosinophilia

Idiopathic pulmonary haemosiderosis

Cyclophosphamide

Interstitial fibrosis

Idiopathic pulmonary haemosiderosis is a rare condition of unknown aetiology characterized by recurrent episodes of

Procarbazine

Pneumonitis Pleural effusion Blood eosinophilia

Chlorambucil

Pulmonary fibrosis

Melphalan

Pneumonitis Pulmonary fibrosis

1

696

intrapulmonary haemorrhage, pyrexia, haemoptysis, and a secondary iron deficiency anaemia. Most cases occur in childhood; patients may die during an acute episode, but more commonly recurrent episodes eventually cause severe pulmonary fibrosis and respiratory failure. Some authorities consider that there is an association with Goodpasture's syndrome (p. 1076). During acute episodes there may be crackles at the lung bases and occasionally hepatosplenomegaly. The chest X-ray during episodes of haemoptysis may show patchy, usually bilateral, shadowing in the middle zones. Following acute pulmonary haemorrhage (of any cause) the diffusion capacity of the lung (TLCO), measured in the lung function laboratory, is increased. Management consists of treatment of the iron deficiency and corticosteroids and/or azathioprine during acute episodes, although their value is as yet unproven. At 3 years from diagnosis one-third of patients will have died, one-third will have active symptomatic disease and onethird will be symptom-free.

MCQ 13.21

TABLE 13.48 Non-cytotoxic drugs which cause pulmonary toxicity Amiodarone Nitrofurantoin Penicillin Erythromycin Tetracycline Streptomycin

Neomycin Ethionamide Isoniazid Suiphonamides ACE inhibitors

in the lung induced by methotrexate may resolve despite continuing treatment with the drug. Cyclophosphamide causes a pneumonitis and fibrosis that can respond remarkably well to corticosteroid therapy. There are also a large number of non-cytotoxic drugs that can cause pulmonary toxicity (Table 13.48). The largest single group is antibiotics, and the most common of these is nitrofurantoin. The antiarrhythmic agent amiodarone is an increasingly important cause of pulmonary fibrosis; frequently the diagnosis is delayed because breathlessness is assumed to be due to cardiac disease rather than to lung damage. ACE inhibitors can cause a chronic dry cough. Some antibiotics (e.g. penicillin, erythromycin, etc.) can cause acute asthma. Pulmonary eosinophilia is associated with drugs such as isoniazid and sulphonamides.

Radiation injury Patients who receive radiation treatment to the thorax may sustain long-term lung injury. The acute injury is characterized pathologically by pulmonary vascular congestion, alveolar oedema and the formation of hyaline membranes. Patients become symptomatic at 4-8 weeks. This may progress, over a period of 6-12 months, to pulmonary fibrosis. Clinically, these processes cause a dry cough, chest wall pain and, depending on the extent of the damage, breathlessness. Radiation injury occurs most frequently in patients receiving treatment for carcinoma of the breast, lymphomas and carcinoma of the bronchus. The chest X-ray in the acute pneumonitis stage shows localized infiltration of the lung in the path of the radiation beam; when fibrosis has developed there is dense shadowing and volume loss. Symptoms and radiological appearances are poorly correlated: some patients with minor infiltrates may be severely dyspnoeic, whereas others with widespread X-ray changes may be asymptomatic. Treatment with corticosteroids may be useful in the acute phase, but once fibrosis is established no treatment is effective. Rarely, phrenic nerve damage causes diaphragm dysfunction. Oxygen toxicity The use of high levels of inspired oxygen causes lung injury indistinguishable from the acute respiratory distress syndrome (ARDS). The injury is related to both the inspired

oxygen fraction (FiO2) and the duration of exposure; a safe level has not yet been established, but an FiO2 of more than 0.7 over a period of more than several days may be associated with X-ray infiltrates and histological change. Toxic oxygen radicals (in particular superoxide anion, O2-) are thought to be important in oxygen-induced lung injury (see also ARDS below).

13

FURTHER READING ON PULMONARY FIBROSIS AND DIFFUSE INTERSTITIAL LUNG DISEASE British Thoracic Society 1999 The diagnosis, assessment and treatment of diffuse parenchymal lung disease in adults. Thorax 54 (suppl. 1). DuBois R M 1994 Diffuse lung disease: an approach to management. Br Med J 309:175-179. Kalica A R, Hunninghake G W 1995 Approaches to the treatment of pulmonary fibrosis. Am J Respir Crit Care Med 151:915-918. Pham S H 1995 New strategies for treatment of pulmonary fibrosis. Thorax 50:415-421. Raghu G 1995 Interstitial lung disease: a diagnostic approach. Am J Respir Crit Care Med 151:909-914. Ziesche R, Hofbaver E, Wittmann K, Petkov V, Block L-H 1999 A preliminary study of long-term treatment with interferon gammaIb and low-dose prednisolone in patients with idiopathic pulmonary fibrosis. N Engl J Med 341:1264-1269.

THE ACUTE RESPIRATORY DISTRESS SYNDROME 1 The term acute respiratory distress syndrome (ARDS) is used to describe a pulmonary condition characterized by progressive hypoxaemia, chest X-ray infiltrates and reduced lung compliance in the presence of a normal left atrial pressure. In many respects the clinical presentation resembles the respiratory distress syndrome of the newborn. The acute respiratory distress syndrome is sometimes called non-cardiogenic pulmonary oedema. It results from acute damage to the alveoli, in response to a wide and seemingly unrelated variety of initiating events (Table 13.49). The pathogenic mechanisms that produce the syndrome (Fig. 13.55) are due to diffuse damage to the pulmonary capillary endothelium, which results in increased 'leakiness' across the endothelial/epithelial barrier, resulting in the passage of plasma and red blood cells into the interstitial space and thence into the alveoli, causing alveolar flooding. The mechanisms responsible for the capillary endothelial damage are poorly understood. Activation of complement with leukocytosis and release of lysosomal enzymes, and interactions between mediators such as leukotrienes, prostaglandins, neuropeptides and superoxides, are all considered to contribute. Damage is therefore due to a complex series of events involving cellular, chemical, neural and vascular responses, all combining to produce the clinical syndrome. The clinical features are outlined in Table 13.50. A typical radiographic progression is shown in Figure 13.56.

697

TABLE 13.49 Conditions associated with the acute respiratory distress syndrome

TABLE 13.80 Clinical features of the acute respiratory distress syndrome

Shock (traumatic, septic etc.)

Initiating event

Infection Pneumonias (viral, bacterial, Pneumocystis carinii) Gram-negative sepsis

Dyspnoea and cough

Trauma (fat emboli, head injuries, lung contusion)

Crackles over lung fields Progressive hypoxaemia Chest X-ray infiltrates

'Stiff lungs' Normal left heart pressures With progression Sepsis Failure of other organs (kidneys, liver etc.)

Aspiration (gastric contents, fresh or salt water) Drug overdose (narcotics, barbiturates) Metabolic disorders (pancreatitis, uraemia) Inhaled toxins (02, smoke, corrosives) Blood disorders (disseminated intravascular coagulation, massive blood transfusion) Miscellaneous Paraquat ingestion Radiation Eclampsia Raised intracranial pressure Seizures High altitude

Management There are as yet no definite predictors as to which patients with the conditions listed in Table 13.48 will develop ARDS. However, the warning signs are deteriorating gas exchange, worsening chest X-ray, and, if the patient is being mechanically ventilated, increasing inflation pressures as the lungs become more 'stiff. Early diagnosis and effective treatment can shorten the course, minimize the long-term sequelae and improve the prognosis of ARDS. Unfortunately, early treatment cannot salvage all patients, many of whom die despite vigorous therapy. As measurements of pulmonary artery pressure, pulmonary artery wedge pressure and mixed venous O2 content are critical, a Swan-Ganz catheter should be passed into the pulmonary artery (p. 726). For the diagnosis of ARDS to be secure, the clinical features of the disorder are combined with a normal pulmonary artery wedge pressure (normal PAWP = 8-12 cm H2O). Once the diagnosis of ARDS has been made the patient is usually so hypoxic that endotracheal intubation and mechanical ventilation is required. The following aspects of the patient's therapy should be considered:

698

FIG. 13.55 Pathogenesis of the acute respiratory distress syndrome (ARDS) The many causes of ARDS have in common the damage to the alveolar-capillary membrane. The pathological changes that follow cause profound hypoxaemia. (FRC = functional residual capacity.)

• Fluid balance • Oxygenation • Drug therapy.

Excessive fluid administration may increase the pulmonary capillary hydrostatic pressure and thus cause further leakage of fluid into the lung interstitium and alveoli; if the patient is allowed to become fluid depleted, however, systemic and pulmonary perfusion may be inadequate, causing renal and other organ damage. A reasonable approach is to administer sufficient fluid to maintain a PAWP of 6-8 cm H2O, and if further circulatory support is required, to administer an inotropic agent such as dobutamine or isoprenaline. Fluid can be given in the form of 5% dextrose solution, or as a crystalloid, and if the patient is severely hypoalbuminaemic, intravenous salt-free albumin should be administered.

Fluid balance Fluid balance is illustrated in Figure 13.57. The physician treating the patient with ARDS is faced with a dilemma.

The first step in improving arterial hypoxaemia is to add supplementary O2. Very high concentrations of O2 may be

Oxygenation

13

FIG. 13.57 The relationship of pulmonary extravascular water and pulmonary capillary pressure In normal circumstances an increase in pulmonary extravascular water (and pulmonary oedema) does not occur until pulmonary artery 'wedge' pressure exceeds 3.3 kPa (25 mmHg). However, increased permeability (as in ARDS), and hypoalbuminaemia (common in intensive care unit patients) can cause pulmonary oedema at normal wedge pressures. (From Hopewell P C, Murray J F The adult respiratory distress syndrome. Ann Rev Med 1976; 27:243-356.)

FIG. 13.56 Acute respiratory distress syndrome The X-rays are of a young man who had sustained severe trauma and blood loss in a road traffic accident; the films cover a period of 5 days from a relatively normal X-ray A to bilateral infiltrates B to bilateral 'white-out' C accompanied by severe hypoxaemia. A Swan-Ganz catheter for measurement of pulmonary artery 'wedge' pressure (as a reflection of left atrial pressure) can be seen in situ on X-ray C. The patient died shortly after the last film.

used initially (up to 100%) but these levels should not be sustained, as they may of themselves cause pulmonary damage. Positive end-expiratory pressure (PEEP) has gained widespread acceptance as an effective method of ventilating patients with ARDS. Positive pressure (usually 5-15 cm H2O) is applied at the end of expiration, thereby preventing airway closure, limiting areas of atelectasis and reducing right-to-left shunting of blood through underventilated areas of lung. The net result of PEEP is usually an increase in Pao2 without increasing the inspired O2 concentration. However, PEEP is not without problems (see Ch. 14, p. 740). By increasing intrathoracic pressure it may reduce venous return to the heart and impair cardiac output. Inotropic support may be necessary, as is careful maintenance and monitoring of intravascular fluid volume. The higher the level of PEEP administered, the greater the risk of inducing trauma to the lung and pneumothorax. The optimum level of PEEP for individual patients can be determined by starting at low levels, say 5cm H2O, and working upwards in small increments, measuring at each level the effect of PEEP on gas exchange, lung compliance, cardiac output and mixed venous-arterial oxygen content difference (see p. 675). In ARDS the lung is 'stiff, a substantial proportion of alveoli are filled with fluid and the accessible lung volume is small. Attempts to deliver a large tidal volume therefore produce high peak inflation pressures, excessive stretching of parts of the lungs, and therefore severe damage. Patients are ventilated by techniques that keep inflation pressures as low as possible. A combination of a prolonged inspiratory phase (inverse inspiratory-expiratory ratio) and pressure-limited ventilation is commonly used. To keep airway pressures and tidal volumes down it is often necessary to allow the arterial CO2 to rise (permissive

699

TABLE 13.51 Relative frequency of primary pulmonary tumours

FIG. 13.58 CT scan in severe ARDS; prolonged mechanical ventilation has produced extensive barotrauma, with numerous bullae within the lungs and loculated pneumothoraces. 1

hypercapnia); provided this is gradual and not excessive it is well tolerated. Despite every effort it is sometimes impossible to ventilate patients with severe ARDS without causing severe lung trauma (Fig. 13.58). In ARDS, abnormalities are most severe in the dependent parts of the lungs, and oxygenation can be improved by periodically ventilating patients in the prone posture. Extracorporeal membrane oxygenation (ECMO) is a technique by which blood is continuously removed from the circulation, passed through a membrane oxygenator, and then returned. The results of controlled studies to date of ECMO in ARDS are discouraging and further technical advances are necessary before this can be considered a realistic option for treatment, except in a small number of selected patients. Drug therapy Sepsis, not respiratory failure, is now the commonest cause of death in patients with ARDS. Pulmonary and abdominal sepsis is particularly common. Steroids may be helpful in the proliferative phase of the illness, but have little effect on established disease. Inhaled nitric oxide is increasingly used as a pulmonary vasodilator, to improve ventilation-perfusion matching and cardiac function. Descriptive studies have shown improvements in oxygenation but controlled trials have yet to be reported, and caution is required when considering NO therapy. The outlook for patients with ARDS remains serious, although it has improved. There is a 40% mortality, and if

1

700

Fig. 13.40

2

MCQ 13.22

Tumour

%

Bronchial carcinoma Alveolar carcinoma Bronchial adenoma Hamartoma Other tumours

95 1-2 1-2 1 1

failure of an organ in addition to the lungs (e.g. the kidneys) supervenes most patients die. Of those who survive, most recover normal lung function; only a small proportion, usually those who have required ventilatory support for many weeks, develop bronchiectasis or pulmonary fibrosis.

FURTHER READING ON THE ACUTE RESPIRATORY DISTRESS SYNDROME Bernard G R et al. 1994 The American-European Consensus Conference on ARDS. Am J Respir Crit Care Med 149:818-824. Tobin M J 2000 Culmination of an era in research of the Acute Respiratory Distress Syndrome. N Engl J Med 342:1360-1361.

LUNG TUMOURS Carcinoma of the bronchus is one of the major causes of death in western countries. The 5-year survival for this common disease is 6-12%. In addition, the lungs and mediastinum are common sites for metastatic malignant disease, with about 30% of cancer patients developing lung secondaries, in half of whom metastases are limited to the lungs. Intrathoracic disease is relatively common in Kaposi's sarcoma. By far the most common primary lung tumour is bronchial carcinoma (Table 13.51).

BRONCHIAL CARCINOMA O In the 19th century bronchial carcinoma was unusual: by the mid-20th century, associated with the widespread habit of cigarette smoking, it had reached epidemic proportions. Lung cancer accounts for more than one-sixth of all cancer deaths worldwide. In the UK, carcinoma of the bronchus is the commonest malignancy in males and is approaching the importance of breast cancer in females; in all, 35 000 patients die each year from this disease (p. 145). The UK has a higher death rate from bronchial carcinoma than any other country in the world (p. 145). In America the incidence of the disease has reached a

plateau; in the UK deaths are starting to fall. Lung cancer rates in many developing countries are rising steeply, reflecting cigarette consumption. Screening programmes with routine chest radiographs or sputum cytology have not been shown to improve survival. There is, however, some recent evidence that early operable asymptomatic lung cancers can be picked up by the use of spiral CT scanning or by molecular genetic markers of cancer in sputum samples. If these reports are confirmed, screening of smokers may possibly produce a better outlook.

Aetiology Atmospheric pollution and industrial exposure to dusts, particularly asbestos, increase the incidence of lung cancer, but the dominant causative agent is tobacco smoke, which is responsible for 90% of cases. The disease is unusual in non-smokers and the increased incidence in smokers is related to the number of cigarettes consumed (Fig. 13.59). The incidence in ex-smokers slowly falls towards that of non-smokers. Passive smoking causes lung cancer in nonsmokers, and in the UK is responsible for approximately 300 deaths each year.

Pathology Bronchial carcinomas are conventionally classified into four main histological groups (Table 13.52). Squamous cell carcinoma Squamous cell carcinomas are most common. The neoplastic epithelial cells produce keratin and frequently arise

from major bronchi. The tumours are thought to originate in areas of squamous metaplasia which progress to carcinoma in situ and then to invasive carcinoma. Squamous tumours frequently cavitate; when peripheral they often invade the chest wall. Hypercalcaemia is a relatively common complication.

13

Adenocarcinomas Adenocarcinomas arise both centrally and peripherally, from the bronchial epithelium or submucosal glands. They do not cavitate and their histology is characterized by tubules and gland-like structures, often producing mucin. Adenocarcinomas are related to cigarette consumption but also occur in non-smokers. There can be diagnostic difficulty, as an apparent primary tumour may prove to be a metastasis from a tumour of the gastrointestinal tract, kidney, ovary, pancreas, thyroid or breast. Tumour markets may help to distinguish primary from secondary tumours. Small cell carcinoma Small cell (oat cell) carcinomas show differentiation into cells with amine-precursor uptake and decarboxylation (APUD) characteristics. Electron microscopy frequently shows neurosecretory granules. The malignant cells are small with little cytoplasm and dark, often oval nuclei. The tumours are usually central, highly invasive and disseminate rapidly. They may secrete a wide range of hormones, particularly ADH and ACTH. Large cell (undifferentiated) carcinoma Large cell undifferentiated tumours constitute a heterogeneous group, including giant cell and clear cell carcinomas, the cells of which show no evidence of maturation. They frequently arise centrally, invade the mediastinum and disseminate widely. Studies of the natural history of bronchial carcinoma and analysis of the time taken for tumours to double in size suggest that by the time a tumour can be diagnosed it may have been present for many years, and has frequently metastasized (Table 13.51).

Clinical features Symptoms of local disease Bronchial carcinomas frequently arise in major bronchi, and therefore common problems include cough, haemoptysis, breathlessness, wheeze and stridor. Some patients have diffuse, poorly localized chest pain, and many have chest infections. Of these symptoms, haemoptysis, progressive breathlessness and persistent respiratory infection are the most important.

FIG. 13.59 Mortality from lung cancer in doctors related to previously recorded smoking habits (Adapted from the classic paper Doll R, Hill A B 1964 Mortality in relation to smoking: 10 years observation of British doctors. Br Med J 1:1399-1414)

Haemoptysis There are many causes of haemoptysis (Table 13.2, p. 612), but in a middle-aged or elderly smoker carcinoma must always be excluded. A normal chest X-ray is unusual in bronchial carcinoma, but does not exclude a tumour.

701

TABLE 13.52 Natural history of untreated lung cancer Years from malignant change to: Doubling time (days)

Earliest diagnosis (1cm)

Usual diagnosis (3cm)

Death (10cm)

13.2

8.4 15,4

17.6

86

7.1

8.2

9.4

29

2.4

2.8

3.2

Histology

%

Squamous Adenocarcinoma Large cell undifferentiated Small (oat) cell

35

88

7,2

21 19

161

25

9.6

From Geddes D M 1979 The natural history of lung cancer: a review based on rates of tumour growth. BrJDisChest73:1-17.

FIG. 13.60 Symptoms and signs of intrathoracic tumour invasion and metastases

Breathlessness This is common in smokers who have developed generalized airways obstruction, but progression is usually gradual; accelerated breathlessness commonly occurs as a tumour narrows a major airway. Such narrowing may produce a localized wheeze, and occasionally stridor if the trachea or major bronchi are affected. Breathlessness may also be due to the accumulation of a malignant pleural effusion or the development of lymphangitis. Chest infections These are not uncommon in chronic heavy smokers and usually resolve quickly when treated with antibiotics. When infection occurs distal to a tumour, however, it frequently persists or relapses. For this reason failure of a pneumonia to resolve is a common presentation of bronchial carcinoma. Symptoms due to intrathoracic infiltration The common sites of local infiltration and associated clinical problems are shown in Figure 13.60.

702

Extrathoracic metastasis In bronchial carcinoma general symptoms such as lethargy,

anorexia and weight loss frequently signify disseminated disease. The most important sites of secondary spread are the liver, bone and brain, but virtually all parts of the body can be affected. In small cell carcinoma the bone marrow is involved in 10-20% of patients presenting with advanced disease, and can produce a leukoerythroblastic anaemia, but in most patients any anaemia is due to the non-specific effects of malignancy (Fig. 23.17, p. 1227). Non-metastatic manifestations Endocrine and metabolic changes occur in 10-15% of cases, but cause symptoms much less frequently. Hypercalcaemia Hypercalcaemia (p. 965) occurs in 6-7% of tumours. In many cases of squamous carcinoma this is due to the production of parathyroid hormone-related peptide elaborated by the tumour. In some patients hypercalcaemia is due to extensive bone metastases. Hyponatraemia The secretion of ADH-like peptides is nearly always associated with small cell cancer, and is a feature in 10% of such tumours (p. 1103). It results in a dilutional hyponatraemia.

Cusking's syndrome The ectopic production of ACTH-like peptides by small cell tumours is common but only occasionally causes symptoms (1% of all cases) (p. 932).

13

Clubbing Clubbing (Fig. 13.9, p. 619) is most frequently seen with squamous cell carcinoma (more than 50%) and is unusual in small cell tumours. Occasionally clubbed patients have hypertrophic pulmonary osteoarthropathy (HPOA) (p. 619). Neurological manifestations These are not common (less than 2%) and most frequently complicate small cell carcinoma. They are described on page 1434. Neurological manifestations can predate any obvious tumour. They may remit spontaneously, but usually progress and, except for rare cases, are little affected by treatment of the associated tumour.

Diagnosis Carcinoma of the bronchus is so common and has such a variety of symptoms and signs that it should be considered in most heavy smokers presenting to a general physician. Diagnosis is usually straightforward and the first important investigation is a chest X-ray. Indeed, 5% of cases are discovered when a chest X-ray has been taken for unrelated reasons. In a smoker, clubbing, haemoptysis, persistent chest infection or weight loss demand an immediate X-ray of the chest. Subsequent CT scanning often provides useful additional anatomical information, which can aid diagnosis and help in accurate staging.

FIG. 13.61 Carcinoma of the bronchus The left hilum is enlarged and more dense than the right. There are strands radiating from the hilum, and this spiculated appearance is typical of tumour.

Common radiological abnormalities Unilateral hilar enlargement Unilateral hilar enlargement (Fig. 13.61) is due to tumour and, frequently, involved lymph nodes. The hilum appears more dense than the opposite side, and the outer border of the hilum is frequently convex. Phrenic nerve invasion produces unilateral elevation of the diaphragm, and left recurrent laryngeal nerve involvement produces vocal cord paralysis with hoarseness. Peripheral solid tumour mass The tumour may be seen as a nodule, a large rounded mass or an irregular density (Fig. 13.62). Streaky shadows often extend into the surrounding lung, particularly towards the hilum, which may be abnormally bulky due to lymphadenopathy. Peripheral cavitating tumour mass A peripheral cavitating tumour mass is characteristic of squamous cell carcinomas. The cavitation is due to necrosis within the tumour, and the cavity wall is irregular and thick. However, a certain diagnosis is not possible on radiological criteria alone, and the differential diagnosis can include tuberculosis, lung abscess (Fig. 13.26), pulmonary infarction complicated by infection, vasculitic lesions and a rheumatoid nodule. Sometimes cavitation

FIG. 13.62 Carcinoma of the bronchus: peripheral (squamous cell) carcinoma The left hilum is enlarged due to adenopathy. Most pulmonary mass lesions more than 4cm diameter are malignant tumours.

703

SUMMARY 8 Common radiological abnormalities in bronchial carcinoma Unilateral hilar enlargement Peripheral solid tumour mass Peripheral cavitating tumour mass Volume loss/collapse affecting a lobe or lung

Abnormal mediastinum Lymphatic invasion Apical (Pancoast) tumour Pleural effusion

occurs in the consolidated lung distal to an obstructing tumour. Volume loss/collapse Bronchial carcinoma frequently arises in a major bronchus and gradually narrows the airway, leading to volume loss in a lung or lobe on the chest X-ray. Volume loss is therefore a critically important radiological sign, particularly if progressive. Eventually the tumour causes the collapse of a lobe or lung, producing an opacity on the chest X-ray (Fig. 13.12). Abnormal mediastinum Bronchial carcinoma frequently involves the mediastinum, and tumour infiltration as well as lymph node enlargement makes the mediastinal outline abnormal (Fig. 13.63). This is common in small cell tumours. Abnormalities of the mediastinum are well demonstrated by a CT scan.

FIG. 13.63 Carcinoma of the bronchus Central (small cell) carcinoma, lateral and inferior to the aortic knuckle. The usual outline of the aortic knuckle and configuration of the left hilar structures is lost. There is a secondary deposit in the left mid-zone.

Lymphatic invasion 1 2 The widespread involvement of the pulmonary lymphatics by tumour produces lymphangitis carcinomatosa (Fig. 13.64). This also occurs in malignant tumours of the stomach, pancreas and breast. Lymphatic invasion is well demonstrated by CT scans. 3 Apical (Pancoast's) tumour Apical tumours frequently invade the ribs, sympathetic chain (Homer's syndrome, p. 1295), brachial plexus and chest wall (Fig. 13.65). Apical tumours are well seen on CT scans or MRI. 4 Pleural effusion A common radiological manifestation of bronchial carcinoma at presentation is a pleural effusion, usually of large volume, which rapidly reaccumulates following aspiration, is commonly bloodstained, and often contains malignant cells. Positron emission tomography (PET) scanning PET scanning, which reflects metabolic activity in malignant tissues, accurately differentiates between benign and malignant nodules and is useful in staging lung cancer and identifying metastases (see p. 155).

704

1

Case 13.3

2

3

Figs 13.42-13.44

Fig. 13.41 4

Figs 13.45-13.47

FIG. 13.64 Carcinoma of the bronchus: lymphangitis carcinomatosa There are widespread linear and nodular opacities due to lymphatic infiltration. Basal septal lines are radiologically similar to those seen in left heart failure. Patients with lymphangitis usually have severe breathlessness.

Confirmation of diagnosis In addition to confirming the diagnosis, a knowledge of tumour cell type affects treatment; chemotherapy is frequently appropriate in small cell cancer but much less useful in non-small cell disease.

surgery, oncology, radiotherapy and palliative care. To achieve good results for patients a multidisciplinary team is therefore essential.

FIG. 13.65 Carcinoma of the bronchus: apical (Pancoast's) tumour The left first rib has been destroyed and there is a soft-tissue mass, due to nnetastatic nodes, in the supraclavicular fossa.

Bronchoscopy Fibreoptic bronchoscopy (p. 627) is the most important technique for confirming a diagnosis of bronchial carcinoma (as well as documenting the location of the tumour). A tumour is visible in about 70% of cases. Sputum cytology This simple, non-invasive technique can produce a positive result in up to 40% of patients. Sometimes the investigation confirms malignancy, but fails to determine the specific nature of the tumour. Similarly, cytology of pleural fluid will usually confirm a diagnosis of malignancy. Needle aspiration lung biopsy Percutaneous needle aspiration biopsy, with X-ray or CT screening to localize the tumour, is particularly effective for peripheral lesions not seen on bronchoscopy. A positive cytological diagnosis is achieved in 75-90% of patients with tumours. Surgical biopsy In a small number of patients surgical biopsy is necessary to establish a diagnosis. Mediastinoscopy and mediastinotomy allow biopsy of tumour within the mediastinum (see p. 630), and thoracoscopy or, occasionally, thoracotomy gives access for lung and pleural biopsy. Staging notation for lung cancer Following diagnosis, patients can be staged according to the extent of their disease (Table 6.6, p. 155). In general, only patients with stage I disease are suitable for surgery. The tumour, node, metastasis (TNM) system is useful when considering surgery for both small cell and non-small cell lung cancer. In small cell carcinoma a categorization into 'limited' disease (disease confined to one hemithorax) and extensive disease can be helpful when planning chemotherapy.

Management of lung cancer The management of lung cancer involves many disciplines: respiratory medicine, radiology, pathology, thoracic

13

Surgery The decision to resect a lung cancer should be made by a multidisciplinary lung cancer management team, and in these circumstances resection rates are often higher than when decisions are made by an individual clinician. The overall results of surgery are not good, with a 5-year survival of about 25%, but surgery represents almost the only possibility of cure in lung cancer. The critical question to be asked for each new patient with bronchial carcinoma is, 'Can this patient be cured by surgery?' The answer to the question of whether surgery is appropriate depends on the answer to several intermediate questions: 1. Is the patient's general health good? Coronary artery disease is common. Severe coexisting medical conditions may preclude surgery. 2. Is pulmonary function adequate to cope with major surgery and pulmonary resection? Most patients have airways obstruction. An FEVi of less than 1.5L usually precludes pneumonectomy. How much 'useful' lung is to be resected? 3. What is the site of the tumour: is surgery technically feasible? 4. What is the tumour histology? Small cell carcinoma is usually disseminated. Most favourable histology is well differentiated squamous cell carcinoma. 5. Is there evidence of local spread of tumour within the thorax? 6. Is there evidence of distant metastases? 7. Is the patient prepared to have surgery? In practice only 10-20% of patients are appropriate for surgical referral. Patients with pulmonary hypertension, hypoxia and hypercapnia should not be operated upon, and those with excessive sputum production often develop postoperative chest infections. Static lung function can be misleading and information on exercise tolerance is helpful. In most cases this information is as easily obtained by the physician walking up the stairs with the patient as by formal exercise testing. In addition to pulmonary function tests the contribution of the lung that will be resected can be obtained by perfusion and ventilation lung scans, and the postoperative FEV\ and VC can be predicted from such information. Recurrent laryngeal nerve invasion, malignant pleural effusion, superior vena caval obstruction, mediastinal lymph node spread and metastasis to the contralateral lung all preclude surgery. In considering surgery, assessment for mediastinal spread is critical. CT scanning can be helpful: a negative scan usually indicates that the mediastinum is clear, but a scan may demonstrate enlarged nodes that are subsequently found to be due to reactive hyperplasia

705

rather than tumour. These cases may require surgical assessment of the mediastinum. The relatively poor results of surgery in bronchial carcinoma reflect the widespread and early metastases so common in this disease. Prior to surgery, biochemical tests may reveal abnormal liver function, and liver secondaries can be confirmed by ultrasound or CT scan. Unfortunately, the available tests for detecting metastases, particularly those in the liver, frequently fail to pick up tumour deposits. A raised alkaline phosphatase may reflect bone secondaries, which can be confirmed by bone scan. CNS symptoms or signs indicate the need for a CT brain scan to exclude cerebral metastases. Results of surgical therapy The staging of disease, cell type, age of the patient and the operation performed all affect the results of surgery (Fig. 13.66). Although the overall survival rates of 25% at 5 years and 17% at 10 years are poor, some patient groups do relatively well. Young, fit patients who have a lobectomy for stage I squamous cell carcinoma have a 5-year survival of 55-60%. Chemotherapy Combination chemotherapy remains the mainstay of treatment of small cell lung cancer. The two most active regimens are etoposide and cisplatin, or cyclophosphamide, doxorubicin and vincristine. Treatment is given at 3-4-weekly intervals over a period of several months. A complete response to therapy with a return to normal of the chest X-ray and disappearance of tumour at bronchoscopy can be achieved in 25-50% of patients. Small cell tumours are rarely completely eradicated, however, and rapidly relapse. Chemotherapy

FIG. 13.66 Survival following surgical resection for non-small cell carcinoma of the bronchus

1 706

MCQ 13.23

increases the overall median survival from approximately 4 months without drug therapy, to 11 months with treatment (Fig. 6.12, p. 156). Approximately 10% of patients with limited disease will continue symptom- and diseasefree for 2 or 3 years. In general, patients with least tumour bulk and good general health have the best prognosis. The great reduction of tumour mass following treatment can produce dramatic symptomatic improvement, often following only two or three courses of drugs. Therapy can relieve breathlessness due to airways obstruction by tumour, bone pain due to metastatic disease, superior vena caval obstruction and haemoptysis. Chemotherapy is well tolerated by most patients, but avoiding the problems of nausea and infection due to marrow suppression requires careful attention to detail. A few more patients may achieve long-term survival if radiotherapy is combined with chemotherapy. However, the outlook for small cell carcinoma remains poor and is unlikely to be substantially improved until better chemotherapeutic agents become available. Combination platinum-based chemotherapy is increasingly being used as standard treatment for locally advanced and metastatic non-small cell lung cancer. Chemotherapy is used in conjunction with radiotherapy in stage III disease and increases survival compared with radiotherapy alone. Platinum-based chemotherapy may also be a useful palliative treatment. The increased use of active new drugs such as gemcitabine and the taxanes in conjunction with cisplatin is likely to result in further improvements in outcome in the next few years. Radiotherapy In bronchial carcinoma the principal purpose of radiotherapy is to produce symptomatic relief. Radiotherapy is particularly good at controlling bone pain, haemoptysis, superior vena caval obstruction and breathlessness due to narrowing of major airways. Following complete collapse of a lobe or lung, radiotherapy is less effective. Cerebral metastases often respond, albeit temporarily, to radiotherapy, with concomitant steroid therapy to control cerebral oedema. In patients with symptoms or signs of spinal cord compression urgent radiotherapy is indicated and can avert paraplegia. Back pain is often the first symptom in such patients. In carefully selected patients, radiotherapy can produce impressive results. For small, localized, well differentiated, particularly squamous cell, tumours in generally fit patients who are not suitable for surgery because of poor lung function, it is possible to achieve a 5-year survival of up to 17% following radical high-dose radiotherapy. Survival in localized small cell carcinoma is not greatly extended by radiotherapy alone, but tumour bulk can be dramatically reduced, with consequent excellent control of symptoms. In patients with small cell carcinoma who respond well to chemotherapy, prophylactic cranial irradiation can largely prevent the distressing development of symptomatic cerebral metastases. Mediastinal irradiation may also be used to prevent local relapse following induction of remission with chemotherapy.

A treatment regimen termed 'continuous hyperfractionated accelerated radiotherapy' (CHART), in which a small dose of radiotherapy is given three times a day over 12 days, may reduce tumour cell recovery between radiotherapy doses. Recent evidence suggests that this approach prolongs useful life. However, there are practical disadvantages, particularly as the treatment has to be given 7 days a week. Radiotherapy can cause a pneumonitis, the severity of which is related to the radiation dose. With radiation pneumonitis patients develop breathlessness and cough, often disproportionately severe relative to the chest X-ray abnormalities, usually about 6 weeks after therapy. Corticosteroids are helpful. Endobronchial therapy In carefully selected patients with obstruction of the major airways not amenable to chemotherapy or external irradiation, symptomatic relief can be achieved by a variety of endobronchial therapies, including endoscopic laser destruction of tumour, endobronchial radiotherapy and tracheobronchial stenting (see p. 628). Advanced bronchogenic carcinoma As with all malignant disease, patients with advanced bronchogenic carcinoma which is no longer responding to treatment require much medical nursing and social support during their last months of life. Support for the family is also important. The large number of patients dying of lung cancer represents an important challenge to hospitals, general practice, community terminal care teams and hospices. The future burden of lung cancer in countries that have only relatively recently adopted high levels of cigarette smoking will become a huge healthcare problem. Pain is a particularly troublesome symptom and radiotherapy may be required for its control. Pain may also require regular analgesics, particularly oral morphine. For intractable pain peripheral nerve blockade may be necessary. Morphine helps reduce the sensation of breathlessness. Poor appetite and lethargy can temporarily be alleviated by steroids, which may also improve the symptoms from cerebral secondaries for a few weeks or months. Weakness, immobility, poor food intake and analgesics all contribute to constipation. In the last weeks of life, pain, constipation, anxiety, breathlessness and weakness can greatly reduce the quality of life. Frequent assessment of the patient's needs is necessary. In this respect the family doctor's contribution is of great importance and the domiciliary terminal care team are able to provide a regular assessment in the patient's home. National guidelines generated by the Department of Health and the British Thoracic Society are now available in the UK and should help to deliver standard high quality management of lung cancer throughout the country. ALVEOLAR CARCINOMA

1

This is an unusual tumour with distinctive pathological and

clinical features. It constitutes 1-2% of primary pulmonary tumours. The sexes are equally affected and the tumour occurs in all races. Alveolar carcinoma is most common in patients above the age of 50, and is not related to smoking tobacco.

13

Pathology Alveolar carcinoma often arises within a damaged area of lung. Histologically the characteristic feature is malignant cells growing along the alveolar walls. On light microscopy it may be difficult to distinguish alveolar cell carcinoma from primary or secondary adenocarcinoma. The good prognosis for resection of early single nodule disease suggests that in most cases multinodular disease represents spread of tumour, rather than simultaneous multiple foci.

Clinical features The presenting symptoms of alveolar carcinoma are similar to other tumours although, because the tumour is not endobronchial, haemoptysis, bronchial obstruction and distal infection are less common. When the tumour mass is large or affecting both lungs, the patient becomes very breathless. In a few cases patients produce large volumes of clear, watery sputum (bronchorrhoea). The chest X-ray usually shows a single, irregular peripheral shadow sometimes resembling pneumonic consolidation, frequently with an air bronchogram within the tumour opacity. Cavitation is uncommon, but effusions occur. Multiple tumour opacities develop, initially within one, and then within both lungs (Fig. 13.67).

FIG. 13.67 Alveolar cell carcinoma of the lung The appearances of 'consolidation' are common with this carcinoma. Bilateral disease is also common.

707

Diagnosis

HAMARTOMA

Malignant cells can be detected in the sputum in some cases, but the distinction from adenocarcinoma is frequently not possible. Bronchoscopy and bronchial biopsy are usually negative, although transbronchial biopsy can provide a diagnosis. Diagnosis can be achieved by needle aspiration biopsy or, more frequently, by surgical biopsy at thoracoscopy or thoracotomy. Diffuse and multiple lesions are often initially thought to be infective, but failure to respond to antibiotic therapy raises suspicion of the true diagnosis.

Management Alveolar carcinoma is often relatively slow growing, and disseminated metastatic disease is less common than in other pulmonary tumours. Surgical resection is therefore the treatment of choice whenever possible, and good results can be achieved, with a 5-year survival of 50% for localized disease. More extensive disease has a poor prognosis and is difficult to treat, the tumour being resistant to chemotherapy and radiotherapy.

BRONCHIAL ADENOMA 1 The term 'adenoma' covers a variety of tumours, approximately 90% of which are carcinoid tumours arising from cells of the APUD series. The diagnosis is often made in young adults and the sexes are equally affected. These tumours are usually endobronchial, arising in larger airways, and causing airway narrowing, volume loss, collapse and distal infection. Common clinical features are cough, haemoptysis, breathlessness, unilateral wheeze and recurrent infections. Most tumours are of low-grade malignancy and symptoms may be present for a long time before a diagnosis is made. Wheezing and dyspnoea may lead to a mistaken diagnosis of asthma. Some tumours become malignant and occasionally (2% of cases) the patients develop the carcinoid syndrome (p. 796). It is important to note that tumour is not usually visible on the chest X-ray and the most common radiological abnormality is volume loss or lobar collapse. Diagnosis is frequently made at bronchoscopy, where 90% of these tumours are visible. The tumour is vascular and bleeds easily and profusely. Rigid bronchoscopy for biopsy is sensible if at fibreoptic bronchoscopy the appearances suggest an adenoma. The best treatment whenever possible is surgical resection, and the prognosis for these tumours is good unless there has been malignant change.

1 708

MCQ 13.23

Hamartomas are tumour-like malformations of disorganized tissue; in some the predominant tissue is cartilage, and in others smooth muscle. Unlike carcinoid tumours, hamartomas usually present over the age of 50. Males are three times more likely to be affected than females. Hamartomas are usually peripheral within the lung parenchyma and seldom within the bronchus. They grow slowly and are benign. Most are asymptomatic and are incidental findings on chest X-ray. The mass of the hamartoma is usually small, less than 4cm in diameter, rounded, well defined and sometimes calcified. Cavitation is rare. In most cases in the absence of typical calcification, and without the benefit of past chest X-rays, a malignant bronchial carcinoma will be part of the differential diagnosis and the correct treatment will be surgical excision. At surgery the hamartoma is felt to be hard, can be removed without resecting lung, and the prognosis is excellent.

THE SOLITARY PULMONARY NODULE

A common diagnostic and management problem arises when the chest X-ray, frequently a 'routine' investigation, shows a solitary pulmonary nodule (Table 13.53 and Fig. 13.68). The larger the lesion, the more likely it is to be malignant; lesions greater than 4 cm in diameter are usually malignant; those less than 8 mm are usually benign.

Management Resection of a single malignant nodule results in surgical cure in more than 60% of cases. Because 40% of all nodules are malignant, surgery should be considered in all patients who are fit for tumour resection, unless there is compelling evidence that the nodule is non-malignant. Nodules that have not grown in size for 2 years or which show calcification (especially central, diffuse or speckled) are likely to be benign. Solitary nodules in patients under the age of 35, especially non-smokers, are likely to be

TABLE 13.53 Causes of a solitary pulmonary nodule Malignant (40%) Bronchogenic carcinoma Alveolar carcinoma Bronchial adenoma Metastasis Benign (60%) Infectious granuloma Non-infective granuloma Benign tumour Other rare causes

25% 3% 2% 10% 50% 2% 2% 6%

13

FIG. 13.68 The 'solitary pulmonary nodule' A Small peripheral 'solitary pulmonary nodule' in the right lower lobe. B When there is doubt, calcification is best demonstrated by CT scan.

benign. The search for past X-rays is often the most rewarding investigation in these patients. A CT scan may demonstrate calcification more clearly or may show multiple pulmonary metastases. Positron emission tomography (PET) is useful to distinguish between malignant and benign pulmonary nodules, and a positive PET scan is strongly predictive of tumour. Fibreoptic bronchoscopy is seldom helpful because the lesion cannot be seen endoscopically. Needle aspiration biopsy is more useful, confirming malignancy in 75-90% of tumours, but a negative result is difficult to interpret. For some tumours and most benign lesions, the precise diagnosis is only revealed by surgical removal.

MEDIASTINAL TUMOURS 1 Mediastinal tumours are illustrated in Figure 13.69. Metastatic malignant disease is the most common cause of a mediastinal mass, most frequently at the hilum. Lym-

FIG. 13.69 Mediastinal mass lesions A The chest X-ray demonstrates a mediastinal mass: the preservation ot the silhouette of the descending aorta suggests'that the mass is placed anteriorly. B This is confirmed by the lateral chest X-ray: the retrosternal space, which is normally translucent, is opaque. The mass was a malignant thymoma. c Location of the more common mediastinal masses.

709

phoma and sarcoidosis are common, with anterior mediastinal adenopathy favouring the former. In addition to the lateral chest X-ray, CT scanning is crucial in elucidating the nature of mediastinal masses; contrast studies demonstrate vascular lesions, and cystic lesions (usually benign) can be identified. Thyroid isotope scans confirm a retrosternal goitre. Solid, non-vascular lesions usually require biopsy for diagnosis (for example, the differential diagnosis of carcinoma, lymphoma and sarcoidosis) by needle aspiration biopsy, mediastinoscopy, mediastinotomy, or thoracotomy. Neurogenic tumours are the most common cause of a posterior mediastinal mass. Symptoms arise from compression of local structures, a small proportion are malignant and treatment is surgical. Occasionally, a neurofibroma is associated with von Recklinghausen's disease. Dermoid cysts arise from ectodermal tissue, whereas teratomas are germ cell tumours that contain ectodermal, mesodermal and endodermal tissue. These teratodermoid tumours arise in the anterior mediastinum, commonly of young adults. Dermoids are usually cystic and benign, whereas teratomas are usually solid and malignant. Treatment is surgery to remove all or as much as possible of the tissue. In malignant cases, radiotherapy and chemotherapy may be necessary. Thymomas can be benign or malignant (about 25% of cases). Whereas many patients with thymoma have myasthenia gravis, only a minority with myasthenia have a thymic tumour. Treatment of all thymomas is surgical excision and, for malignant cases, postoperative radiotherapy.

FURTHER READING ON LUNG TUMOURS British Thoracic Society and Society of Cardiothoracic Surgeons of Great Britain and Ireland 2001 Guidelines on the selection of patients with lung cancer for surgery. Thorax 56:89-108. Doll R, Hill A B 1950 Smoking and carcinoma of the lung. British Med J 2:739-748. Matthay R A (ed) 1993 Lung cancer. Clin Chest Med 14 (1) Pieterman R M et al 2000 Preoperative staging of non-small cell lung cancer with positron-emission tomography. N Engl J Med 343 (4): 254-261. Report of a working group for the Standing Medical Advisory Committee 1994 Management of lung cancer; current clinical practices. London: HMSO.

PLEURAL DISEASE The visceral pleura, composed of a layer of mesothelial cells supported on connective tissue rich in lymphatics and blood vessels, covers the lung. The parietal pleura lines the

thoracic cavity, covers the diaphragm and meets the visceral pleura at the hilum. The visceral pleura has no pain fibres, whereas irritation of the parietal pleura induces the pain characteristic of pleurisy. Normally, the two pleural layers are adjacent to one another, separated by a thin film of fluid, continually formed at the parietal and drained at the visceral pleural surface. As a consequence of the elastic property of the lung, the pressure within the pleural space is negative, approximately -5cm H2O at resting endexpiration.

PLEURAL EFFUSION 1 Water and electrolytes from the pleural space are taken up by the visceral pleura and cleared by the pulmonary and systemic (bronchial) circulation, whereas the proteins of pleural fluid are mainly cleared via the lower mediastinal lymphatics. Excess fluid accumulates within the pleural space when visceral or parietal pleural capillary pressure is raised (e.g. left or right heart failure), when capillary permeability is increased (e.g. infection), and when lymphatic drainage is decreased (e.g. lymphatic invasion by tumour). The turnover of pleural water and electrolytes is rapid, even when a large effusion is present, hence the rapid response to diuretics of effusions due to cardiac failure. Protein turnover is less rapid, contributing to the slow clearance of effusions with high protein content. The accumulation of pleural fluid reduces the volume of the underlying lung, causing a restrictive ventilatory defect. Large effusions will lead to breathlessness, particularly if they develop rapidly, and, when unilateral, cause mediastinal shift.

Radiology The history, and particularly the clinical examination, may suggest pleural fluid, but the most helpful, simple investigation is a chest X-ray (Fig. 13.70). Radiology may suggest fluid but cannot determine its nature (e.g. a clear transudate, pus or blood). When there is doubt about pleural fluid a film taken in the lateral decubitus position will demonstrate pleural fluid adjacent to the lateral chest wall. Very small effusions are clearly seen on CT scanning. On occasion, pleural fluid becomes loculated, particularly when infected; ultrasound is a simple technique for confirming that fluid is present and locating its position. The collection of fluid in the pleural space beneath the inferior surface of the lung (subpulmonic effusion) can mimic an elevated hemidiaphragm. A lateral decubitus X-ray or ultrasound examination confirms the diagnosis of subpulmonic fluid.

Differential diagnosis 1

710

MCQ 13.24

2 Fig. 13.48

There are many causes of pleural effusion (Table The accumulated fluid may be a transudate, as a quence of alteration in the pressure relationships pleural space (as in cardiac failure), or it may

13.54). conseof the be an

TABLE 13.54 Causes of pleural effusion

13

Transudates (hydrothorax) Cardiac failure Hypoproteinaemia Constrictive pericarditis Myxoedema* Peritoneal dialysis and ascites (including Meigs' syndrome) Superior vena cava obstruction* Exudates Infections Malignancy Pulmonary infarction Rheumatoid arthritis Systemic lupus erythematosus Acute rheumatic fever Polyarteritis nodosa* Scleroderma* Dermatomyositis* Postmyocardial infarction syndrome Ovarian hyperstimulation Pancreatitis Subphrenic abscess Asbestosis* Familial Mediterranean fever Hepatic and pulmonary hydatid disease Hepatic amoebiasis Trauma Drugs Sarcoidosis* Yellow nail syndrome Radiotherapy* Chylous effusion Haemothorax latrogenic (e.g. misplaced subclavian infusion line) * Denotes that it is unusual for that disease.

FIG. 13.70 Pleural effusion A Small left effusion. The left hemidiaphragm cannot be visualized and the fluid appears to rise laterally and medially: on a lateral chest X-ray it rises anteriorly and posteriorly. (The patient had the yellow nail syndrome.) B Large effusion. Pleural fluid is homogeneous and without vascular markings or air bronchogram. The mediastinum is shifted to the right. Note the surgical clips in the left axilla. (The patient had metastatic pleural disease from breast carcinoma.) ©

exudate, as a result of increased capillary permeability (as in infection). Transudates have a protein content of less than 3 g/dL, are clear or light yellow, and often bilateral. Exudates have more than 3 g/dL of protein, are dark yellow or orange in

colour, often slightly cloudy, may clot on standing, and are usually unilateral. Trons udates Left heart failure, right heart failure, or a combination of both, are common causes of pleural effusion. Initially, the fluid may be unilateral, usually right-sided; when bilateral, the right effusion is usually larger than the left. In constrictive pericarditis the pleural effusions are less significant than the ascites. Any cause of hypoalbuminaemia may cause pleural effusion plus oedema and ascites; the nephrotic syndrome is often the cause. In Meigs' syndrome the pleural effusion is associated with ascites due to an ovarian tumour (usually a benign fibroma, occasionally

711

malignant disease), and resolution occurs if the tumour is removed. Effusions can occur with peritoneal dialysis and ascites. The mechanism of the hydrothorax is the direct passage of fluid through the diaphragm into the pleural space. Small communications have been observed at postmortem in patients with hydrothorax and cirrhosis.

712

Exudates Postinfective pleural effusions are most common with bacterial infections, particularly pneumococcal, and, if inadequately treated, may progress to an empyema. Even when not obviously purulent the fluid contains many polymorphonuclear leukocytes and is often associated with persistence of fever and leukocytosis. Mycoplasma is rarely responsible. Effusions following viral infections occasionally occur, particularly as part of the pleuropericardial involvement of Coxsackie B infections. Effusions have been described in glandular fever. Rarely, actinomycosis and other fungal infections can cause a pleural effusion. Tuberculosis, once a common cause of effusion, is now relatively unusual in the UK. An important aspect of diagnosis is pleural biopsy for histological examination and culture of the mycobacterium. Malignant pleural effusion may be due to primary pleural tumours, most commonly mesothelioma (see below) and rarely a benign pleural fibroma. However, the most common cause is metastatic pleural malignancy from carcinoma of the bronchus, as well as breast, ovary and other extrathoracic tumours. Effusions, usually bilateral, may complicate lymphangitis carcinomatosa (p. 704). Occasionally, lymphomas involve the pleura. Involvement of ribs and adjacent parietal pleura by myeloma may produce a pleural effusion. It is characteristic of malignant effusions that they develop rapidly and recur quickly after aspiration; they are also commonly, though not invariably, bloodstained. A pleural effusion is an important clinical feature of pulmonary infarction (p. 675). Pleural effusion is relatively common in some of the collagen vascular disorders (p. 690). In SLE and MCTD pleurisy and pericarditis are not uncommon, particularly in young and middle-aged women. Aspirated fluid contains lymphocytes, occasionally lupus erythematosus cells, and more often antinuclear factor and rheumatoid factor. In active rheumatoid arthritis 5% of patients have a pleural effusion (p. 1137). Pleural effusions are rather uncommon in polyarteritis nodosa, systemic sclerosis and dermatomyositis. Acute rheumatic fever is complicated by pleural effusions, usually small, in 10% of cases, most of which also have pericarditis. In the postmyo cardial infarction syndrome pericarditis and fever develop several days or weeks after acute infarction, and occasionally there is also a pleural effusion. Sarcoidosis is a rare cause of pleural effusions. Recurrent, self-limiting effusions can be a feature of familial Mediterranean fever (familial paroxysmal polyserositis). In the yellow nail syndrome effusions are associated with a yellow discoloration of nails, plus lymphoedema and occasionally

bronchiectasis and sinusitis. Lymphatic hypoplasia underlies the disorder, and this is thought to explain the effusions. Drugs can cause pleurisy with effusions: hydralazine (as part of a lupus syndrome), nitrofurantoin, sulphonamides, PAS, methotrexate, practolol and methysergide have all been incriminated. Radiotherapy to the lung can produce a radiation pneumonitis and small bilateral effusions. Following chest trauma and pulmonary contusion a pleural effusion occasionally collects. However, a haemothorax must first be excluded. Pathological processes below the diaphragm can cause pleural effusions, a common complication of acute pancreatitis, when it is usually painless, left-sided and haemorrhagic. The pleural fluid has a high amylase concentration, higher than in the blood. Occasionally similar effusions are seen in patients with chronic pancreatitis or pancreatic pseudocysts. Subphrenic abscess is frequently accompanied by a pleural effusion, most often on the right, with low lateral chest and shoulder pain as a consequence of diaphragm inflammation. On chest X-ray, in addition to the effusion, there may be elevation of the hemidiaphragm, below which there is sometimes a fluid level. If treatment is delayed, an empyema develops. Prior antibiotic therapy may produce a chronic clinical picture.

Investigation of pleural fluid Whenever there is doubt about the cause of an effusion, aspiration should be performed. Loculated or small effusions are best aspirated with ultrasound localization. Bloodstained effusions, rather than the frank blood of a haemothorax, are most common with malignancy, but also occur in pulmonary infarction and following chest trauma. Rare causes of a bloody effusion include cirrhosis, tuberculosis and leukaemia. Infected pleural fluid is turbid, and with advanced empyema becomes purulent. With a chylous effusion the fluid is a milky white colour. Analysis of pleural fluid protein concentration broadly separates transudates (less than 3g/dL) from exudates (more than 3g/dL), but in a large number of cases borderline, unhelpful values are obtained. Cytology is a key investigation, not simply for malignant cells. The presence of numerous polymorphonuclear leukocytes suggests bacterial infection; lymphocytes predominate in tuberculosis but are also common in any chronic effusion; eosinophils are numerous in the effusions of pulmonary eosinophilia, polyarteritis nodosa and Hodgkin's disease. Eosinophils can be the prominent cells following pulmonary infarction, and may also accumulate following bleeding into the pleural space from any cause. Pneumonias and rheumatoid arthritis are less usual causes for an eosinophilic effusion. Malignant cells are identified in up to half of patients with tumour involving the pleura. Bacteria may be demonstrated in postpneumonic effusions, particularly in those progressing to empyema. In tuberculosis acid-fast bacilli are only occasionally identified, but culture of pleural fluid for the mycobacteria is often useful.

In rheumatoid arthritis the pleural fluid contains a characteristically very low concentration of glucose. When pleural aspiration is not diagnostic pleural biopsy should be performed. Biopsy is particularly helpful in malignant disease of the pleura and in tuberculosis. Pleural biopsy An Abrams' biopsy needle is most commonly used and the investigation should not normally be undertaken when there is no fluid present, lest the biopsy needle puncture the visceral pleura. If the pleura is much thickened, biopsy with a cutting needle is safe even in the absence of fluid. Sometimes, diagnosis may require a large biopsy sample; for example, in mesothelioma a surgical biopsy is often indicated. Video-assisted thoracoscopy is very useful, and is increasingly used in the investigation of pleural disease (p. 630). Multiple biopsies increase the diagnostic yield, and one of the biopsy samples should be sent to the laboratory for culture if a diagnosis of tuberculosis is possible.

Management Infected pleural fluid should be promptly and completely drained; in many cases this is best done with an intercostal tube. Any pleural effusion causing breathlessness requires drainage. Malignant effusions may be large and recurrent, and therefore require pleurodesis. Re-expansion pulmonary oedema is rare, provided large volumes of fluid are not rapidly aspirated; passive drainage seldom causes problems. A variety of sclerosing agents can be introduced into the pleural space in an attempt to fuse the visceral and parietal layers, thereby avoiding the reaccumulation of fluid. Tetracycline, bleomycin and talc slurry can produce satisfactory results. If a large pleural effusion reaccumulates despite attempts at medical pleurodesis, surgical pleurodesis using talc powder or pleurectomy should be considered. This can be achieved by thoracoscopy or thoracotomy. Medical pleurodesis is successful in approximately 50% of patients, whereas surgery is effective in virtually all cases.

fluid. Fluid rapidly reaccumulates, and repeated drainage soon leads to severe wasting, hypoproteinaemia and lymphopenia. In up to 50% of patients with chylothorax there is spontaneous healing of the fistula, but if this does not occur the prognosis is poor unless the leak of chyle can be stopped. Following diagnosis a lymphangiogram is useful to determine the site of the fistula; if a leak is demonstrated the patient should be submitted to thoracotomy and ligation of the thoracic duct below the leak. If no leak can be demonstrated at lymphangiography, repeated aspiration is performed and the patient is supported by parenteral feeding, in the hope that the fistula will heal spontaneously. When healing does not take place, pleural drainage combined with pleurodesis is sometimes successful.

13

HAEMOTHORAX Haemothorax is usually the result of a penetrating injury or blunt trauma to the chest. Occasionally, a haemothorax develops without trauma. Such a spontaneous haemothorax, almost always left-sided, may occur with acute aortic dissection. Bleeding can occur with a pneumothorax, from the rupture of vessels within pleural adhesions, and in some cases the air may have been absorbed by the time of presentation. Bleeding disorders, heparin therapy, vascular pleural metastases and pleural endometriosis are rare causes of haemothorax. Haemothoraces, unless small and stable, should be drained by a wide-bore intercostal tube introduced in the midaxilla and connected to an underwater seal. If bleeding continues, the patient requires a thoracotomy. If blood is not removed from the pleural space, infection progressing to empyema can be a complication. In the long term an intense fibrous reaction to undrained blood can occasionally lead to a grossly thickened pleura and an encased lung, which then requires decortication if there is not to be substantial ventilatory impairment.

EMPYEMA CHYLOTHORAX Chylothorax most frequently follows rupture of the thoracic duct after trauma, particularly during thoracic surgery. Damage to the thoracic duct in the lower half of the mediastinum causes a right-sided chylothorax, whereas damage to the duct in the upper mediastinum produces a left chylothorax. When chylothorax is not due to trauma, malignancy involving the thoracic duct is the most common cause, particularly metastatic disease from carcinoma of the stomach, and lymphoma. Rarely chylothorax can complicate chylous ascites. Patients with chylothorax present with symptoms and signs of a large pleural effusion and the diagnosis is only apparent following aspiration of the milky white

Empyema refers to infected pleural fluid, whether a turbid effusion or frank pus. In many cases the diagnosis only becomes apparent on aspiration of the fluid. Most commonly, an acute empyema complicates a bacterial pneumonia, particularly pneumococcal, but other causes require consideration (Table 13.55). Staphylococcal pneumonia is an important cause in children. Anaerobic infection is sometimes difficult to diagnose, and it is likely that anaerobes are responsible for many 'sterile' empyemas. Tuberculous empyema is now unusual. It is not uncommon for empyema fluid to contain more than one organism. Empyema following thoracic surgery, and those associated with abdominal sepsis are frequently due to Gram-negative organisms. Amoebic liver abscess can rarely perforate the diaphragm and enter the

713

TABLE 13.55 Causes of acute empyema Thoracic surgery Penetrating chest injuries Sub-diaphragmatic infection Haematogenous spread of infection

Pneumonias Lung abscess Bronchiectasis Oesophageal perforation Pulmonary infarction

bronchus, thereby producing a bronchopleural fistula. Patients then cough large volumes of purulent sputum and on X-ray the empyema has fluid levels. Untreated, an empyema may discharge through the chest wall (empyema necessitatis). Empyema, particularly when due to staphylococcal infection, can be complicated by a pneumothorax (pyopneumothorax).

Management Antibiotic therapy will depend on the bacteriology of aspirated fluid, and regimens will often require metronidazole to cover anaerobic organisms. The pleural fluid must be completely drained as soon as possible. With early infection and freely moving fluid, needle aspiration can be successful. However, in most cases a large-bore intercostal drain should be inserted. The drainage of pleural fluid can be enhanced by instillation of intrapleural streptokinase (250 000IU daily for 3 days). If such treatment fails to produce rapid resolution of symptoms and complete drainage of the empyema, surgery is required. At surgery, the empyema is evacuated and the lung decorticated so that it may fully expand and thereby obliterate the pleural space.

Chronic empyema

FIG. 13.71 Empyema On the lateral chest X-ray the configuration of the posterior loculated pleural fluid is typical of empyema. The left hemidiaphragm is well seen, with the gastric bubble beneath it. The posterior curve of the right hemidiaphragm is obscured by the empyema.

pleural space, as can actinomycosis and hydatid disease. The empyema that occasionally complicates the mediastinitis of oesophageal rupture is usually left-sided. The clinical features of acute empyema include fever, night sweats and chest pain, with signs of a pleural effusion. The most common clinical setting is that of a pneumonia that has failed to respond to therapy. The white cell count is elevated (15000-20000) with a polymorphonuclear leukocytosis. Pleural fluid aspiration is essential and diagnostic; the fluid may be turbid, with numerous polymorphs or frank pus. Empyemas are usually posterior and lateral above the diaphragm. Initially, the pleural fluid moves freely in the pleural space and the chest X-ray demonstrates the appearances of a simple effusion, but the infected fluid rapidly becomes loculated and the chest X-ray then shows localized collections of fluid (Fig. 13.71). An acute empyema can breach the lung and enter a 1 714

MCQ 13.25

2

Fig. 13.8

Occasionally, the diagnosis of acute empyema is missed or the condition is inadequately treated and a stable chronic empyema develops. Infection is particularly likely to persist when there is an associated bronchogenic carcinoma or a bronchopleural fistula. Patients are usually not acutely ill. Fever is low-grade and there is likely to be clubbing, weight loss, anaemia, and the clinical signs of extensive pleural thickening. The chest X-ray shows dense pleural shadowing, perhaps fluid levels, and occasionally periosteal reaction in overlying ribs. A later radiological development is pleural calcification and a late clinical development is secondary amyloidosis. The treatment of chronic empyema is surgical, but in patients too frail for surgery drainage by a wide-bore drain, progressively shortened as the empyema cavity reduces, can produce satisfactory resolution.

PLEURITIC CHEST PAIN WITHOUT EFFUSION A variety of disorders can cause pleuritic chest pain, often with a pleural rub, but without a pleural effusion. Some of these conditions remain 'dry' but in most fluid may accumulate, and therefore many of the causes of a pleural effusion must be considered in the differential diagnosis of dry pleurisy. The most important causes of pleurisy without fluid are listed in Table 13.56. Pleuritic pain is a common feature of many radiologically obvious pneumonias, but can also occur with minor bacterial and viral pulmonary infections not radiologically

TABLE 13.56 Important causes of dry pleurisy Pneumonia (particularly pneumococcal pneumonia) Pulmonary infarction Bronchogenic carcinoma Lung abscess Bronchiectasis Tuberculosis Rheumatoid arthritis Systemic lupus erythematosus

Uraemia Coxsackie B virus (Bornholm's disease) Familial Mediterranean fever Trauma Asbestos pleural disease, including mesothelioma

apparent. Epidemic myalgia (Bornholm's disease), most commonly due to the Coxsackie B virus, is primarily an infection of intercostal muscle but occasionally the pleura is involved. The intercostal muscles are tender. There may be associated pericarditis, myocarditis or orchitis. Pain may relapse and remit several times before finally settling.

TABLE 13.57 Conditions predisposing to spontaneous pneumolhorax

13

Chronic bronchitis and emphysema Asthma Cystic fibrosis Tuberculosis Marfan's syndrome, Ehlers-Danlos syndrome Congenital cysts Honeycomb lung

pneumothorax). Most commonly, spontaneous pneumothorax develops in an otherwise fit person as a result of the rupture of a small apical pleural bleb, probably congenital in origin. Young men are most frequently affected and are often tall and of slim build.

Clinical features PNEUMOTHORAX 1 Pneumothorax is air in the pleural space, as a consequence of which there is partial or complete collapse of the lung. Owing to the recoil of the chest wall and the lung, the pressure within the pleural space is normally negative. When the pleural membrane (visceral or parietal) is breached, air is sucked into the pleural cavity and the lung collapses. When the defect in the pleura seals, the pneumothorax is closed and there is no movement of air in or out of the pleural cavity. When there is a persistent defect in the pleura the pneumothorax is open. When this defect is of the visceral pleura there is then a bronchopleural fistula, and air moves in and out of the pleural space during breathing. Occasionally the damaged visceral pleura acts as a valve, permitting air to enter the pleural space on inspiration but not to leave it on expiration, leading to a tension pneumothorax. A pneumothorax may be spontaneous, or follow chest trauma, mechanical ventilation, ruptured oesophagus, or artificial induction. Causes of traumatic pneumothorax include blunt trauma to the chest, commonly external cardiac massage, and penetrating chest injuries (e.g. stab wounds or needle aspiration biopsies that breach the visceral pleura from without) and transbronchial biopsy and positive-pressure ventilation that breach the visceral pleura from within. A spontaneous pneumothorax may be primary, in which case the patient is otherwise healthy, or may be secondary to some other disorder (Table 13.57). Of these, airways obstruction, particularly with associated bullous emphysema, is the most common predisposing cause. Rarely, in young women, pneumothorax may follow the passage of air from the abdominal cavity into the pleural space (usually the right) through fenestrations in the diaphragm. The air enters the abdominal cavity from the fallopian tubes at the time of menstruation (catamenial

Characteristically, there is a sudden onset of chest pain laterally, sometimes radiating to the shoulder. With substantial collapse of the lung there is associated breathlessness, and a dry, irritating cough is common. Sometimes the patient is aware of the partially collapsed lung flopping about within the thorax. On examination the most striking findings are of reduced breath sounds and hyperresonant percussion. A small left-sided pneumothorax may be associated with a clicking noise with each heartbeat, noted by the patient and on occasions loud enough to be heard by others. A large pneumothorax in a normal person causes breathlessness, and in the presence of pre-existing pulmonary disease a small one may cause severe respiratory distress. Tension pneumothorax is a medical emergency. The valve action of the pleural tear results in a progressive increase in the size and pressure of the pneumothorax. The underlying lung is totally collapsed. The mediastinum is shifted to the contralateral side, compromising the function of the opposite lung, and the high intrathoracic pressure elevates the jugular venous pressure, reduces venous return and causes tachycardia, low cardiac output, and eventually circulatory collapse and death. With the exception of a tension pneumothorax, when action may be required immediately, a chest X-ray is necessary in all patients in whom the diagnosis of pneumothorax is a possibility. © Small apical pneumothoraces may be difficult to see and an expiratory chest X-ray, which enhances the contrast between lung tissue and air in the pleural space, is helpful. For small pneumothoraces the edge of the visceral pleura is often easier to see when the chest X-ray is viewed on its side. The X-ray allows an assessment of the size and position of the pneumothorax (the air may be loculated if there are pleural adhesions), any abnormality in the underlying lung, and the presence of mediastinal or subcutaneous air. A loculated pneumothorax can be difficult to distinguish from a bulla, in

715

which case a CT scan is helpful. CT can also detect pneumothoraces too small to be visible on chest X-ray. A small volume of pleural fluid is commonly seen, but when there is a larger collection it should be sampled to exclude bleeding and, occasionally, infection (pyopneumothorax). Recurrence of pneumothorax is common; a young person with spontaneous pneumothorax has a 20-30% chance of a second one, usually on the same side; with each pneumothorax the probability of further episodes increases. The longer the time that elapses following a pneumothorax, the less chance there is of a recurrence.

Management

716

Any pneumothorax sufficient to cause or increase breathlessness requires drainage. Many patients, particularly young, otherwise fit males, present with a small or moderate pneumothorax, and whatever breathlessness was experienced acutely has settled. Such closed pneumothoraces do not require therapy and the patient can be monitored as an outpatient, provided he is instructed to seek medical advice if breathlessness returns. A 20-30% collapse of a lung normally takes several weeks to expand fully. Simple aspiration is an effective treatment for many pneumothoraces. If aspiration is not effective (because of a bronchopleural fistula) an intercostal drain is required. In a tension pneumothorax, if necessary, prior to the insertion of an intercostal drain, tension can be released by inserting an intravenous cannula into the pleural space in the second interspace in the midclavicular line. Unless a pneumothorax is loculated, the best site for a drain is in the midaxilla at the level of the sixth intercostal space. When the catheter is connected to the underwater sealed drainage bottle, any pleural air under positive pressure bubbles out. Careful and regular inspection of the level and movement of the water meniscus in the draining tube provides important information in the management of a pneumothorax (Fig. 13.72). In a closed pneumothorax, coughing or other manoeuvres that produce a positive intrathoracic pressure drain the air and the pneumothorax rapidly resolves. When there is a large defect in the visceral pleura, producing an open pneumothorax and a bronchopleural fistula, any manoeuvre producing positive intrathoracic pressure causes much bubbling, but because the pleural defect remains the pneumothorax does not resolve. The chest drain is kept in situ in the expectation that the pleural defect will close and the pneumothorax convert from an open to a closed type and then resolve. If resolution does not occur within a few days, it is common practice to apply suction to the pneumothorax. In practice, only a minority of open pneumothoraces - usually those in which the pleural leak is small - can be treated in this way. Frequently, the application of suction simply serves to hold the pleural defect open. When air no longer bubbles from the catheter even when the patient coughs, a chest X-ray should be taken to confirm that the lung has expanded fully. If this is the case,

FIG. 13.72 Pneumothorax [A] In a closed pneumothorax the meniscus in the drain rises and falls with breathing, and at resting end-expiration the small negative intrathoracic pressure supports a column of water. B In an open pneumothorax air bubbles into the drainage bottle, with each breath, during active expiration, c If pleural fluid blocks the drainage tube the bubbling of air ceases and there is little movement of the meniscus with breathing. This situation can therefore mimic that of a resolved pneumothorax.

the drain should be clamped for several hours, and if the lung remains fully inflated the drain can be removed. Where a drain has been inserted for a recurrent pneumothorax a medical pleurodesis should be undertaken prior to removal of the catheter, most commonly with tetracycline, bleomycin or talc slurry. If an open pneumothorax persists for a week despite a period of suction the patient should be referred for thoracoscopy or thoracotomy, closure of the bronchopleural fistula, and parietal pleural abrasion or pleurectomy.

MediasHnal emphysema (pneumomediastinum) Air can enter the mediastinum following penetrating neck injuries, and after oesophageal or bronchial rupture, but the most common cause is rupture of alveoli. Air enters the interstitial tissues and tracks to the mediastinum. High intrapulmonary pressures, for example due to vomiting, are frequently the cause. In asthma, gas trapping, alveolar distension and high intrapulmonary pressures during coughing make mediastinal emphysema a relatively common complication. Positive-pressure mechanical ventilation can produce mediastinal emphysema as well as pneumothorax. Mediastinal emphysema is usually asymptomatic but can cause chest pain, which may be worse on swallowing or breathing. There may also be a crunching sound coincident with each heartbeat. Radiology may show air in the upper mediastinum, around the heart, within the soft tissues of the neck, and in 30% of cases an associated pneumothorax. Treatment is of the predisposing cause, following which the mediastinal emphysema rapidly resolves. Resolution can be hastened by breathing a high oxygen concentration, during which the more soluble oxygen replaces the insoluble nitrogen.

13

TUMOURS OF THE PLEURA The pleura can be the primary site of benign or malignant tumours or, more commonly, the site of metastatic malignant disease.

Fibroma The pleura - usually the visceral surface - may be the site of a single benign fibrous tumour (localized fibrous mesothelioma). The tumour, which is well encapsulated and often pedunculated, is frequently first noted as a peripheral pleural mass on routine chest X-ray. A pleural effusion is uncommon. These tumours are not associated with asbestos exposure. Fibromas can become large and thereby cause breathlessness as well as chest pain. Patients are generally middle-aged or elderly, and clubbing, as well as hypertrophic pulmonary osteoarthropathy, is a frequent association. A definite diagnosis is seldom possible without thoracoscopy or thoracotomy and biopsy, and surgical resection is the correct management. If left untreated, these tumours may eventually become locally invasive.

Diffuse mesothelioma Diffuse mesothelioma, a highly malignant tumour, is related to asbestos exposure in most - perhaps all - cases. The patient's exposure to asbestos is, on average, 40 years prior to the malignancy becoming apparent. Most commonly, exposure to asbestos is in the shipbuilding, construction and demolition industries. Crocidolite (blue asbestos) is most commonly incriminated. In the UK 1300 patients each year are diagnosed as having malignant mesothelioma and the number is likely to rise to 3000. The tumour spreads over the pleural surface, causing great thickening of the pleura by malignant tissue, and the lung becomes encased. The tumour spreads into the mediastinum, through the diaphragm to involve the peritoneum, and invades the chest wall. An effusion is characteristic and is often bloodstained. Distant metastases occur, although relatively infrequently, and cause major clinical problems. The histology is pleomorphic and a definite diagnosis, is often difficult on a small biopsy sample. Malignant mesothelioma can be divided into three subtypes: epithelioid, sarcomatoid and biphasic. Differentiation between epithelioid and metastatic adenocarcinoma can be difficult. The use of adenocarcinoma and mesothelioma markers can facilitate immunohistochemical discrimination. Patients with malignant mesothelioma present at an average age of 60, most commonly with persistent dull chest pain. There is progressive breathlessness because of pleural thickening and the accumulation of fluid. Finger clubbing occurs but is relatively unusual. The chest X-ray commonly shows a large effusion, with the pleural tumour often only apparent following aspiration. CT scanning clearly demonstrates the extent of pleural disease, and

FIG. 13.73 Mesothelioma [A] The chest X-ray shows bilateral pleural thickening and fluid, B The CT scan shows extensive pleural thickening, plus fluid, and pleural calcification posteriorly on the right (arrow).

when pleural calcification is also present the appearances are highly suggestive of the diagnosis (Fig. 13.73). Although cytology of aspirated fluid may demonstrate malignant cells, and Abrams' or cutting needle pleural biopsy may also show malignancy, a large surgical biopsy is sometimes needed to confirm the diagnosis. However, surgical biopsy can be followed by infiltration of the thoracotomy scar by tumour. The diagnosis is commonly achieved by thoracoscopic biopsy. A very small number of patients with localized disease can be cured by surgery; for most, treatment is palliative. Overall, no therapy, whether surgery, cytotoxic drugs or radiotherapy, has been demonstrated to prolong life. Recurrent symptomatic pleural effusions can be managed by talc pleurodesis or thoracoscopic pleurectomy. Chest wall pain, but not breathless-

717

ness, can often be helped by palliative radiotherapy. Life expectancy is 1-2 years following presentation. In the UK patients with a history of asbestos exposure are eligible for industrial injuries benefit.

Metastatic malignant pleural disease Pleural metastases with an associated effusion are common, particularly with bronchial carcinoma. The effusion is usually painless, but causes breathlessness and may require pleurodesis (p. 713). Pleural involvement is also common following malignant disease of the breast and ovary. Less commonly, pleural involvement occurs in other carcinomas, lymphangitis carcinomatosa, multiple myeloma, malignant lymphoma (particularly Hodgkin's disease), Kaposi's sarcoma and Waldenstrom's macroglobulinaemia.

TABLE 13.58 Neuromuscular disorders which cause respiratory muscle weakness Neurogenic Motor neuron disease Surgical trauma of phrenic nerves Polyneuropathy Neuralgic amyotrophy Poliomyelitis Multiple sclerosis Traumatic tetraplegia Acute porphyria Charcot-Marie-Tooth disease Following radiotherapy

Myopathic Muscular dystrophy/atrophy Inflammatory myopathies Myotonic dystrophy Acid maltase deficiency Thyroid myopathy Metabolic disturbance (e.g. hypophosphataemia)

Neuromuscular junction Myasthenia gravis Lambert-Eaton syndrome

FURTHER READING ON PLEURAL DISEASE British Thoracic Society 2001 Statement on malignant mesothelioma in the United Kingdom. Thorax 56:250-265. Davies R J O, Traill Z C, Gleeson F V 1997 Randomised controlled trial of intrapleural steptokinase in community acquired pleural infection. Thorax 52:416-421. Miller A C, Harvey J E on behalf of the Standards of Care Committee of the British Thoracic Society 1993 Guidelines for the management of spontaneous pneumothorax. Br Med J 307:114-116. Peto J, Hodgson J T, Matthews F E, Jones J R 1995 Continuing increase in mesothelioma mortality in Britain. Lancet 345:535-539. Reid P T, Rudd R M 1993 Management of malignant pleural effusion. Thorax 48:779-780.

NEUROMUSCULAR AND SKELETAL DISORDERS NEUROMUSCULAR DISORDERS Ventilation depends on adequate central respiratory drive, and depression of central nervous system function can therefore cause ventilatory failure. In such patients the diagnosis is usually self-evident and the resulting hypercapnia easily documented. Central nervous system dysfunction may also cause sleep apnoea syndromes (p. 668). In addition, a wide range of neurological and muscular disorders can cause ventilatory impairment on the basis of respiratory muscle weakness (Table 13.58).

The respiratory muscles

718

In a normal person at rest inspiration is active, whereas expiration is a passive process; most of the respiratory muscles are therefore inspiratory in their action. The diaphragm is the principal muscle of inspiration and is responsible for 80% of ventilation at rest. Diaphragm

contraction causes descent of the diaphragmatic dome, anterior displacement of the abdominal wall, elevation of the lower ribs and expansion of the lower thorax. At resting lung volume (functional residual capacity) the external intercostal muscles are inspiratory and the internal intercostal muscles are expiratory. The abdominal muscles are the most powerful muscles of expiration. The accessory muscles of respiration are conventionally considered to be the sternomastoid, scalene and pectoral muscles, which act mainly to fix or elevate the upper ribcage. However, many other muscles contract vigorously during maximum respiratory efforts.

The respiratory muscles in neuromuscular disease Clinical assessment Isolated involvement of the respiratory muscles by neuromuscular disease is unusual, and most patients will have evidence of more widespread involvement, with, for example, weakness of other muscle groups, bulbar problems and impaired cough, as well as breathlessness. On examination there may be muscle wasting, weakness and fasciculation. Patients with bilateral severe diaphragm weakness or complete diaphragm paralysis present an unusual and striking clinical picture (Fig. 13.74). Severe weakness of the inspiratory muscles reduces the lung volume on the chest radiograph, but this can be an unreliable sign; unless the radiologist is already aware of the diagnosis, it is likely that it will be assumed that the patient has chosen not to take a full inspiration rather than that he or she is incapable of doing so. Diaphragm paralysis elevates the hemidiaphragm involved. Radiological screening, which should be performed with the patient supine, will demonstrate the upward movement of a paralysed hemidiaphragm during

FIG. 13.74 Paralysis or severe weakness of the diaphragm In patients with severe diaphragm weakness, when supine the contraction of the intercostal muscles lowers intrathoracic pressure, sucks the diaphragm into the chest and the anterior abdominal wall moves inwards. Patients with diaphragm paralysis therefore have severe orthopnoea.

FIG. 13.75 The effect of respiratory muscle weakness on lung volumes The principal effect of substantial weakness is to reduce vital capacity. (TLC = total lung capacity; TV = tidal volume; VC = vital capacity; RV = residual volume.)

a sharp sniff. Diaphragm movement can also be assessed by ultrasound. Lung function Respiratory muscle weakness reduces both inspiratory and expiratory capacity (Fig. 13.75). Total lung capacity is reduced, residual volume (RV) is increased, and both of these changes decrease vital capacity (VC); this simple measurement is therefore excellent for the detection of respiratory muscle weakness. VC is an ideal measurement in the follow-up of patients with weakness and in muscular dystrophy; for example, the VC correlates well with both the clinical staging of disease and the assessment of prognosis. In neuromuscular disorders that can cause rapid profound weakness (e.g. Guillain-Barre syndrome) large falls in VC warn that assisted ventilation may be required. The measurement of VC is particularly useful when assessing diaphragm weakness. In the upright posture VC is reduced, but when the patient lies flat there is a further marked fall in VC, characteristically by 50% with total paralysis of the diaphragm. Lung function in patients with respiratory muscle weakness shows no evidence of airways obstruction and the ratio of FEVj to FVC is normal. Although overall gas transfer may be reduced, when corrected for the reduction in lung volume, gas transfer coef-

ficient (KCO) is normal or raised. With the progression of neuromuscular disease patients gradually develop hypoxaemia, but hypercapnia is a late development associated with a poor prognosis. Hypercapnic ventilatory failure is unusual before the VC has fallen to below 50% of normal. With generalized respiratory muscle weakness, particularly when the diaphragm is severely involved, ventilatory failure is more pronounced when supine, and may be particularly severe during sleep. Sleep studies are therefore very useful in the assessment of such patients.

13

Respiratory pressures A normal supine VC excludes severe inspiratory muscle weakness, but a reduced VC is not specific for weakness and must be assessed in the context of possible coexisting lung disease. A more specific and sensitive index of global respiratory muscle strength is provided by measurement of maximum static mouth pressures. Pressures are measured by a transducer or gauge; expiratory mouth pressures are recorded during a maximum expiratory effort at TLC against an obstructed airway, and inspiratory mouth pressures are measured during a maximum inspiratory effort from RV. In patients with generalized neuromuscular diseases respiratory pressures are commonly reduced to 50% of normal, even though patients are often not particularly breathless. Hypercapnia is unusual unless inspiratory mouth pressures are reduced to below 30% of normal. Diaphragm weakness The phrenic nerves are commonly invaded by tumour, particularly carcinoma of the bronchus. Viral infections, including herpes zoster, can cause a phrenic neuropathy, as can the neurological disorders of motor neuron disease, poliomyelitis, neuralgic amyotrophy, Charcot-Marie-Tooth disease and many peripheral neuropathies. Trauma, particularly surgical, is important, and phrenic nerve dysfunction is quite common following cardiac surgery. In the past, phrenic crush and phrenic avulsion for the treatment of tuberculosis was a common cause of phrenic neuropathy. Unilateral phrenic nerve palsy causes a 20% reduction in ventilatory capacity and a reduction in maximum transdiaphragmatic pressures, but patients have few symptoms: perhaps transient breathlessness, in the absence of additional pathology. Bilateral phrenic nerve paresis, on the other hand, causes shortness of breath on exertion and severe orthopnoea; indeed, patients are quite unable to remain flat for more than a few seconds. Many patients sleep propped up or on their sides to assist ventilation. Diaphragm strength can be specifically and accurately measured by recording transdiaphragmatic pressure with oesophageal and gastric balloon catheters during a maximum sniff effort. A non-volitional assessment of phrenic nerve and diaphragm function is achieved by electrical or magnetic phrenic nerve stimulation in the neck and measurement

719

CASE STUDY 13.8 DIFFICULTY IN SLEEPING IN A PATIENT WITH MOTOR NEURON DISEASE A 55-year-old woman was referred for a neurological opinion with a history of progressive pseudobulbar palsy developing over a 6-month period. She had difficulty with speaking and swallowing. She had not noticed weakness in her limbs. On examination there was facial weakness and a brisk jaw jerk, a brisk gag reflex and reduced tongue movements. The movements of the tongue were slow and spastic, and the tongue was wasted with fasciculation. There were also widespread fasciculations in all muscle groups of the upper limbs. In the lower limbs occasional fasciculation was noted, but the power was normal. Neck flexion was weak. There were no sensory abnormalities. Reflexes were brisk in the upper limbs and normal in the lower limbs, and the plantar responses were flexor. An MRI scan of the brain was normal and neurophysiological studies demonstrated denervation in all four limbs and the facial muscles. A diagnosis of motor neuron disease was made.

Questions

1. From a respiratory point of view, wiiat tests are useful at this stage? 2. What ire the most common causes of death in MND?

Discussion Most patients with motor neuron disease die because of ventilatory failure caused by weakness of the respiratory muscles. In some patients aspiration and chest infections are important causes of morbidity and mortality. Abdominal muscle weakness reduces cough effectiveness. Inspiratory muscle

720

weakness reduces inspiratory capacity and expiratory muscle weakness reduces expiratory capacity, and therefore vital capacity reflects overall respiratory muscle weakness and is a predictor of survival. Some patients have difficulty with the vital capacity test and alternative measurements may be required. However, for many patients with motor neuron disease (and other neuromuscular disorders) the measurement and monitoring of vital capacity is an essential part of assessment. The vital capacity in this patient was 98% of predicted, indicating that at this stage there was no substantial weakness of the respiratory muscles. Over a period of 4 months the patient's speech progressively deteriorated and she became virtually unintelligible. She also had severe problems with pooling of saliva and with swallowing. A feeding gastrostomy (PEG) was inserted. Nine months after referral the vital capacity was reduced to 40% of predicted. Although her limbs remained reasonably strong and she was able to walk unaided, there was breathlessness after 500 yards. The patient was sleeping poorly, waking frequently at night, and in the morning felt tired. First thing in the morning there were headaches, which cleared over a period of approximately half an hour. The patient was uncomfortable with breathlessness when lying flat on her back, and preferred to sleep at 45°. Throughout the day she felt lethargic and tired, and frequently fell asleep. Questions

3. Why is the patient's sleep disturbed? 4. What is the significance of her ofthopnoea?

Patients with respiratory muscle weakness eventually hypoventilate during sleep. Initially this is episodic, occurring first in REM sleep. The hypoventilation causes hypoxia and hypercapnia and leads to arousal. Sleep is therefore fragmented and of poor quality. As a result of sleepdisordered breathing, patients are excessively tired and sleepy during the day. The hypercapnia causes headache on wakening, which persists until the CO2 is reduced as ventilation increases. As weakness progresses, ventilatory failure becomes established during the day as well as the night. The orthopnoea in this patient signifies diaphragm weakness. When supine, the weak diaphragm is unable to descend and displace the abdominal contents against gravitational forces. When upright it is much easier for the diaphragm to descend. The diaphragm is the most important muscle of inspiration, and therefore orthopnoea is an important symptom. In general, orthopnoea develops when diaphragm strength is reduced to approximately one-third of normal. Lung function tests and respiratory muscle strength tests showed FEVi 0.4 L (predicted 2.4), VC 0.7 L (2.8); the patient had difficulty in performing these tests and therefore more detailed studies of respiratory muscle strength were undertaken, involving the measurement of oesophageal, gastric and transdiaphragmatic pressures during inspiratory efforts and following stimulation of the phrenic nerves. In addition, the pressure generated at the nostril during a maximum sniff effort (sniff nasal inspiratory pressure) was measured. These measurements confirmed that the inspiratory muscles, including the diaphragm, were profoundly weak approximately 20% of normal strength. Expiratory muscle strength

13

CASE STUDY 13.8 CONTINUED was also reduced to about 20% of normal. Blood gas analysis showed a Pao2 7.0 kPa, Paco2 9.2 kPa, pH 7.38 and bicarbonate 40.4.

Question 5. What do these blood gases indicate? The patient has chronic ventilatory failure with a compensated respiratory acidosis. The chronicity and severity of the ventilatory failure make oxygen therapy (and sleep) hazardous. The patient was instructed in techniques to augment coughing by the physiotherapists. Treatment was initiated with non-invasive positivepressure ventilation delivered via a nasal mask (Case Fig. 13.8.1). The patient tolerated this therapy well, and over the next few days the Pao2, when off the ventilator, increased to 9.09 kPa and the CO2 fell to 7.0 kPa. The patient was discharged from

CASE FIG. 13.8.1 Patient receiving non-invasive ventilation, via a nasal mask, from a small bed-side ventilator. Nocturnal domiciliary non-invasive ventilation improves the quality of sleep and enhances quality of life in many patients with neuromuscular disease.

hospital and continued non-invasive ventilation at home at night. The quality of her sleep improved dramatically and she felt much better, with resolution of her tiredness by day and no further headaches. The Paco2 and bicarbonate progressively fell. The patient and her carers were

of diaphragm electromyogram, phrenic nerve conduction time, and twitch transdiaphragmatic pressures.

Management Respiratory muscle weakness due to neuromuscular disease may improve spontaneously, as in the GuillainBarre syndrome, or may respond to appropriate therapy, as in myasthenia gravis. Prior to improvement, some patients may require a period of mechanical ventilation. In most patients, however, deterioration is relentless, particularly in those with muscular dystrophy or motor neuron disease, and these patients die of respiratory infection and ventilatory failure. In some patients deterioration may be relatively slow or a plateau may be reached, with respiratory muscle weakness causing ventilatory failure and disabling breathlessness. For carefully selected patients assisted ventilation may be appropriate, particularly at night. Non-invasive positive-pressure ventilation (via a nasal mask) and, less commonly, negative-pressure ventilation (with tank ventilators, cuirasses and jackets) can both be used, usually in the patient's home, and long-

in no doubt that the quality of her life was improved by nocturnal ventilation. Gradually the patient developed limb weakness and increasing immobility, and eventually, 1 year after starting ventilatory support therapy, she died.

term survival and/or improved quality of life can make this treatment worthwhile. In patients with quadriplegia from high cervical cord lesions, the diaphragm and other respiratory muscles are paralysed and the patient is dependent on mechanical ventilation. If the phrenic nerves are intact, as shown by adequate diaphragm contraction with percutaneous phrenic nerve stimulation, diaphragm pacing via the phrenic nerves can sustain ventilation and mechanical ventilation is no longer required. A common disorder affecting diaphragm function is hiccup: usually this is transient, but can occasionally be a chronic and disabling problem. Many therapies have been tried, but the most effective is baclofen.

Metabolic disorders causing weakness Acidosis impairs the contractility of both skeletal and cardiac muscle. Hypercapnia, perhaps mediated through acidosis, adversely affects muscle function and it is probable that the interaction of hypercapnic acidosis and hypoventilation contributes to the rapid deterioration of patients with ventilatory failure and a low pH. Hypophos-

721

phataemia is a well-recognized cause of ventilatory failure due to respiratory muscle weakness and is an important factor on the intensive care unit. Reversible weakness of the respiratory muscles also occurs with thyroid disease, hypokalaemia and disorders of magnesium and calcium metabolism.

Weight loss and muscle weakness Patients with a wide variety of medical and surgical illnesses have profound weight loss and muscle atrophy, with consequent weakness. This loss of muscle bulk can involve the respiratory muscles, which therefore also become weak. Adequate nutrition (p. 748), particularly in hospitalized patients, can minimize wasting, as will the reversal of catabolic factors such as sepsis, fever, protein-losing states and steroid therapy. Immobilization leads to rapid atrophy of skeletal muscle, and it is possible that prolonged mechanical ventilation causes disuse atrophy of the respiratory muscles.

Respiratory muscle function in patients with airways obstruction

722

Chronic bronchitis, emphysema and asthma cause airways obstruction and hyperinflation, both of which require the patient to generate greater than normal pressures during tidal breathing. The capacity of the respiratory muscles to produce the required large negative intrapleural pressure is compromised by muscle shortening, decreased diaphragm curvature, increased velocity of shortening and abnormal ribcage geometry. Respiratory muscle contractility may also be impaired by reduced oxygen delivery to the muscle, as a consequence of hypoxaemia and cardiovascular disease, and also by the respiratory acidosis associated with acute hypercapnia. The increased work of breathing combined with the reduced capacity of the respiratory muscles to generate tensions limits ventilatory reserve, causes breathlessness, and may predispose to the development of respiratory muscle fatigue and hypercapnic ventilatory failure as, for example, in severe acute asthma. Patients with airflow limitation require a high respiratory centre output to maintain ventilation, and depression of neural output (for example by sedatives) can markedly impair breathing. The work of breathing is reduced and the function of the respiratory muscles improved by therapy that reduces airflow limitation and lung volume. In patients with chronic lung diseases, specific respiratory muscle training has been undertaken with a view to improving both the strength and endurance of the respiratory muscles. Although some studies have reported a modest benefit from training, the overall value of such programmes remains a matter of debate. Attempts have also been made to improve respiratory muscle contractility with drugs. It has been postulated that theophylline may have a very small inotropic effect, but it is doubtful

whether under most circumstances this is of any clinical importance.

Respiratory muscle fatigue All skeletal muscles fatigue when subjected to sufficiently high load, and the respiratory muscles are no exception. Careful studies of the stressed respiratory muscles of normal subjects have demonstrated fatigue in terms of electromyographic changes, slowing of relaxation rate, and reduced contractile response to nerve stimulation. However, these complex techniques have not easily been applied to patients with ventilatory failure, and the clinical importance of respiratory muscle fatigue remains to be established. Data from patients on ICU failing to wean from mechanical ventilation indicate that muscle fatigue may play a part. Some authorities, particularly in North America, advocate the 'resting' of the respiratory muscles by assisted ventilation to allow fatigue to resolve. More clinical studies on the detection of respiratory muscle fatigue and its clinical importance are necessary before widespread use of this approach could be advocated.

THORACIC DEFORMITY Adequate ventilation requires that the thoracic cage be of adequate volume, rigidity and mobility. Many of the disorders causing thoracic deformity are now rare (for example rickets), and the most common problems are flail chest (a surgical and anaesthetic emergency), ankylosing spondylitis, scoliosis, thoracoplasty and fibrothorax.

Ankylosing spondylitis Ankylosis of the ribs is characteristic of ankylosing spondylitis (p. 1148) and chest expansion is frequently reduced. Respiratory problems occur relatively late in the course of the disease and most patients are 50-70 years old before developing symptoms. In addition to the classic radiological changes affecting the spine (p. 1150), patients can also develop upper zone pulmonary fibrosis with cavitation and occasionally an aspergillus fungal ball (mycetoma). Despite the reduced expansion of the ribcage, ventilation is relatively little impaired because of good diaphragm function and, in the absence of lung involvement or coexisting cardiac disease, ventilatory failure does not occur. As a consequence of the importance of diaphragm function, patients tolerate badly upper abdominal surgery or any other factors impairing diaphragm movement.

Scoliosis Scoliosis refers to a lateral curve of the spine (Fig. 13.76) and in most cases, perhaps three-quarters, it has no known cause. A wide variety of disorders are responsible for the remaining cases (Table 13.59). Although a congenital

13

TABLE 13.59 Disorders causing chronic scoliosis Idiopathic Disorders of bone (e.g. osteogenesis imperfecta) Neurological diseases (e.g. cerebral palsy, poliomyelitis) Muscle disease (e.g. muscular dystrophy) Connective tissue disorders (e.g. Marfan's syndrome, Ehlers-Danlos syndrome) Thoracic surgery (e.g. pneumonectomy, thoracoplasty) Chronic infection (e.g. empyema)

vertebral abnormality produces scoliosis in childhood, most scolioses develop in adolescence during the rapid growth spurt. Congenital heart disease is common in scoliotic patients. Scoliosis makes the thoracic cage stiff. The deformity also reduces the volume of the lungs and adversely affects the mechanical advantage of the respiratory muscles. Lung function tests show a restrictive ventilatory defect, the severity of which is related to the angle of scoliosis, with high thoracic curves causing the greatest impairment. Patients with moderate or severe scoliosis are breathless, and with more severe deformity they eventually develop hypoxaemia, hypercapnia, pulmonary hypertension and cor pulmonale. Patients with scoliosis secondary to neurological disorders are frequently the most severely affected, particularly if diaphragm function is impaired. Ventilatory failure first becomes apparent during the night, particularly as a consequence of the hypoventilation associated with REM sleep. With the onset of ventilatory failure, the patients require assisted ventilation. Positivepressure ventilation, via a nasal mask, is very effective. Negative-pressure devices, such as a purpose-built cuirass, a Tunnicliffe jacket, pneumosuit or body wrap, can also be used. Such domiciliary ventilation is usually only needed at night and can produce remarkable long-term wellbeing, with control of respiratory failure by day, reversal of pulmonary hypertension and resolution of cor pulmonale. During acute episodes of deterioration more prolonged ventilatory support may be required. From a respiratory point of view the role of surgery to straighten a scoliosis is controversial. Most surgery is undertaken in adolescence, with the aim of avoiding deterioration of the curvature and ventilatory function that otherwise occurs at this time. Thoracoplasty, surgical collapse of the ribcage to compress the cavities of tuberculosis, was a relatively common surgical procedure in the treatment of pulmonary tuberculosis prior to modern drug management. It produces a similar ventilatory defect to scoliosis. Many patients with

FIG. 13.76 Chest X-ray showing severe kyphoscoliosis Eventually such patients commonly develop ventilatory failure.

thoracoplasty develop breathlessness and eventually ventilatory failure in later life. Some of these patients have been treated with nocturnal domiciliary ventilation, with excellent long-term results comparable to those for scoliosis. Such is the success of assisted ventilation programmes that all patients with severe or progressive chest wall deformity (and some with neuromuscular disease) require careful specialist long-term follow-up.

FURTHER READING ON NEUROMUSCULAR AND SKELETAL DISORDERS Guelaud C, Similowski T, Bizec J-L, Cabane J, Whitelaw W A, Derenne J-P 1995 Baclofen therapy for chronic hiccup. Eur Respir J 8:235-237. Hill N S 1995 Long term nasal ventilation. Editorial. Thorax 50:595-596. Lyall R A et al 2001 A prospective study of quality of life in ALS patients treated with non-invasive ventilation. Neurology 57: 153-156. Polkey M I, Green M, Moxham J 1995 Measurement of respiratory muscle strength. Thorax 50:1131-1135. Shneerson J M 1995 Disorders of the thoracic skeleton. In: Brewis R A L, Corrin B, Geddes D M, Gibson G J (eds) Respiratory medicine, 2nd edn. London: W B Saunders 1537-1546.

723

This page intentionally left blank

14 Critical Care

Medicine John Goldstone, Geoffrey Bellingan and Martin Smith

Introduction 725

Organ donation 747

Cardiovascular system 725

General care of the critically ill 749

Sepsis 733 Respiratory problems 736 Central nervous system problems 742

nial pressure or jugular venous bulb saturation, is undertaken as required. Infusions, urine output and total fluid balance can be measured over periods of minutes, hours or days. Routine laboratory investigations are often performed on the ward itself, enabling sequential analysis of arterial blood gases, electrolytes and other biochemical data. Many hospitals have developed high-dependency units (HDUs) that allow detailed monitoring, either as part of the ICU or run as a separate ward in close liaison with ICU, often with the same clinical team. Patients requiring close observation rather than organ support can be admitted to the HDU, and if they deteriorate they can then transfer to full ICU care. Central to this process of monitoring and adjustment of the patient's physiology are the intensive care nurses, who are highly trained and usually allocated one per patient on the ICU (1:2 on HDU, depending on the patient mix).

CARDIOVASCULAR SYSTEM

Sedation 749 Outcome 750

INTRODUCTION Critical care medicine has its origins in the treatment of failing organ systems; it began with respiratory support, using techniques developed in anaesthesia, and was accelerated in the 1950s by the need to ventilate large numbers of patients with paralytic polio. Patients are now admitted to the intensive care unit (ICU) not only to receive mechanical ventilation, but also for the management of a wide variety of severe illnesses. Thus the critical care team will care for medical, surgical, obstetric or paediatric patients, or indeed any patient who has a failing organ that needs treatment or simply close monitoring. The most common causes for admission to the ICU are listed in Table 14.1. The skills required by an 'intensivist' are thus extensive and the work demands the expertise of a team of medical staff, nursing staff and others, including physiotherapists, pharmacists and speech therapists. This team functions most effectively when led by an enthusiastic and committed critical care specialist. By necessity, the ICU concentrates facilities in one area, with an array of technology. Facilities are available for immediate resuscitation. Many physiological variables are monitored continuously with the aid of electronic monitors at each bedside, including oxygen saturation, respiratory rate, pulse rate, ECG, continuous intra-arterial blood pressue monitoring and central venous pressure (Fig. 14.1). More detailed monitoring, such as cardiac output, intracra-

Patients admitted to the ICU frequently require assessment, monitoring and therapy directed at the cardiovascular system (CVS). Cardiovascular failure may be due primarily to a cardiac event, e.g. myocardial infarction, or secondary to other organ failure. The function of the heart is to deliver oxygenated blood and nutrients to peripheral organs and tissues, and when perfusion is reduced, secondary effects such as renal failure ensue. Indeed, normal cerbral function, urine output and blood pH remain the most basic indices of a functioning circulation. The aim of management is to determine the nature of the cardiovascular problem, monitor the effect of treatment and make adjustments to therapy as appropriate.

ASSESSMENT Whether assessment is by clinical examination or complex monitoring, the approach to the circulation should involve an evaluation of filling status, cardiac output and peripheral perfusion. The optimization of the cardiovascular system may be urgent, to avoid organ damage. In uncomplicated cases that respond rapidly to treatment, invasive measurements may not be required, but in many patients the need for physiological information to guide treatment requires invasive techniques. Traditionally, disorders characterized by hypotension are termed 'shock' (hypovolaemic, cardiogenic or septic). The clinical picture of reduced peripheral perfusion (cold, vasoconstricted limbs), reduced vital organ blood flow (oliguria or anuria, mental confusion) and compensatory tachycardia due to sympathetic overactivity are common to cardiogenic and hypovolaemic shock. In sepsis, peripheral vasodilation causes warm limbs with bounding pulses, making the clinical

725

FIG. 14.1 A typical bedside in ICU The patient is surrounded by monitoring equipment but central to his care is the careful charting of observations and constant adjustment of treatment by the intensive care nurse.

assessment misleading; however, the blood pressure is low and can lead to end-organ dysfunction (Fig. 14.2). As peripheral perfusion fails oxygen delivery to the tissues is reduced, and as respiration continues an oxygen debt builds up. Anaerobic respiration leads to the production of lactic acid. If this process continues the chances of survival decrease as the oxygen debt and serum lactate rise (Fig. 14.3). In these complicated cases, cardiovascular performance is optimized by manipulating filling pressures (intravascular infusions), systemic and pulmonary vasoconstriction (nitrates, ACE inhibitors, calcium antagonists, oxygen, nitric oxide) and cardiac contractility (inotropes, chronotropes), all of which may require continual adjustment and intervention. In addition, cardiac function can be manipulated by reducing workload (mechanical ventilation and paralysis) and by improving coronary perfusion, by using an intra-aortic balloon pump or surgically (with angioplasty or vein grafting).

Filling pressures The force of contraction of cardiac muscle is determined by its length:tension relationship, and force generation falls if the muscle is excessively stretched (increased preload). Function is also impaired by excessive afterload,

1

726

MCQ 14.1

TABU 14.1 Common causes of admission to ICU, UCLH Diagnosis

Incidence (%)

Postoperative care Pneumonia Chronic obstructive lung disease Sepsis Myocardial infarction and heart failure ARDS

46 16 10 6.5 5 3

which prevents cardiac muscle shortening (see Ch. 12). For a given preload and afterload, contractility is also influenced by ionic and hormonal factors (e.g. acidosis, hyperkalaemia, hypocalcaemia). It is not possible to measure fibre length directly on the ICU, but it is possible to measure pressure within the heart during cardiac filling. During diastole, atrial pressure is related to fibre length, and the filling pressure rises as the heart muscle fibres stretch. Initially the response to a rise in filling pressure (increased preload) is an increased stroke volume, but excessive pressures cause a fall in output. Pressure in the right atrium can be simply measured through a central venous catheter and a manometer; the height of the column of water is measured above the level of the right atrium, taken as being the midaxillary line when supine. Electromanometers convert the pressure signal into an electrical output, and this can be visualized

14

FIG.

14.3

When perfusion fails, outcome is related to its severity. This is indicated by an oxygen deficit and the arterial lactate concentration.

TABLE 14.2 Requirements for accurate measurement of CVP • • • •

FIG. 14.2 Management of hypotension

on the bedside monitor. Requirements for accurate measurement of central venous pressure (CVP) are shown in Table 14.2. Central venous access can be achieved from many sites. The most common method is to cannulate the internal jugular vein. To reduce the incidence of trauma during the procedure, a guidewire is passed through a small-gauge needle and the central venous catheter is threaded over the wire. Most catheters are multilumen, allowing several infusions to be administered simultaneously and safely. The position of the line and the possibility of a pneumothorax should be checked with a chest X-ray after insertion. The CVP is influenced by circulating blood volume and vascular tone. In the hypovolaemic patient filling pressures

Freely falling meniscus when connected to the patient Meniscus has fallen to steady baseline Meniscus swings with respiration/ventilation Reference point and body position are standardized between measurements

may initially be maintained by an increase in smooth muscle tone within the venous circulation. Under these circumstances a normal CVP may be misleading. Further blood loss or vasodilation (e.g. administration of opiates) will lead to low filling pressures and hypotension. A single estimate of central venous pressure is unreliable in indicating blood volume status, as are many commonly measured indices of haemodynamics (blood pressure, heart rate, mean arterial pressure). Volume status is more adequately assessed by measuring the haemodynamic response to a dynamic change. A key manoeuvre on the ICU is thus the fluid challenge. During continuous CVP monitoring, the impact of a rapidly administered bolus of fluid (200-400 mL of colloid) is observed. In the hypovolaemic patient the response to fluid challenge is a transient rise in pressure, but this is not sustained and CVP falls rapidly back to the baseline value (Fig. 14.4, line A). As fluid replacement continues, further volume challenges result in more gradual rises and falls in filling pressure until the point is reached where an increase in filling pressure is sustained (Fig. 14.4, line C). At this stage fluid replacement is adequate. 1 Volume status can also be assessed by observing the effect of either a volume load induced by leg elevation or the effects of vasodilatation induced by a bolus of a vasodilator (e.g. nitrates). Stroke volume can be continuously assessed during both these manoeuvres from the velocity profile of blood flowing in the descending aorta measured with an oesophageal Doppler ultrasound probe. With these techniques it is possible to demonstrate whether or not further venous filling is appropriate in

727

FIG. 14.4 The effect of a fluid challenge on central venous pressure (CVP) Line A occurs when CVP is unaffected by a bolus of colloid, and represents hypovolaemia. In line B a transient rise in filling pressure occurs, but is not sustained. In line C , filling pressure rises after the fluid challenge and is sustained.

rapidly changing circumstances, e.g. when inotropic agents are being used. Volume loading may enable inotropic support to be reduced, especially if vasoconstriction is a prominent feature. There is now excellent evidence that maintaining appropriate volume loading in high-risk patients prevents further complications and improves outcome. In health, right and left atrial pressures change in parallel, allowing assessment of left ventricular function from right-sided filling pressures. This relationship may be altered in disease (Table 14.3), requiring the more direct assessment of left atrial pressures. Left heart filling pressure can be measured by the insertion of a pulmonary artery catheter (see below), which can also be used to determine cardiac output and therefore the construction of filling pressure/cardiac output curves.

Pulmonary artery catheters

728

Pulmonary artery (PA) catheters are introduced via a great vein and advanced into the right atrium. Passage of the catheter into the right ventricle and out through the pulmonary valve into the pulmonary artery is enhanced by an air-filled balloon at the tip of the catheter. The flow of blood into the right atrium and out into the pulmonary artery carries the catheter into smaller arteries until it becomes wedged in a narrow vessel. When the balloon is deflated the catheter recoils proximally, restoring perfusion to the distal lung segment. Running through the catheter is a lumen that emerges at the tip. The pressure at the catheter tip is displayed continuously, and as the catheter is advanced the characteristic pressures of the right atrium, right ventricle and pulmonary artery are observed (Fig. 14.5). The occluded

FIG. 14.5 Pulmonary artery (Swan-Ganz) catheter in position in a patient with severe pneumonia As the catheter is advanced through the great veins into the right atrium (RA), right ventricle (RV) and pulmonary artery (PA), the characteristic pressure waves are displayed. The catheter is sited in the PA, and when measurement of pulmonary artery wedge pressure (PAWP) is required, the balloon at the tip of the catheter is inflated.

TABLE 14.3 Common circumstances where left- and right-sided filling pressures do not agree • Valvular heart disease (e.g. mitral stenosis) • Myocardial infarction • Pulmonary hypertension (e.g. secondary to lung disease, pulmonary embolus)

pressure is termed the pulmonary artery wedge pressure (PAWP) and is related to the filling pressure of the left side of the heart. Pulmonary artery catheters commonly have several lumina, so that cardiac output can be measured. Specialized fibreoptic catheters measure oxygen saturation at the tip, giving a continuous output of mixed venous oxygen saturation (SVo2). There is currently great debate as to the role of PA catheters, arising from a study demonstrating that certain groups of patients in whom the catheters were inserted had a worse outcome than those who did not have PA catheters. Despite ongoing debate about this report it is clear that great attention needs to be focused on who inserts the catheter, the clinical indications, the quality of the measurements made and, most importantly, the therapeutic responses made to the information derived.

Cardiac output Cardiac output can be determined in patients on the ICU by measuring the rate of disappearance of a dye introduced into the right atrium (dye dilution technique). The 'dye' conventionally used is cold saline and the fall and recovery in temperature is sensed at the tip of the catheter (in the pulmonary artery) by a miniature temperature

14

FIG. 14.6 Filling pressure versus cardiac output

FIG. 14.7 Poor contractility

Patients are normally in the middle of the filling curve. In the hypovolaemic state, the response to fluids is to move the filling pressure from point 1 to point 2, increasing the cardiac output.

Relationship between cardiac index (Cl) and filling pressure (PAWP) for the normal (solid line) and poorly contracting (dashed line) heart. Compared with normal the PAWP/CI curve (dashed line) is almost flat over a wide range of filling pressures.

gauge. In patients with a low cardiac output the rate of disappearance of the 'dye' is slow, and the resultant curve of dye concentration against time is low and flat. If the cardiac output is high, dye is quickly washed away from the heart and the curve is peaked and narrow. With sophisticated computer technology employing fuzzy logic, cardiac output can be measured continuously from right heart catheters, which slightly heat blood passing over the catheter and detect the temperature of the blood at the tip in the pulmonary artery. Cardiac output can also be determined using the oesophageal Doppler, by a lithium dilution technique, and clinically with reasonable accuracy by determining peripheral perfusion (warm to the fingers and toes, to the hands and feet or to the knees and arms, allied to an assessment of the strength of the arterial pulse). Disorders of the circulation can be considered by constructing a graph of filling pressure versus cardiac output (Fig. 14.6). To allow for differing sizes of patients, cardiac output is normalized by dividing by the body surface area, producing a cardiac index (CI) which is expressed in litres/minute/m2. The normal heart responds to a raised filling pressure (increased preload) by a rapid increase in CI. At the upper limits, further stretching or dilatation of the heart results in a reduction in cardiac output. When cardiovascular function is abnormal, the position of the curve is shifted (Fig. 14.7). The graph can be used to define the problem, monitor the effect of treatment and predict outcome. The filling pressure of the heart can be reduced either by a reduction in circulating blood volume due to haemorrhage or severe dehydration, or by loss of venous return

if the circulation is suddenly vasodilated, as would occur with sepsis. Baroreceptors detect the fall in filling pressure, stimulating the release of catecholamines from the adrenal gland and sympathetic nervous system. Catecholamines act on the myocardium to increase the heart rate and force of contraction (positive inotropic effect), and on the peripheral vessels to maintain venous return by venoconstriction and blood pressure by vasoconstriction (increased systemic vascular resistance). Other neurohormonal reflexes are also stimulated. Fluid loss from the kidney falls initially because of a reduction in glomerular filtration rate (GFR). Aldosterone is released, and salt and water reabsorbed, leading to the production of small urine volumes (<750mL/day) with low urinary sodium concentrations (<20mmol/L).

Peripheral perfusion When peripheral perfusion is poor, blood flow to skin, subcutaneous tissue and muscle is reduced, leading to cold clammy extremities. Blood flow is also diverted from mesenteric organs, such as the kidney, leading to oliguria. This response is mediated by baroreceptors and catecholamine production is stimulated. Peripheral perfusion may also be compromised during sepsis-induced vasodilatation, despite an often elevated cardiac output. In this circumstance perfusion is limited not by a failure of stroke volume or cardiac output, but by hypotension and preferential blood flow through arterial venous shunts in peripheral organs; the exact mechanisms leading to organ failure in sepsis are, however, not clear.

729

A metabolic acidosis is the fundamentally important observation common to all causes of poor peripheral perfusion. Sequential measurement of acid-base status is crucial; continuing acidosis or a worsening base deficit indicates continuing haemodynamic problems. Such sequential measurements are facilitated on the ICU by the use of an intra-arterial catheter, commonly in the radial or dorsalis pedis artery, from which arterial blood gas samples can be taken quickly and simply. With effective treatment and optimization of the circulation metabolic acid is cleared from the circulation, although frequently this process does not occur as readily as its accumulation, leading to a delay in the restoration of a normal pH. Several attempts have been made to measure tissue hypoperfusion directly, most successfully for the gut and brain. Measurements of gut mucosal pH can be made with balloon tonometers passed nasogastrically, and intracellular pH can be calculated to assess the perfusion defect at the tissue level. Although this and other methods of directly measuring tissue oxygen consumption are helpful, they have not yet replaced the more general measurement of acidosis in the arterial circulation. During anaerobic metabolism glucose is metabolized to pyruvate, but this substrate cannot enter the oxidative pathways and is converted to lactate. Lactate concentration in the circulation can be measured directly. Raised lactate levels indicate a decompensating circulation. Serial measurements of lactate greatly enhance the diagnosis of peripheral hypoperfusion. Lactate measurement is especially useful to exclude other causes of arterial acidosis, such as renal failure. As cardiac output declines, more oxygen is removed from arterial blood in order to maintain tissue consumption. In health, 25% of the available oxygen that is delivered to the tissues is removed, and the venous blood is 75% saturated when it returns to the heart. In hypoperfusion more oxygen is extracted, and the blood returning to the heart therefore has a lower saturation. Global venous oxygen saturation can be measured from samples of blood drawn from the pulmonary artery. Using specialized pulmonary artery catheters, mixed venous saturation (Svo2) can be measured continuously in the pulmonary artery. When cardiac output is low, oxygen extraction reaches a maximum and the mixed venous saturation falls. An Svo2 of <50% indicates a very poor prognosis. Cold peripheries can be assessed by measuring the temperature gradient between the core and the periphery (great toe) with two simple electrical thermometers. As perfusion falls the peripheries cool and the temperature gradient increases. A gradient greater than 4°C indicates hypoperfusion, and this measurement can be used to guide treatment, especially in the hypovolaemic patient.

1

730

MCQ 14.2, 14.3

SUMMARY 1 Non-circulatory causes of metabolic acidosis • • • •

Ketoacidosis Renal failure Drugs, e.g. salicylates Poisons, e.g. ethylene glycol

Fluid replacement Restoration of blood volume should be rapid to avoid organ damage. The speed of transfusion is often inadequate, especially in the elderly, where there is a fear of precipitating heart failure. As a guide to the adequacy of fluid replacement, a fluid challenge or other dynamic test of blood volume adequacy should be employed (Fig. 14.4). The ICU allows the rapid restoration of circulation because patients can be closely monitored. Choice of replacement fluid (Table 14.4) The choice of replacement fluid depends on the cause of the decreased filling pressure. In haemorrhage, replacement with blood is essential, although blood substitutes may be used until blood is available. Severe dehydration requires predominantly salt and water replacement. Reduction in filling pressure due to vasodilatation can be corrected with a plasma substitute. If the circulation is restored quickly the metabolic effects of an inadequate circulation will reverse spontaneously, but acidosis, hyperkalaemia and hyperglycaemia require careful management in patients in whom hypovolaemia has been prolonged.

Fluid overload Critically ill patients often retain fluid because of renal dysfunction, excess antidiuretic hormone (ADH) secretion or drug therapy (e.g. steroids). Volume overload may occur as a result of injudicious intravenous fluid therapy in patients who cannot produce a compensatory diuresis. The filling pressure is raised, stretching cardiac muscle to the point where contractility declines, at the plateau of the PAWP/CI graph. The clinical signs can be similar to those of the poorly contracting ventricle (see below), both right- and left-sided filling pressures being raised. Treatment is to reduce filling pressures rather than to infuse inotropes. Diuretic therapy reduces the volume of the circulation and promptly relieves oedema. If filling pressures are critically high, vasodilator therapy can reduce them acutely. Diuretic therapy will only be successful if cardiac output is sufficient to perfuse the kidneys and renal function is not impaired. If renal function is inadequate, volume reduction can be achieved by haemofiltration or venovenous dialysis (p. 1059).

14

TABLE 14,4 Replacement fluids

Solution

Stability

Half-life in circulation

Crystalloid

Good

10-15 min

Only a quarter to a third will remain intravascular

Haemaccel

Good

2-4 hours

Hespan

Good

6-12 hours

Fresh frozen plasma

Once thawed, clotting factor activity diminishes after 2 hours. Good

Several hours

Initially intravascular. As metabolism and excretion occurs, oncotic pressure declines and the original gain in intravascular volume diminishes With increased permeability, Hespan redistributes and can remain in tissues. As a result, colloid and tissue oncotic pressure can rise substantially. Intravascular

Albumin

1 hour

Redistribution

Mainly intravascular but low molecular weight results in tendency to dissipate within tissues and this is enhanced if permeability is increased

Poor contractility The relationship between filling pressure and CI when left ventricular function is depressed is shown in Figure 14.7. As filling pressure rises, the response of the ventricle is reduced and the CI is low. This typically occurs after myocardial infarction (cardiogenic shock). The crucial point is that rises in the filling pressure do not improve the CI. Pulmonary oedema occurs when the filling pressure is raised. The pressure at which fluid extravasates into the lung is determined by the balance of forces driving the fluid out of the circulation (hydrostatic pressure) and the osmotic pressure of the plasma, acting to retain fluid within the vasculature. As the filling pressures rise, this point is exceeded and oedema fluid develops (p. 699, Fig. 13.57). Oedema develops at low pressure if plasma proteins are reduced or capillary permeability is increased. Patients with pump failure have a low-volume pulse, raised jugular venous pressure, a third heart sound, and sometimes signs of pulmonary oedema. As a result of the low CI, urine output may be reduced. In such patients monitoring of the circulation is necessarily invasive, and includes continuous arterial pressure monitoring through a cannula inserted into a peripheral artery, cardiac output monitoring (PA catheter or oesophageal Doppler) and measurement of urinary output via a urinary catheter as an index of renal perfusion.

Management In pump failure the aim of management is to increase the

Colloid osmotic pressure

Clotting factors

Zero

None

None

28mmHg

None

25mmHg

High in most clotting factors

25mmHg

None

Comment Very good with long shelf-life, mainly used in USA for volume expansion Often used for initial volume replacement Stable in solution with a long shelf life With large transfusion volume a coagulopathy may occur. This colloid has advantages when capillary leak due to inflammation occurs Needs to be stored as frozen and therefore requires thawing

Not recommended for use on the ITU as its molecular weight is low and in inflammation it quickly leaks from the circulation

output of the heart for a given filling pressure, thereby altering the slope of the PAWP/CI curve. Positive inotropic agents The naturally occurring inotropic agents are adrenaline (epinephrine), secreted by the adrenal medulla, and noradrenaline (norepinephrine), produced by the sympathetic nervous system. Other inotropic drugs include the adrenaline precursor dopamine and the synthetic dobutamine. These drugs increase the force of cardiac contraction by stimulation of (3 receptors. A further category of drugs, the phosphodiesterase inhibitors (enoximone and milrinone), produce an inotropic response by increasing intracellular cyclic AMP. These drugs are also vasodilators, leading to a reduction in peripheral vascular resistance. These newer agents are more expensive and, unlike the catecholamines, their half-life is measured in hours, making it difficult to finely adjust their effect. Positive inotropes are detailed in Table 14.5. Preload reduction with vasodilators is most effective when the filling pressure/cardiac output curve is flat, and changes in filling pressure produce a reversal of pulmonary oedema without large decreases in the CI. Careful monitoring of the circulation is essential, as slight alterations in vasodilating dose may produce large reductions in cardiac output. All vasoactive drugs should be given by volumetric pumps, and the shorter-acting drugs are an advantage. Alterations in peripheral vascular resistance can be particularly effective if combined with inotropic support, improving forward flow while maintaining myocardial perfusion pressure. 1

731

TABLE 14.5 Positive inotropes (dosages shown are an initial guide) Agent

Receptors

Dose range ( /kg/min)

Actions

Comments

Adrenaline (epinephrine)

a and

1-10

Noradrenaline (norepdinephrine) Dopamine

a

0.1-4.0

Low dose: HR ++, contractility ++ Increasing dose: vasoconstriction, SVR++ Vasoconstriction ++

Dopaminergic, , -nergic

Isoprenaline

3 (low dose) 5-10 (intermediate) 10-40 (high dose) 1-10

Low dose: renal and mesenteric vasodilation Intermediate dose: HR and contractility raised High dose: vasoconstriction HR++, contractility ++

Useful alternative to dobutamine, but vasoconstriction may reduce peripheral perfusion Increased blood pressure at expense of perfusion, not an inotrope Used to promote renal perfusion. Often used in conjunction with dobutamine

Dobutamine

1-10

HR and contractility increased

1-10

Inodilator

Dopexamine

DA, P, NA reuptake inhibitor

SUMMARY 2 Factors that reduce 02 delivery to the myocardium • • • • •

732

Diastolic pressure <60mmHg Overdistension of ventricles Tachycardia Vasoconstriction Hypovolaemia

Negative inotropes The force of contraction of the myocardium is depressed by acidosis, hypokalaemia, hypoxaemia and hypercapnia. These negative inotropes, particularly acidosis, offset the effectiveness of treatment. The dose response of catecholamines is adversely affected by acidosis. Correction of these disturbances may depend on restoration of peripheral circulation, although in the short term reversal of acidosis by hyperventilation (in patients receiving mechanical ventilation) and bicarbonate may be necessary. Another group of negative inotropes used on the ICU are sedative drugs, commonly continuously infused in mechanically ventilated patients. It is important that these drugs are infused at the minimum dose necessary to keep the patient comfortable. Oversedation has many serious consequences, particularly reduction of cardiac performance and airway protection. The major determinant of oxygen delivery to heart muscle is coronary blood flow which, if reduced, impairs contractility. Myocardial blood flow, especially to the subendocardial layer, is reduced by an increase in ventricular wall tension. This is the case in the dilated heart and in one contracting against a raised peripheral vascular resistance. Flow to all layers of the myocardium is crucially dependent on diastolic pressure and duration, and both are reduced by tachycardia. Improvement of coronary

Improves contractility but increases myocardial 02 consumption Preferred initially as vasodilatation enhances cardiac output Action limited by tachycardia; possible preferential splanchnic vasodilator

blood flow can be effected by a reduction in afterload (reducing ventricular wall tension), if it is achieved without markedly decreasing diastolic pressure or causing a tachycardia. Mechanical support Myocardial function may remain depressed despite conventional therapy. Mechanical devices have been developed to enhance heart action by decreasing left ventricular work and improving coronary blood flow. One such is the intra-aortic balloon pump (LABP). The pump is connected to a sausage-shaped balloon that is introduced percutaneously into the femoral artery and placed in the descending aorta. The gas-filled balloon is inflated during diastole, and blood flow in the aorta away from the heart is reversed, increasing coronary blood flow. The balloon collapses at the start of systole, decreasing peripheral vascular resistance. The IABP therefore reduces afterload but maintains coronary perfusion. It has had considerable success in the management of pump failure immediately before or after cardiac surgery, but much less success in patients with a low CI, for example after myocardial infarction. New techniques, including artificial hearts, may in the future allow time to prepare selected patients for cardiac transplantation, although these are only available in a few centres. With severe pulmonary oedema the work of breathing is greatly increased. Overall oxygen consumption (and myocardial work) is reduced by mechanical ventilation. Although mechanical ventilation may depress the CI when patients are hypovolaemic, positive-pressure breaths may enhance cardiac function in the failing heart. Improvements in cardiac function may also occur by controlling pulmonary oedema, reducing the work of breathing and allowing for a stable period during which other therapies can be pursued.

SEPSIS Sepsis is now the 13th most common cause of death in the USA but, with the associated sepsis-induced multiorgan failure, is the most common cause of death on the ICU. It is thus a great challenge in modern critical care medicine. As our understanding of the clinical syndrome has been illuminated by basic science research, the need for stricter definitions has led to a series of terms: • Systemic inflammatory response syndrome, SIRS. This is a clinical condition with at least two of the following: temperature >38°C or <36°C; heart rate >90 beats/min; respiratory rate >20 breaths/min or a low Paco2; WCC >12 x 109 or <4 x 109, or >10% immature forms. Note that in 50% of patients SIRS is initiated from a noninfective cause, such as pancreatitis, haemorrhage or burns. • Sepsis. SIRS and a positive blood culture. • Severe sepsis (this has had many other names, such as multiple organ dysfunction syndrome, MODS). This is sepsis associated with organ dysfunction, hypoperfusion abnormalities (e.g. raised creatinine, low platelet count, DIC, mental confusion, hypoxaemia and cyanosis) or hypotension. • Septic shock. This is sepsis-induced hypotension, despite fluid resuscitation.

Clinical presentation The clinical appearance of sepsis may be a progression of the entities described above, with the illness beginning with SIRS. In aggressive cases the illness may present at a later stage. There is increasing recognition of the importance of Gram-positive infections in the aetiology of sepis in the ICU and these, together with mixed infections and fungal infections (which are also increasing in incidence), account for more cases of sepsis than the traditionally recognized Gram-negative infections. Many cases of sepsis occur without positive blood cultures, and other severe illnesses may initiate the inflammatory response. These may all result in clinically indistinguishable conditions.

Tissue hypoperfusion Within the terms used to describe sepsis, tissue hypoperfusion is an important clinical stage. It may be manifested by oliguria or by a metabolic acidosis, and serum lactate levels may be elevated (>2.0mmol/L). In contrast to other forms of hypoperfusion, tissue oxygen uptake is limited and oxygen extraction decreases. As a result, more oxygen is present in blood returning to the heart and mixed venous O2 levels are elevated. The relationship between oxygen supply to the tissue and uptake changes. Oxygen uptake may be linearly related to supply, and no 'plateau' exists as supply is increased. This has been interpreted as an oxygen handling problem at the level of the cell, although there

RECENT ADVANCES IN THERAPEUTIC INITIATIVES IN SEPSIS

14

The manipulation of individual mediators released during sepsis has been an attractive concept, especially as manufacture of monoclonal antibodies is possible. Unfortunately, initial enthusiasm has been tempered by lack of clinical success. It is possible that NO acts as a final pathway during sepsis, especially inducing hypotension. Although L-arginine analogues have been used clinically, no large studies are available describing a beneficial effect of such novel drugs on outcome. Plasmapheresis is now being more widely applied in adult medicine, and in fulminant infection (e.g. meningitis) this may be of value, especially acutely. Surface cooling during fever often results in vasoconstriction and a decrease in drug therapy, as well as decreasing peripheral oxygen consumption. Despite negative trials for steroids in early sepsis, they may be beneficial in fulminant cases.

may also be a microcirculatory component, with shunting through some capillaries and reduced perfusion through others. Cardiac output may be raised as initially myocardial contractility is maintained. With organ dysfunction cardiac output will decline, and perfusion often requires inotropic support.

Therapeutic strategies Throughout the 1980s investigators sought to manipulate the effects of sepsis, and several therapeutic initiatives have been extensively studied. Basic support of the cardiovascular system involves optimization of fluid administration, as the patient is often functionally hypovolaemic with a vasodilated and leaky circulation. Fluid therapy needs to be aggressive, large quantities are required, and the ICU is the ideal environment for monitoring filling status (changes in stroke volume or central venous pressures in response to filling) along with close monitoring of blood pressure and end-organ effects, such as arterial pH and urine output. The choice of fluid is often debated, but colloids are generally used in the UK. Synthetic colloids have much to commend them, as albumin often leaks away from the circulation, dissipating its effect.

Optimization of the circulation It is now widely accepted that simple measurements such as vital signs and urine output do not predict survivors from non-survivors in the critically ill. Studies in highrisk surgical patients demonstrated that data obtained from Swan-Ganz catheters describing cardiac output, the amount of oxygen delivered to and consumed by the tissues, do predict survivors. It is possible to describe the

733

lowest level of these parameters associated with survival, and these parameters were then used as a target for patients to achieve - so-called 'goal-directed therapy'. Three typical patterns emerged from these studies. First, a group of patients who naturally respond to critical illness by raising oxygenation indices: these patients tended to survive. Second, patients who had suboptimal indices but in whom it was possible to achieve the target oxygen delivery by raising the cardiac output with either fluid therapy or inotropic support, or increasing transport capacity by blood transfusion. Lastly, some patients were never able to achieve the target, and this group had a very poor prognosis. Much of this information has been derived from studies in high-risk surgical cases whose haemodynamics were studied in a controlled theatre environment. When this approach was applied to the general population of critically ill on the ICU, many workers found that the 'goals' that predict survival could not be achieved, often because cardiac output could not be raised sufficiently. It is now recognized that the distinction lies between those who can be optimized prior to a major inflammatory insult (e.g. major surgery) and those in whom the inflammatory processes are already established (e.g. SIRS, severe sepsis and septic shock). In these latter patients attempts at achieving goals were often associated with a worsening prognosis. Nevertheless, the importance of filling to maintain the cardiac output, and attention to treatment of the circulation (especially when patients are acidotic), should be emphasized. Many patients with sepsis will require intropic support in addition to fluid therapy, although aggressive cooling of septic patients with a pyrexia may prevent or reduce the need for inotropes. The choice of inotrope depends on cardiac output, with noradrenaline (norepinephrine) being the agent of choice for hyperdynamic sepsis and adrenaline (epinephrine), with or without intravenous nitrates, being a suitable option for those with a low cardiac output.

Inflammatory mediators The inflammatory response seen in sepsis is frequently initiated by endotoxin derived from surface molecules of Gram-negative bacteria. The binding of endotoxin complexes to cell surface receptors begins mediator release, characterized by production, amplification and numerous positive feedback loops between the mediators. This cascade can often continue when the precipitating agent (infection) appears to be controlled. One group of mediators involved are metabolites of arachidonic acid, the eicosanoid family, and they share a similar hairpin-shaped molecule with two fatty acid side

1

734

MCQ 14.4

arms and a central ring structure. Arachidonic acid is broken down by cyclooxygenase or lipoxygenase pathways. They include the prostaglandin, thromboxane and leukotriene groups. Many of the clinical signs of sepsis can be induced by these molecules, and their synthesis is increased in septic animals and humans. It is possible to ameliorate the effect of sepsis by altering eicosainoid production or by the use of anti-inflammatory drugs, strengthening the suggestion that this group of agents is of central importance. Further effects during sepsis come from neutrophils, macrophages and the B and T lymphocytes. Many cytokines are released from these cells, including tumour necrosis factor (TNF), interleukins, especially IL-1, IL-4, IL-6, IL-8, IL-10 and IL-13, interferon, and the colonystimulating factors G-CSF, GM-CSF and M-CSF. These cytokines have multiple roles in the inflammatory response, including increasing inflammatory cell numbers through increased bone marrow production and release; many also delay leukocyte apoptosis. These mediators also prime and activate inflammatory cells, regulate adhesion molecule expression and stimulate the production of other mediators, such as free radicals, the eicosanoids and platelet-activating factor (PAF) (Fig. 14.8A). Numerous attempts have been made to block specific mediators and thus to alter the course of the septic process, including antiendotoxin antibodies, anti-TNF therapy, the use of natural inhibitors such as IL-1 receptor antagonist and antagonists against PAF, although so far these approaches have not been successful. This failure is due partly to the enormous redundancy of the inflammatory cytokine network, with huge overlap in the effects of many cytokines such that no single mediator occupies a pivotal role that cannot be replaced by other cytokines. 1

Nitric oxide Many patients present with hypotension, and indeed the term septic shock was used to describe this ominous development. Hypotension is associated with a worse prognosis and over 50% of patients may not survive. Nitric oxide (NO) is a ubiquitous molecule produced from L-arginine by nitric oxide synthetase (NOS), of which there are several isoforms, distinguished by the cells that synthesize them and by whether enzyme levels can be induced by proinflammatory mediators. NO is produced constitutively by the endothelial enzyme eNOS (also termed NOS III), where it is central to regulation of normal vascular tone. Another NOS is produced by neuronal cells and called nNOS or NOS I, but the most important NOS in sepsis is believed to be an inducible NOS produced by macrophages and other cells and called iNOS or NOS II. It is believed that this leads to excessive production of NO in inflammatory conditions, where it contributes to the severe hypotension so characteristic of sepsis. NO functions as an intracellular messenger, involved in many subcellular signalling functions, and can directly inhibit

14

FIG. 14.8 Mediator response to sepsis A Many mediators are produced after an insult, leading to an inflammatory response, B The response to sepsis falls into three types. Little response or overwhelming response may both result in death. C eNOS is constitutive and requires Ca2+ to be activated, allowing it to be turned on and off quickly, whereas iNOS needs to be transcribed, and this process happens 2-4 hours after activation or injury. When activated, the amount of iNOS is substantially higher, leading to a many-fold amplification of activity.

735

cellular function, for example acting as a competitive inhibitor of the mitochondrial electron transfer chain. It is also cytotoxic and is part of the normal armoury of the neutrophil and macrophage for killing phagocytosed bacteria. Its cytotoxic actions may, however, also contribute to the organ damage seen in severe sepsis. Of great therapeutic interest is the finding that refractory severe hypotension has been treated with NOS inhibitors, analogues of Larginine, which block NO production and enable the blood pressure to rise (Fig. 14.8B). However, a recent multicentre trial of an NOS inhibitor in severe sepsis failed, with an excess mortality in certain subgroups, although this may have been due to reductions in cardiac output in the highdose groups. Paradoxically there are circumstances where it may be advantageous to deliver NO rather than to block it. In acute lung injury, hypoxaemia is often associated with patchy collapse and atelectasis, and perfusion to many parts of the lung is reduced. NO (a gas) can be administered mixed with the gas from the ventilator and acts to vasodilate the pulmonary circulation, but only the blood vessels associated with ventilated alveoli. This leads to a rapid increase in the Pao2, although it is unclear whether this (transient) rise in oxygenation translates into a persisting clinical benefit. The inhaled NO is rapidly bound to haemoglobin, clearing it from the circulation and preventing systemic hypotension. The effect of NO in terms of mortality in acute lung injury is the subject of current clinical studies.

RESPIRATORY PROBLEMS Respiratory failure has many causes (p. 670, Table 13.32). Severe hypoxaemia with or without hypercapnia is a common indication for transferring a patient to the ICU for monitoring and, where necessary, mechanical ventilation. A particularly common error leading to poor management is restricting the amount of oxygen delivered to patients in respiratory distress. Loss of hypoxic drive to breathe does exist but affects only a small minority of patients. Despite this, it is common to find patients with low saturations being given restricted oxygen therapy. Far more appropriate is to give higher levels of oxygen to patients in respiratory distress and observe them closely for any signs of hypoventilation. It should be recognized that elevations in Paco2 are not usually due to loss of hypoxic drive to breathe but to many other factors, of which exhaustion, with small tidal volumes, is particularly common. These patients will benefit from unrestricted

1

736

MCQ14.5

2

MCQ14.6

SUMMARY 3 Tissue oxygenation - important

measurements

Tissue perfusion - cardiac output Oxygen carriage in blood - Pao2, Sao2, Hb Tissue uptake of oxygen - mixed venous oxygen tension and saturation Metabolic acidosis and lactic acid accumulation

oxygen therapy, with good physiotherapy and correct positioning (sitting up), and restricting their oxygen to 24 or 28% will not correct their tissue hypoxia and will thus not help in correcting the underlying problem. Another common error is to take a blood gas on air rather than on oxygen. There is no need to take the patient off oxygen just to see how low the Pao2 falls, as this can be detrimental and is unnecessary, as the Pao2/Fio2 (see below) provides a clear indication of the oxygenation.

Assessment of ventilation and tissue oxygen delivery During steady-state conditions, when the amount of CO2 from the tissues is constant, arterial CO2 tension is dependent on alveolar ventilation. The adequency of ventilation can therefore be judged by Paco2. As ventilation becomes inadequate, CO2 is retained and eventually the patient becomes acidotic. Over a number of days acclimatization to hypercapnia may occur, and the respiratory acidosis is compensated for by a metabolic alkalosis (p. 1115). Despite the undoubted value of CO2 measurements in the assessment of ventilation, it should be emphasized that in some patients (e.g. those with severe asthma) there can be acute respiratory distress with CO2 levels that are normal or reduced. Adequacy of oxygenation is more difficult to assess. Arterial oxygen tension should be judged against the added oxygen the patient is breathing. In disease, as gas exchange deteriorates so the difference between the inspired oxygen tension and that of arterial blood increases (the A-ao2 gradient widens; p. 670). The size of the A-ao2 difference indicates the failure of the lung to perform gas exchange (see Table 14.6). An indication of this deficit is easily assessed by calculating the Pao2/Ho2 ratio. In health, if the blood gas is measured in kPa this ratio is greater than 65 (450 when blood gas is measured as mmHg) (e.g. 13.3/0.21 = 66.5), and severe disease is indicated when the ratio falls, with the Pao2/Fio2 for acute lung injury being less than 40kPa (300 mmHg) and for acute respiratory distress syndrome (ARDS) less than 26.7kPa (200mmHg). Reduction in the ratio may be due to impaired diffusion or ventilation/perfusion (V/Q) mismatch, and added oxygen can correct these defects. However, if the hypoxaemia is due to blood bypassing ventilated alveoli (physiological shunt), then increasing the oxygen fraction will not dramatically improve blood gases. Both the level of added oxygen that is required and the rate at which it is neces-

TABLE 14.6 Assessment of oxygen uptake by the lung The difference between alveolar and arterial oxygen tension is normally small. The extent of the alveolar-arterial (A-ao2) gradient provides a useful measure of the severity of respiratory failure. Alveolar-arterial gradient (A-a) = oxygen tension in alveolar air (PAo2) minus oxygen tension in arterial blood: A - ao2 = PAo2 - Pao2 The oxygen tension in alveolar air (PAo2) = oxygen tension in inspired air (P102) minus the arterial C02 tension (Paco2)/Respiratory Quotient (R): PAo2 = Pio2 - Paco2/R Thus in a normal subject. PA02

- [0.21 x (760 - 47)1 - 40/0.8 = 100mmHg A-ao 2 = 100-98 = 2mmHg (normal gradient is small, rising with age, and is less than 15mmHg; 2kPa) For a patient with severe respiratory failure on the ICU, ventilated with 50% oxygen who has an arterial Pao2 of 80mmHg, Paco2 of 48mmHg, the results could be: PA02

- [0.5 x (760 - 47) - (48/0.8)] = 297mmHg A-ao 2 = 297-80 = 217mmHg Serial measurement of the A-a gradient provides excellent information on oxygenation and allows day-to-day comparisons despite changes in fio2 and Paco2. * Water vapour pressure.

sary to increase this fraction reflect the severity and progression of disease. Arterial blood gases provide a single respiratory assessment of the patient, and can be misleading when the patient's circumstances are changing rapidly. Repeated samples are often preferable and require the insertion of an arterial cannula. The clinical context of the arterial sample (e.g. Fio2, ventilator settings) should always be documented. Oximeters that continuously and non-invasively measure oxygen saturation (Sao2) are very useful. Oxygen must be delivered to the tissues, and therefore the state of oxygen transport and uptake from the circulation is important. Transport of oxygen depends on the amount of haemoglobin and its ability to take up oxygen (O2 dissociation curve, p. 606). The cardiac output and state of the peripheral microcirculation then determine the delivery of oxygen to the respiring cells. 1

ACUTE RESPIRATORY DISTRESS SYNDROME (ARDS) This is one of the success stories for intensive care, as mortality has declined steadily from 70 to 90% in the 1970s to approximately 40% in the late 1990s. ARDS is defined as a refractory hypoxaemia (Pao2/Fio2 <26.7kPa) with diffuse shadowing on the chest X-ray resulting from a known

acute precipitating cause and not related to left heart failure. ARDS is classically viewed as beginning with a phase of capillary leak, proceeding to an exudative phase and then to a fibrotic phase. This view is simplistic, and there is good evidence that inflammation and fibroproliferation occur in parallel and both are present at the very earliest stages of the condition. At present there are no specific therapies to prevent progression to ARDS, although there is increasing evidence that the use of a prolonged course of methylprednisolone in those who are not improving after 1 week does improve outcome, probably through an effect on fibroproliferative mechanisms. Otherwise the management of ARDS is conservative, with most (but not all) patients requiring mechanical ventilation while awaiting a recovery of lung function. This period of support can be prolonged and often is complicated by pneumothoraces and episodes of ventilator-associated pneumonia (see below).

14

INDICATIONS FOR MECHANICAL VENTILATION Intubation and ventilation are easier to undertake as a considered manoeuvre than as an emergency event. The entire clinical situation is relevant, and ventilation is seldom indicated solely on blood gas results. An increasing demand for oxygen over a few hours, failure to oxygenate the patient adequately with a face mask, or mental deterioration, indicate the likely need for mechanical support. Patients who are initially distressed often have a normal or slightly lowered Paco2, with rapid shallow respirations. Further distress, an increase in the respiratory rate or increasing sympathetic activity (tachycardia, hypertension) indicate greater difficulty in eliminating CO2, and exhaustion, with associated hypercapnia, is a common end-point prior to ventilation. The clearance of secretions by the patient is dependent on the ability to cough vigorously, and sputum retention may worsen or precipitate respiratory failure. Endotracheal intubation allows efficient removal of secretions by suction, protects the upper airway from aspiration, and should be considered in all patients with a Glasgow Coma Scale (GCS) of less than 8. Many postoperative patients are ventilated electively to provide a period of stabilization. Some patients are intubated during resuscitation. Mechanical ventilation is sometimes required to provide a stable setting during the management of the pulmonary oedema of left heart failure, where it reduces the workload of the myocardium and can improve oxygenation. For artificial ventilation to be initiated, the patient must be anaesthetized and intubated. Patients may already be critically ill; if so, the cardiovascular effects of the anaesthetic agents, loss of reflexes guarding the larynx, mechanical problems in passing the tube and the effects of positive intrathoracic pressure on venous return make intubation and ventilation a potentially dangerous procedure. 2

737

CASE STUDY 14.1 PROGRESSIVE HYPOXAEMIA IN A 38-YEAR-OLD MAN WITH PNEUMONIA A 38-year-old banker presented to casualty with a 3 day history of fever, pleuritic left-sided chest pain and breathlessness. He had no significant previous medical history and was a non-smoker and non-drinker. On examination he was clearly unwell, flushed and his temperature was 38.9°C; the pulse was 120 and regular, and blood pressure was 105/45 mmHg. He was tachypnoeic with a respiratory rate of 32 breaths per minute, and there was marked bronchial breathing at the left lung base. The oxygen saturation was 94% on 60% inspired oxygen and arterial blood gases showed a pH of 7.28, Paco2 4.1 kPa, Pao2 9.7 kPa and a base deficit of -7. The white count was elevated but his biochemistry was normal. A clinical diagnosis of left lower lobe pneumonia was made and confirmed on chest X-ray. He was admitted to the high dependency unit for close monitoring where he was reviewed by the intensive care team. He was commenced on intravenous cefuroxime and erythromycin (changed the next day to intravenous benzylpenicillin when blood cultures grew S. pneumoniae}. He had a central venous line inserted and required 3 fluid challenges of 200 ml of colloid each to raise his mean blood pressure to 60 mmHg. He was managed with continuous humidified oxygen (60%) and his oxygen saturations, blood pressure and clinical state were closely monitored. Analgesia was prescribed to reduce the pleuritic chest pain and enable an effective cough. Overnight he remained breathless and his saturations began to fall below 90% despite maximal facemask oxygen. The chest X-ray now showed progressive shadowing with diffuse changes throughout both lung fields (Case Fig. 14.1.1). To achieve adequate oxygenation he was intubated and mechanically ventilated; with a tidal volume of 375 ml, positive end expiratory

738

pressure (PEEP) of 15 cm H2O, and a fractional inspired oxygen concentration of 0.85 which give a Pao 2 of ll.lkPa.

Question 1. What is the diagnosis and wliat js the correct rttianagement?

The diagnosis is now that of the acute respiratory distress syndrome (ARDS), as the Pao2 to Fio2 ratio is <26.7kPa (11.1/0.85 = 13.1), there is diffuse chest X-ray shadowing, and a recent precipitating cause (pneumonia). A CT scan taken a few days later showed typical ARDS changes (Case Fig. 14.1.2). For complete accuracy the possibility of left ventricular failure as a cause of these changes needs to be eliminated (this can be achieved with a pulmonary artery catheter, demonstrating a pulmonary artery

wedge pressure of <18mmHg, or by echocardiography). Correct management is treatment of the underlying cause (S. pneumoniae) and supportive therapy. A recent large multi-centre trial has shown that limiting tidal volumes and airway pressures leads to a 25% reduction in mortality. The accepted upper limit for tidal volumes is 0.5 ml/kg. Oxygenation is maintained using high PEEP, with values of 10 to 20 cm H2O being appropriate for this patient (although there is little definitive evidence for how PEEP should be titrated). Early enteral nutrition is beneficial for these patients. Management of fluid balance is controversial; although most clinicians would aim to maintain patients appropriately filled (best done by monitoring stroke volume and response to fluid challenges), some would aim to restrict fluids and maintain pressures using inotropes, and this issue is currently being addressed in a large multi-centred trial in the USA.

CASE FIG. 14.1.1 Chest X-ray. There is dense consolidation in the left lower lobe and diffuse shadowing throughout both lung fields. The patient developed a right pneumothorax (relatively common in ventilated ARDS patients) requiring tube drainage.

14

CASE STUDY 14.1 CONTINUED Question 2. What rescue therapies are available? A number of trials have been conducted examining the use of different ventilatory strategies (including extracorporeal membrane oxygenation - ECMO), surfactant replacement, inhibitors of inflammatory pathways (e.g. cyclooxygenase inhibitors) and free radical scavengers. As a treatment for severe sepsis (rather than ARDS), activated protein C has been shown to be beneficial; however, to date, as a specific therapy for ARDS, the only trial with convincing data showing improved outcomes (aside from the limitation of tidal volumes) is that for the late use of low doses of steroids. This trial suggests that in those patients who fail to improve, or in whom there is a deterioration by day 7 after initiating mechanical ventilation, a trial of methyl prednisolone (2 mg/kg daily for 14 days) should be considered. There is reasonable evidence that this improves lung function and survival although the risk of infection needs to be considered. In those that respond, steroids can be weaned off slowly over the next 16 days, and the non-responders (usually less than 30% of cases) can be weaned more rapidly. However, the evidence is based on a trial with relatively small numbers, and the definitive role of steroids in ARDS awaits the results from an ongoing trial run by the National Institutes of Health in the USA.

Question 3. What is the outcome for patients with ARDS and what is the long term prognosis for iiirvivors?

Mortality from ARDS has fallen steadily over the last few decades and is now approximately 40%. Patients with a high number of organ failures, older patients and those with liver failure do worse. The ICU and hospital stays for survivors are still prolonged, with many patients being hospitalized for three months or more. The pulmonary fibrosis so characteristic of ARDS can lead to persisting respiratory problems after patients leave ICU. Recovery in lung function can be slow, with some

patients taking up to twelve months to return to baseline, while a few may show some degree of chronic abnormality including low diffusion capacity and elevated pulmonary artery pressures. In the majority, however, lung mechanics recover fully suggesting that the pulmonary fibrosis in ARDS is reversible. The prolonged period of hospitalization and slow recovery, along with the nature of the initial injury and the invasive therapies required to support life on the ICU, can lead to significant psychological stress and many patients benefit from counselling and from close follow-up after discharge from hospital. The patient described in this case survived, but when receiving ventilation he developed a pneumothorax and in all required 7 weeks of ICU support. One year after his illness the lung function was virtually normal.

CASE FIG. 14.1.2 CT Scan. Widespread airspace consolidation of ARDS, most marked in the dependent parts of the lungs.

METHODS OF VENTILATION The ventilator provides the motive power to drive oxygenenriched air into the patient's lungs. This may need to be done at high pressure as, for example, when ventilating an asthmatic with severe bronchoconstriction. With modern

ventilators it is possible to determine the respiratory rate, tidal volume, driving pressure, and the timing and waveform of the inspiratory cycle. Many patients on the ICU are deeply sedated and sometimes paralysed. All the ventilatory requirements of such patients are provided by the mechanical ventilator: this is termed controlled mandatory ventilation (CMV). Other patients often require partial

739

support from the ventilator, and may breathe spontaneously between machine breaths. In order to avoid a machine breath coinciding with spontaneous exhalation the inspiratory cycle of the ventilator is synchronized to the patient's own inspiratory effort. This is sensed at the start of each breath, and the inspiratory cycle of the machine is initiated. The machine rate and tidal volume are set to provide adequate alveolar ventilation; this is termed synchronized intermittent mandatory ventilation (SIMV). Newer ventilators can now assist each spontaneous breath, the level of assistance ranging from partial support to full ventilation. This is performed by providing a constant, positive pressure to the patient with each spontaneous breath (inspiratory pressure support, IPS). The amount of IPS can be adjusted up and down. However, this form of respiratory support is critically dependent on the patient's spontaneous effort, and will provide no support if the patient becomes apnoeic. A further feature of ventilators is the capacity to add a positive (greater than atmospheric) pressure throughout the expiratory phase of the respiratory cycle. The purpose of such positive end-expiratory pressure (PEEP) is to prevent collapse of peripheral airways and subsequent atelectasis, thereby improving gas exchange. The addition of PEEP - commonly 5-10cmH2O - usually improves arterial oxygen tension, but impedes venous return and can reduce the CI. Oxygen delivery to the tissues may therefore be decreased despite an improvement in Pao2. PEEP may also increase pressure damage to the lung. The optimum level of PEEP for a particular patient is that which maximizes tissue oxygen delivery.

NON-INVASIVE VENTILATION

740

Mechanical ventilation does not require endotracheal intubation, and respiratory support is frequently provided with a tightly fitting face mask during resuscitation and by anaesthestists in the operating theatre. This technique has been adapted in order to support patients on the ICU, using a simple ventilator and accepting that some of the ventilated gas may leak away. Such support is non-invasive and has many advantages. The patients do not need to be intubated, are not sedated, and may spontaneously cough and expectorate their own secretions. Additionally, the face mask may be removed intermittently and the patient may be able to speak and communicate during therapy. This type of treatment is now termed non-invasive ventilation (NIV). NIV is a generic term used to describe any form of positive-pressure facemask therapy; it begins with CPAP (continuous positive airway pressure) and also includes volume- or pressure-controlled mechanical ventilation. NIV was initially used in patients with severe chronic hypercapnic respiratory failure as an alternative to mechanical ventilation with an artificial airway. Success in this area is widespread, and the treatment is no longer restricted to the ICU. Rather, ward-based care is

usual. Frequently patients require long-term treatment and domiciliary NIV is provided, with specifically designed ventilators. NIV is used in acutely ill patients in respiratory failure. NIV is particularly successful when used at an early stage when the patient can cooperate effectively. Several studies have demonstrated that NIV decreases morbidity and mortality in acute ventilatory failure due to chronic destructive pulmonary disease (COPD), possibly owing to the decrease in pneumonia associated with endotracheal intubation. Not all patients are suitable for this form of therapy. Patients with decreased level of consciousness, those who are unable to cough and expectorate, or who are unable to swallow or protect their upper airway are difficult to manage with NIV and may need to be intubated. NIV demands an effective team of nurses, doctors and physiotherapists. In the acutely ill the first few hours of treatment demand intensive input from the team. The case for NIV in hypoxaemic respiratory failure is more complex, and some studies have suggested that treatment may not prevent more invasive therapy involving endotracheal ventilation. Nevertheless, the use of NIV is widespread.

COMPLICATIONS OF VENTILATION There is significant morbidity and occasional mortality associated with mechanical ventilation. Pressure damage to the lung is termed barotrauma, and the incidence is increased when high airway pressures are required to ventilate the lungs. A recent large multicentre trial in the USA has demonstrated improved survival in patients with acute lung injury or ARDS who are ventilated at lower tidal volumes (5-6mL/kg) rather than standard tidal volumes (10-12 mL/kg), although the exact mechanisms whereby this benefit is achieved are not yet clear. Pneumothorax is not uncommon and may present as gas in the pleural space, mediastinum or soft tissues (surgical emphysema). A tension pneumothorax can occur. When drained, the pleural leak may persist (bronchopleural fistula) and, if severe, the volume of gas escaping from the leak may exceed the capacity of the ventilator, leading to hypoventilation. Selective ventilation of each lung (differential ventilation) can isolate the side with the leak. High levels of oxygen produce lung damage and cause or exacerbate the acute respiratory distress syndrome (ARDS, p. 697). Patients on ventilators are often deeply sedated and therefore totally dependent on the correct functioning of equipment. They are also dependent on a supply of humidified, oxygen-enriched air. Constant nursing care is crucial. Secretions in the respiratory tract must be aspirated by the ICU staff, and airway collapse and atelectasis are common complications. The artificial airway is a portal for the entry of pathogens. Patients can be ventilated via an endotracheal tube for several weeks, but because the tube is a constant source of

TABLE 14.7 Major indications for tracheostomy • • • •

Provision of prolonged positive pressure ventilation Control of secretions To prevent aspiration To bypass upper airway obstruction

irritation to the mouth, a tracheostomy is often performed when it is likely that an artificial airway will be required for a prolonged period (Table 14.7).

TABLE 14.8 Metabolic abnormalities that impair respiratory muscle contractility • • • • • • • • •

14

Acidosis Hypophosphataemia Hypomagnesaemia Hypokalaemia Hyper- and hypothyroidism Steroid therapy Other drugs (e.g. drug-induced myasthenic syndromes) Hypoxia Hypercapnia

WITHDRAWAL OF MECHANICAL VENTILATION As patients on the ICU recover, the aim is to withdraw mechanical ventilation. For this to be feasible the patient's cardiovascular system must be stable, with good peripheral perfusion. Gas exchange must be good, with satisfactory oxygenation and an inspired oxygen concentration of less than 50%. The patient should not be septic. There should be no gross acid-base or electrolyte abnormality. In patients where the period of ventilation has been short (hours), the withdrawal of assisted ventilation usually poses few problems. Sedation is withdrawn, the patient wakes up and breathes spontaneously from a source of oxygen, and is then extubated. In patients who have been ventilated for a number of days it is often not possible to withdraw ventilation suddenly, and a period of gradual withdrawal, or weaning, is necessary. Successful weaning requires careful consideration of the following interrelated questions: • Is central drive optimal? • Is the respiratory muscle pump as effective as possible? • Has the load on the respiratory muscle pump been reduced to the minimum? • Given the capacity of the respiratory muscle pump (Fig. 14.9), can the ventilatory load be sustained or is fatigue likely? Gradual withdrawal of ventilation can be achieved in many different ways. When the different methods are compared, SIMV is consistantly worse than using pressure support or a simple T-piece breathing system. Considerable effort has gone into the development of 'smart' ventilators which can detect when the load applied to the respiratory muscles changes. This allows the ventilator to automatically compensate, increasing ventilatory support. Alternatively, if the load diminishes, support may be reduced. A gradual reduction in the number of machine breaths, using SIMV, allows the patient to breathe spontaneously with a background of support from the ventilator. Alternatively, the inspiratory pressure support (IPS) can be gradually reduced so that the patient progressively takes

FIG. 14.9 The respiratory muscle pump When assessing the likelihood of successful weaning from mechanical ventilation the key factors are ventilatory load, drive and capacity. If the demands made of the pump exceed capacity, weaning will fail.

over the work of breathing. Another approach is to allow the patient to breathe spontaneously and unaided for successively longer periods, alternating with machine ventilation. In some patients the work of breathing can be reduced and weaning facilitated by continuous positive airway pressure (CPAP), which can be continued post extubation via a face or nasal mask. Some patients can be weaned by using non-invasive ventilation. During weaning it is important to optimize the function of the respiratory muscle pump. Any metabolic disturbances known to reduce the contractility of the respiratory muscles should be corrected (Table 14.8). In patients with airways obstruction bronchodilatation is helpful in reducing hyperinflation and improving muscle efficiency, as well as reducing ventilatory load. Treatment of pulmonary oedema also reduces the work of breathing. The effort of breathing through the tubing, ventilator and endotracheal tube should be minimized and the airway made as comfortable as possible. The patient should be sitting upright, preferably in a chair, and the abdomen should be free to allow the diaphragm to descend. During weaning, nutritional support must be maintained.

741

fying, especially as the patient has difficulty with communication, and much psychological support is needed. In the early stages of weaning the patient should receive full ventilation during the night, and the benefit of a full night's sleep. Sedative drugs are in general best avoided, but may be required if the patient is sleep deprived. 1

CENTRALNERVOUSSYSTEMPROBLEMS Following an injury, the brain is at risk from the development of secondary damage that is essentially ischaemic in nature. Many factors can contribute to secondary brain injury: these include expanding intracranial haematomas, systemic hypotension, hypoxaemia and raised intracranial pressure (ICP). ICU therapy is aimed at protecting the brain from further damage to allow the best chance of neurological recovery. Treatment includes maintenance of cerebral oxygenation and perfusion, cerebral protection, and management of raised ICP. Appropriate monitoring allows a rational approach to therapy.

PATHOPHYSIOLOGY OF BRAIN INJURY Cerebral ischaemia

FIG. 14.10 An algorithm for weaning from mechanical ventilation

Injury to the brain initiates a cascade of ionic and metabolic changes that cause brain swelling and a reduction in cerebral blood flow (CBF), leading to a vicious cycle of worsening cerebral ischaemia. As blood flow is reduced there is sequential loss of cellular function that leads to uncontrollable movement of ions (especially calcium and potassium) across the cell membrane, and ultimately to cell death. Ischaemic brain tissue releases inflammatory mediators and excitatory amino acids that worsen cerebral injury.

Cerebral blood flow and metabolism Several factors distinguish patients who will sustain spontaneous ventilation from those who are ventilator dependent. A simple test to assess the likelihood of weaning success is to assess rapid shallow breathing by measuring the f/Vt ratio (f = frequency, Vt = tidal volume measured in litres). If this ratio exceeds 105 (i.e. rate of 40, Vt = 0.31,f/Vt = 133), weaning is unlikely. A suggested application of this test is shown in Figure 14.10. The patient should not be allowed to become distressed off the ventilator. The sensation of breathlessness is terri-

1

742

Case 14.1

The effects of severe head injury on CBF and metabolism are variable. The majority of patients suffer an initial reduction in CBF that puts them at risk from further ischaemic damage. This is usually followed by a return to normal or a period of higher than normal flow (hyperaemia). Normally CBF and cerebral metabolism are coupled, to allow blood flow to be directed to metabolically active areas. After head injury this relationship may be lost, and in particular may vary in different parts of the brain. Under normal circumstances CBF is also controlled by autoregulation of cerebral vessels. After severe head injury autoregulation may be abolished and CBF changes directly with cerebral perfusion pressure. Systemic hypotension therefore increases the risk of cerebral ischaemia and is associated with a poor neurological outcome after head injury.

Intr a crania I pressure The cranium is a rigid box containing brain, cerebrospinal fluid (CSF) and blood, and the pressure inside the head (the ICP) is determined by the contents. Figure 14.11 shows the relationship between intracranial volume and pressure. In normal individuals a small increase in the volume of one compartment of the intracranial contents is compensated for by a reduction in another, thereby maintaining a constant pressure (point A in Figure 14.11). For example, as the brain swells, CSF and blood are driven out of the intracranial cavity. However, if there is continued brain swelling a stage is reached where compensation can no longer take place, and further small increases in volume cause large increases in pressure (points C-D in Figure

14.11). The blue lines in Figure 14.11 show the changes in pressure that occur when small aliquots of fluid are added via a catheter placed in a lateral ventricle. At the beginning of the curve, addition of a small volume of fluid causes a small, transient rise in ICP. Further along the curve, the addition of a similar volume of fluid causes a substantial rise in ICP although the baseline pressure has only risen slightly. Rises in ICP are important for two reasons. The cerebral perfusion pressure (CPP) is related to the mean arterial blood pressure (MAP) and ICP in the following way:

14

CPP = MAP - ICP

Increase in ICP may therefore reduce the supply of blood to the brain. A reduction in CBF below a certain critical level results in cerebral ischaemia. Intractable rises in ICP may also cause direct pressure effects on the brain and a downward shift of brain substance. This is termed coning and leads to recognizable clinical features, including the Cushing response (hypertension and bradycardia), ipsilateral pupillary dilatation and abnormalities of respiratory pattern, including apnoea. It should be noted that these changes occur late, often when irreversible brain damage has occurred. Coning therefore adversely affects outcome and is often fatal.

Causes of raised ICP

FIG. 14.11 The intracranial pressure-volume relationship • Between points A and B an increase in the volume of one component of the intracranial contents does not cause a rise in intracranial pressure because of a compensatory reduction in volume of the other two. • Between points C and D the intracranial compliance is critically low and a small rise in volume causes a substantial rise in pressure. • The intracranial compliance can be determined by a small volume challenge (blue lines)-see text.

Rises in ICP may occur for a variety of reasons. Spaceoccupying lesions, such as a haematoma, can expand. The ICP can also rise because of brain swelling, and this is usually due to cerebral ischaemia. Dilation of cerebral vessels causes an increase in cerebral blood volume (CBV) and ICP. Cerebral hyperaemia may occur as a direct result of brain injury (see above), but also because of dilatation of cerebral vessels secondary to a rise in Paco2. Cerebral vessels are exquisitely sensitive to changes in arterial carbon dioxide tension, and to a lesser degree to changes in oxygen tension. Figure 14.12 shows that a small rise in

FIG. 14.12 Cerebral blood flow [A] Effect of arterial C02 tension (Paco2) on cerebral blood flow. CBF changes in a linear fashion with Paco2 between 3.3 and 10.0kPa. B Effect of arterial 02 tension (Pao2) on cerebral blood flow. CBF does not change until Pao2 falls below 5.0kPa.

743

TABLE 14.9 Indications for cranial CT scan following bead injury 1. Immediate referral - Patient in coma - GCS <8 - Deterioration in conscious level - GCS falls by >2 - Focal pupil or limb signs 2. Urgent referral - Contusion persisting for >8 hours - Depressed or compound skull fracture - Base of skull fracture - Signs of rising ICP

TABLE 14.10 Targets for treatment of head injured patients • • • • •

Pao 2 >13.5kPa Paco2 4.0-4.5 kPa Mean arterial blood pressure >100mmHg Cerebral perfusion pressure >70mmHg Intracranial pressure <20mmHg

haematomas or the presence of brain swelling. The indications for an urgent CT scan are listed in Table 14.9. Management

CO2 can lead to a substantial rise in CBF whereas a low arterial CO2 cause a reduction in CBF. It is therefore crucial that Paco2 is maintained within normal limits during management of brain-injury. Severe hypoxaemia also increases CBF (Fig. 14.12) and ICP. Brain swelling or space-occupying lesions may obstruct CSF outflow and lead to a secondary hydrocephalus that may also contribute to raised ICP. 1

INITIAL ASSESSMENT It can be difficult to assess the central nervous system in patients on the ICU because the usual signs may be masked by sedation. Conscious level is usually assessed by the Glasgow Coma Score (GCS), based on the best verbal, motor and eye-opening responses. The score is described in detail on page 1283. The GCS is a measure of conscious level and therefore it is also important to document pupillary responses and focal neurological deficits. The initial GCS after resuscitation is a useful prognostic indicator after traumatic head injury, and sequential measurements give an indication of response to treatment. An assessment of the cardiovascular system should be made in all patients following brain injury, and serial recordings made of blood pressure and heart rate. About 40% of head-injured patients have other injuries that might affect outcome, and a full trauma survey should be carried out and appropriate investigations ordered. Up to 10% of patients with severe head injury have an associated cervical spine injury and precautions should be taken to protect the cervical spine (e.g. cervical collar) until this can be excluded by high-quality radiographs or CT scan. A chest radiograph should also be obtained because pneumothorax is a continued cause of mortality after head injury. CT examination of the head is the investigation of choice after severe head injury, and may reveal intracranial 1

744

MCQ14.7

ICU therapy for brain-injured patients is aimed at providing ventilatory and circulatory support to ensure adequate cerebral oxygenation and perfusion. Treatment may also be required to reduce ICP. General targets for therapy are shown in Table 14.10. Ventilation Hypoxaemia occurs in up to 65% of severely head-injured patients, and endotracheal intubation and mechanical ventilation are mandatory to ensure maintenance of adequate oxygenation. Ventilation should also be adjusted to maintain normal PaCO2 levels. Hypercapnia may cause rises in ICP secondary to cerebral vasodilatation, and hypocapnia may lead to critical reductions in CBF and the development of cerebral ischaemia. Blood pressure Prolonged periods of even modest hypotension (systolic BP < 90mmHg) are associated with poor neurological outcome after head injury. Maintenance of a stable cardiovascular system is therefore essential to protect the brain from secondary damage. Fluid resuscitation can be carried out with crystalloids or colloids, but 5% dextrose should be avoided unless hypoglycaemia is suspected. CPP should be maintained >70mmHg, and if the ICP is elevated it will be necessary to increase systemic blood pressure. If adequate blood pressure cannot be achieved with fluid resuscitation alone, inotropes or vasopressors should be introduced. Maintenance of adequate CPP is one of the most important aspects of treatment of brain-injured patients on the ICU. Control of intracranial pressure Provision of adequate analgesia and sedation (see below) prevents rises in ICP during mechanical ventilation and invasive interventions on the ICU. If patients are well sedated but continue to cough and strain, muscle relaxants may be used. Good patient positioning, with a 20° head-up tilt, the neck in the neutral position and neck veins free from obstruction, allows maximum venous drainage from the brain and a reduction in ICP. Surgical evacuation is the primary goal in the presence of an expanding mass lesion causing effects on ICP.

However, in some cases the lesion may be inaccessible or the brain swelling diffuse, in which case other methods must be used. Insertion of a catheter into the lateral ventricle of the brain allows small amounts of CSF to be withdrawn to control raised ICP. Drugs such as mannitol have been used for some time and decrease brain bulk by an osmotic effect. In an acute situation a single dose of mannitol (0.5g/kg) can be lifesaving because it 'buys time' prior to definitive surgical treatment. It can also be used on the ICU to treat increases in ICP, but long-term treatment does not improve outcome. As well as having an osmotic effect, mannitol also lowers blood viscosity and might increase cerebral microcirculatory flow. Excessive use of mannitol results in fluid depletion and hypernatraemia. Barbiturates have been the mainstay of treatment for raised ICP for many years. They increase cerebrovascular resistance and reduce CBF and ICP secondary to dose-dependent suppression of brain electrical activity and oxygen consumption. However, controlled studies have not shown that barbiturates improve outcome after head injury. They cause hypotension that might be detrimental to the injured brain, and are also long-acting. It can therefore be difficult to assess neurological status because of residual sedation when the drug is discontinued. Propofol also causes a dose-dependent fall in ICP and, because of its rapid metabolism and favourable pharmacological profile, is associated with good control of sedation level and ICP. It allows rapid wake-up when it is discontinued, and has become the first-line sedative for brain-injured patients on the ICU. For many years hyperventilation has been used as a treatment for raised ICP. Lowering Paco2 reduces CBF and ICP secondary to constriction of cerebral vessels. However, excessive hyperventilation should be avoided because it can cause a dangerous reduction in CBF and precipitate cerebral ischaemia. It is now known that overzealous hyperventilation is associated with worsened neurological outcome after head injury. Modern practice uses a target Paco2 of 4.0-4.5 kPa. In selected patients lower levels of Paco2 might be beneficial, but hyperventilation to such levels should only be undertaken in association with appropriate monitoring to check the adequacy of cerebral oxygenation (see below).

CEREBRAL PROTECTION There have been many attempts to develop a 'magic bullet' to block the cellular events that occur during cerebral ischaemia and therefore prevent secondary brain damage. It is unlikely that such a solution will be reached in the near future, but there are many strategies available on the ICU that protect the brain from further damage. Maintenance of cerebral perfusion and oxygenation is the cornerstone of cerebral protection, and control of cerebral metabolic rate with sedative drugs prevents oxygen demand outstripping delivery. It is also important to

prevent hyperglycaemia. A high level of plasma glucose leads to elevated levels of lactate in areas of ischaemic brain, and this is associated with poor neurological outcome. Plasma glucose levels should be maintained within the normal range. High temperature is also associated with worsened neurological outcome, and pyrexia must be treated aggressively. It has been suggested that modest reductions in temperature to 33-34°C might improve neurological outcome. Large studies are under way to determine the exact role of hypothermia in the management of brain injury.

14

GENERAL THERAPY Other organ system failure is common in patients with brain injury. Acute lung injury occurs in about 25% of head-injured patients because of pulmonary complications such as aspiration, or because of direct complications of the brain injury itself, such as neurogenic pulmonary oedema. Electrolyte disturbances occur in almost 60% of patients after severe head injury. In particular, disorders of sodium balance occur secondary to diabetes insipidus (DI) or the syndrome of inappropriate antidiuretic hormone secretion (SIADH). Urine and plasma electrolytes must be closely monitored and appropriate treatment initiated (see Ch. 17). Early provision of an adequate calorie intake is associated with improvement in outcome after traumatic head injury, and enteral feeding should be instituted soon after admission to the ICU. An orogastric tube should be used for feeding in the presence of a fracture of the skull base. Because many brain-injured patients require longer-term enteral nutrition, early consideration should be given to placement of a percutaneous gastrostomy. Coagulation abnormalities can also be associated with brain injury because the ischaemic brain releases tissue thromboplastin. Coagulation studies guide replacement of clotting factors. Brain-injured patients fall into a high-risk group for the development of venous thromboembolism. Heparin prophylaxis may increase the risk of intracranial bleeding following craniotomy, and so physical methods of prevention, such as antitrombotic stockings and calf compression, should be used. The physiotherapist plays a key role in the management after brain injury, and in the early stages will be active in the prevention and treatment of chest infection. Early physiotherapy to the limbs and trunk prevents later contractures, and the patient must be sat and stood up as soon as their general condition allows. Such manoeuvres ensure the return of autonomic control of blood pressure and heart rate reflexes, and maintain body and limb alignment.

MONITORING Routine monitoring should include ECG, Spo2, measurement of direct arterial blood pressure, CVP, temperature,

745

urine output and regular blood gas analysis. In the presence of cardiovascular instability an oesophageal Doppler probe or pulmonary artery catheter will allow optimization of circulatory support. Treatment of brain-injured patients with appropriate amounts of sedation removes the clinical signs that would otherwise indicate changes in their neurological status. Brain monitoring is therefore crucial to assess the effects of treatment. Monitoring of ICP allows measurement of CPP and guides therapy. The 'gold standard' for ICP monitoring is measurement of the pressure in the lateral ventricles via a catheter and pressure transducer system. However, under most circumstances ICP is monitored using a miniature sensor placed in the brain tissue via a small burr hole through the skull. Such monitoring devices are safe and reliable. A well as measuring ICP and CPP, pressure monitors are able to detect abnormal ICP waveforms that reflect decreased intracranial compliance. Persistent rises in ICP above 25 mmHg are associated with poor outcome. Measurement of cerebral oxygenation is now standard practice in the ICU management of brain-injured patients. When perfusion of the brain is not adequate, the brain tissue extracts more oxygen from the blood than usual. This results in a reduction in the amount of oxygen in the blood leaving the brain, reflected by a fall in the cerebral venous oxygen saturation. A small catheter can be placed in the internal jugular vein so that its tip lies in the jugular venous bulb. Intermittent blood samples can be withdrawn via the catheter to measure jugular venous oxygen saturation (Sjo2) and venous lactate levels. A reduction in Sjo2 associated with a rise in lactate suggests inadequate cerebral perfusion. Specialized catheters, incorporating fibreoptic technology, allow continuous measurements of Sjo2 to be made in a similar manner to the measurement of mixed venous saturation by pulmonary artery catheters. This allows reductions in cerebral saturation to be monitored in real time and therapy to be initiated early. A reduction in Sjo2 to <50% or multiple episodes of desaturation are associated with poor neurological outcome. Two new technologies have recently been used to measure cerebral oxygenation. Small multiparameter probes that measure brain tissue Po2, ,Pco2, pH and temperature have recently become available. They are inserted into the brain in a similar manner to an ICP monitor, and reductions in tissue Po2 are associated with reductions in focal CBF and poor outcome. Near-infrared spectroscopy (NIRS) is also able to measure cerebral oxygenation by measuring the absorption of near-infrared light by oxygenated and reduced haemoglobin in the brain. The advantage of NIRS is that it is non-invasive and allows continuous monitoring. Both NIRS and brain tissue probes are still

research tools, but offer exciting possibilities for the future of cerebral monitoring. It is difficult to measure CBF at the bedside and the available techniques are invasive or require the use of radioisotopes. Transcranial Doppler (TCD) allows bedside measurement of the flow velocity of blood in the major cerebral vessels. If the diameter of the vessel remains unchanged, flow velocity can be used as an indirect measure of CBF. The TCD waveform can also be analysed to give additional information, such as the presence of vasospasm. The detection of epileptiform activity can be difficult in sedated and paralysed patients. Seizures can occur after head injury, and are associated with a massive rise in cerebral metabolic rate that leads to further secondary damage if untreated. Characteristic changes in the electrical activity of the brain during seizure activity can be monitored using the electroencephalogram (EEG). Analysis of the 16-lead EEG requires specialist training, but processed EEG techniques allow easy recognition of seizure activity on the ICU.

OUTCOME AFTER HEAD INJURY Outcome can be improved by appropriate therapy in the acute phase. No intervention can attenuate the primary injury, but recovery and survival can be jeopardized by secondary complications. The mortality after head injury remains high: of those patients who reach hospital 50% will make a good recovery, whereas 50% will either die or remain severely disabled. It is not possible in the early stages after injury to predict which patients will do well, but prevention of the factors known to cause secondary brain injury is of crucial importance. The major predictors for outcome after head injury are the initial GCS, pupillary size and reaction, and age. This last is by far the most important factor, with young adults having the best chance of a good outcome. GCS < 8 after resuscitation is associated with a worse prognosis than higher scores, and about 40% of patients in coma after 6 hours will die. The prognostic importance of the GCS is shown in Figure 14.13. Recovery from brain injury has three main components, physical, mental and social. Most physical recovery takes place within the first 6 months, but improvements in mental state and social integration may be slower. In general terms, about 80% of recovery is achieved within the first 6 months.

BRAINSTEM DEATH 1

746

MCQ 14.8

Many of the physiological functions of the body can be maintained artificially, but brain function cannot be replaced. Patients who have sustained irreversible struc-

TABLE 14.11 Diagnosis of brainstem death in the UK

14

1. Preconditions • Diagnosis compatible with brain stem death • Presence of irreversible structural brain damage • Presence of apnoeic coma 2. Exclusions • Drug effects (sedatives, hypnotics, muscle relaxants) • Hypothermia • Metabolic abnormalities (sodium, glucose, pH) • Endocrine abnormalities • Intoxication (alcohol, drugs) FIG. 14.13 Post-resuscitation Glasgow coma score and prognosis following traumatic head injury

tural damage to the brain stem as a result of head injury, stroke or cerebral anoxia lie in a deep coma without the capacity to breathe. Prompt medical attention on the ICU may have taken over ventilation and, because of continued oxygenation, the heart continues to beat. Recovery, however, is impossible and this state has become widely known as brainstem death (BSD). In the UK the uncertainty of how to confirm death under such circumstances has been removed by the development of strict guidelines for the diagnosis of BSD. Confirmation of BSD is essential to prevent the continuation of useless treatment, and is a prerequisite for organ donation. Diagnosis of BSD is in three stages (Table 14.11). Certain preconditions must be met. Irreversible structural brain damage compatible with the diagnosis of brainstem death must be present and the patient must be in apnoeic coma (i.e. unresponsive and dependent on mechanical ventilation). Reversible causes of coma must then be excluded. In particular, sedative drug effects, hypothermia and metabolic or endocrine disorders should be considered. Alcohol intoxication or drug overdose should also be excluded. The final part of the diagnosis of BSD is designed to demonstrate the absence of brainstem reflexes and confirm the presence of persistent apnoea by clinical tests (Table 14.11). The diagnosis should be made by two of the most senior doctors available, at least one of whom should be a consultant. If the patient is to become an organ donor, no member of the transplant team may be involved in performing the tests. Two sets of tests for the confirmation of BSD should always be performed to remove the risk of observer error. There is no prescribed time interval between the two sets of tests: this should be determined for each patient individually and is a matter for clinical judgement. In general, the interval will depend upon the primary pathology and the clinical course of the patient. The clinical findings must be documented in the case notes. Although death is not confirmed until the second set of tests have been completed, the legal time of death is when the first set of tests confirms brainstem death. 1

3. Clinical tests • Confirmation of absent brain stem reflexes - No pupillary response to light - Absent corneal reflex - No motor response within the cranial nerve distribution - Absent gag reflex - Absent cough reflex - Absent vestibulo-ocular reflex • Test for confirmation of persistent apnoea - Preoxygenation with 100% oxygen for 10 minutes - Disconnect from ventilator - Insufflate oxygen at 6 L/minute via endotracheal tube to maintain adequate oxygenation during test - Allow Paco2 to rise to >6.65kPa - Confirm no spontaneous respiration - Reconnect ventilator

TABLE 14.12 Suitable organ donors • • • • • • •

Age 0-70 years Patient has suffered irreversible brainstem death Ventilator-dependent No malignancy (except primary brain tumour) No systemic sepsis Hepatitis and HIV negative ABO compatibility

ORGAN DONATION Patients who are BSD might be suitable to become organ donors (Table 14.12). In the UK this is coordinated by UK Transplant (UKT). Regional transplant coordinators are available to support and advise relatives, to offer advice on the management of the potential donor, and to organize the transplant of suitable organs. These can include heart, lungs, liver, kidneys and pancreas, as well as corneas, skin and bone.

747

TABLE 14.13 Physiological complications after brainstem death Hypotension Diabetes insipidus Disseminated intravascular coagulation Cardiac arrhythmias Pulmonary oedema Hyperglycaemia Acidosis Seizures

There is a high incidence of complications likely to jeopardize vital organ function after BSD (Table 14.13). The many physiological changes that occur following cessation of brainstem function cause the multiple organ donor to become very unstable and active support is required. Following the diagnosis of BSD there is a change in the emphasis of care. Therapy has previously been aimed at preserving residual brain function, but after BSD has been confirmed the emphasis of care switches to optimizing organ function for subsequent transplantation. If it is the family's wish that their relative should become an organ donor, it is the duty of the ICU staff to provide organs in optimum condition.

RENAL PROBLEMS

The development of renal failure is a common and serious complication in the critically ill, associated with high mortality. The most common cause of renal failure is relative hypotension (prerenal), although this often proceeds to acute tubular necrosis. Although the major insult to the kidney may have occurred prior to admission, rapid and effective resuscitation may prevent further deterioration of renal function. Incipient and established renal failure are discussed in Chapter 20. Sepsis, sometimes occult, is a common cause of renal failure on the ICU. Other causes must not be forgotten, however, and a renal ultrasound to exclude obstruction is important at an early stage. Commonly, the first indication of renal dysfunction is reduced urinary volume. Assessment and manipulation of the CVS are often important. Despite hypovolaemia, the blood pressure may be normal because of vasoconstriction, but the urine will be concentrated, with a urinary sodium often less than 20mmol/L.

1 748

MCQ14.9

2

MCQ 14.10

Once an adequate circulation has been established some clinicians try to enhance urinary output with lowdose dopamine (<5ug/kg/min), but this is not of proven benefit. A further reduction in urine output is likely to herald established renal failure; although output may be enhanced by high-dose furosemide (frusemide), this may only postpone eventual anuria. Glomerular filtration will be depressed by severe acidosis. Fluid intake should be adjusted to the falling urine output. The multiplicity of intravenous infusions makes volume overload a possible error. Drug dosages should be reviewed in the light of the patient's renal function, and nephrotoxic drugs avoided (e.g. aminoglycoside antibiotics, NSAIDs, furosemide (frusemide)). Indications for dialysis include the control of extracellular volume, electrolyte imbalance and acidosis, and uraemia. Intermittent haemodialysis or peritoneal dialysis can be difficult in the seriously ill. Newer techniques have revolutionized the management of renal failure. Haemofiltration filters can be perfused with blood and require only a small driving pressure to produce a urine-like filtrate. Although this continuous filtration can control extracellular volume, the clearance of urea and electrolytes requires high volumes of filtrate per day and, without skill, fluid imbalance can lead to circulatory collapse. By passing a solution across the membrane against the direction of blood flow, dialysis of urea and other molecules occurs by diffusion and convection gradients. This is now achieved by a continuous-flow machine, removing blood from the venous circulation and returning it into the second lumen of a multichannel central venous catheter. Devices are now available to automatically compute the amount of fluid removed, and fluid balance can be adjusted by simply setting the difference between filtration and concomitant fluid replacement. 1

NUTRITION In normal individuals there is a limited reserve of rapidly available energy substrates (glucose and glycogen), which are depleted within 24-36 hours of starvation. Prolonged fasting leads to the production of glucose from protein catabolism via gluconeogenesis, and metabolism is shifted towards the utilization of fats. In starvation the metabolic rate is decreased and nitrogen loss (from protein breakdown) reduced, with the result that previously well-nourished patients can fast for several weeks. In the severely ill, not only is protein catabolized to glucose at a faster rate, but calorie requirements are also raised, by as much as 125% in those with extensive burns (Fig. 14.14). Many patients are compromised by a degree of malnutrition prior to admission to the ICU. Nutritional failure is associated with increased morbidity and mortality. Synthetic capability is reduced, one of the

TABLE 14.14 Sedative drugs used in ICU Agent

Properties

Comments

Benzodiazepines

Hypnosis and anxiolysis

Narcotics

Analgesia and hypnosis

Propofol

Intravenous induction agent

Ketamine

Intravenous induction agent

Isotlurane

Volatile anaesthetic agent

Inexpensive but may accumulate especially in the critically ill. This is of less importance when illness is prolonged Often successfully combined with benzodiazepines. Low cost advantageous. Alfentanil (newer synthetic narcotic) is used in renal failure Smooth, rapidly reversible, expensive agent. Not used in children but excellent for short-term sedation where rapid changes of consciousness are wanted receptor stimulant effect useful in bronchospasm although cost and dreaming/ psychosis limits its use Very effective and quickly reversible. Bronchodilating properties used in asthmatic emergencies

FIG. 14.14 Calorie requirements in normal individuals compared with those starved or critically ill and hypercatabolic

consequences of which is decreased immunocompetence and increased susceptibility to infection. Nutritional status requires careful assessment in the critically ill, and the history and length of illness should be noted. Precise measurement of nutritional status and requirements is difficult. Easily available measurements of synthetic capacity (albumin, total proteins) are rather variable and discriminate poorly between nutritional states (nutrition is discussed fully in Chapter 5). If possible, feeding should be continuous and controlled using a fine-bore nasogastric tube. This route is simple, economical and less likely to lead to infection. The incidence of gastric erosions is reduced in the enterally fed. However, absorption is often decreased by gastric stasis, or by the volume that can be given limited by diarrhoea. Several strategies may overcome this problem. Often, reduced gastric motility occurs when absorption in the small bowel is adequate. Gastric motility is increased by metoclopramide and erythromycin. If this fails, nasojejunal tubes can be inserted with a gastroscope, bypassing the stomach. If these efforts fail, a minority of patients require parenteral nutrition (p. 134), which requires the placement of a dedicated intravenous line. Patients often tolerate high fluid loads poorly, and feeding therefore requires concentrated solutions which may only be administered via a central vein. The constituents of intravenous feeding are ideal culture media and there is always a risk of infection. The volume of fluid, even when made up of highly calorific fats and concentrated glucose solutions, can precipitate volume overload. Careful attention to detail is required to provide appropriate quantities of trace elements and vitamins. During cellular respiration, fats and carbohydrates, when metabolized, produce differing amounts of CO2, to be excreted mainly by the lungs. High-carbohydrate feeds increase CO2 production and can contribute to CO2 retention and breathlessness in patients with marginal ventilatory function. 2

14

GENERAL CARE OF THE CRITICALLY ILL

Care on the ICU must be total as well as intensive. The patient who recovers from serious illness may have decades of life ahead and any residual handicap will be important. Attendant injuries, especially to the peripheral limbs, can be overlooked in the initial effort to resuscitate the patient, but may result in significant disability once the drama of intensive care is over. The maintenance of limb mobility and the prevention of contractures are important in bedbound supine patients. Skin and pressure area care is crucial. Poor peripheral perfusion commonly contributes to skin breakdown. Healing is enhanced when pressure is relieved, and this can be facilitated by air beds. Reflexes are lost during paralysis and sedation. The eyes and mouth are vulnerable. If the patient is ventilated the artificial airway must be kept clear and clean.

SEDATION There is no ideal sedative drug for patients on the ICU. The administration of sedation does not imply unconsciousness, and levels of sedation can be difficult to judge if the patient is receiving neuromuscular blocking drugs. For this reason local anaesthesia should always be used during painful procedures. Discussions around the bedside should be conducted as if the patient were aware (Table 14.14).

749

TABLE 14.15 Outcome Outcome from a single treatment technique for a defined disease • IPPV for severe asthma: >80% of patients discharged from hospital • IPPV for poisoning: >80% of patients discharged from hospital • IPPV for pneumonia or non-traumatic coma: <35% of patients discharged from hospital Outcome for organ system failure (e.g. cardiac, respiratory or renal failure) • Failure of one organ system for more than 1 day: mortality approximately 35% • Failure of two organ systems for more than 1 day: mortality approximately 60% • Failure of three or more organ systems for more than 3 days: mortality approximately 90%

Deep sedation is also harmful to the patient, and contributes to the depression of haemodynamics in the critically ill. It is easy for sedative drugs to accumulate, especially when renal or hepatic failure is present. The best management involves the use of a clinical scoring system which can be recorded frequently. If deep sedation is detected, the sedation can be stopped and the level of consciousness reassessed. Despite this approach, some patients, especially the elderly with multiple organ failure, can take many days to recover. The effect of sedation can be monitored by sophisticated processing of the EEG, and in the future it may be practical to use this technique to monitor the effect of sedative drugs in the critically ill. Patients are often mentally obtunded, or may experience hallucinations and paranoia, causing severe agitation. A nighttime sleeping regimen (or, better still, a daytime wakeful regimen) is of great help in the care of the longerterm patient on the ICU, especially during weaning from mechanical ventilation.

OUTCOME The outcome of patients admitted to the ICU can be assessed in different ways (Table 14.15). Survival is, in general, related to the severity of the presenting illness and the response to therapy. Patients with multiple organ failure have a particularly poor prognosis. Great care should be taken, however, when judging the prognosis of an individual patient, and the application of scoring systems to assess prognosis for a particular case has limited value. All staff on intensive care units have seen occasional patients survive against massive odds. The morale of patients and staff is central to the success

1

750

MCQ 15.1, MCQ 15.2

of an intensive care unit. The sustaining of morale, along with high standards of patient care, is a great challenge for the critical care team.

FURTHER READING Abraham E 1993 Sepsis: cellular and physiological mechanisms. New Horizons 1:1-159. A general article introducing the cellular biology of sepsis. The American-European Consensus Conference on ARDS. Definitions, mechanisms, relevant outcomes and clinical trial co-ordination. Bernard G R, Artigas A, Brigham K L et al. 1994 Am J Respir Crit Care Med 149(3 Pt l):818-824. Bone R C, Sibbald W J, Sprung C L 1992 The ACCP-SCCM consensus conference on sepsis and organ failure. Chest 101(6):1481-1483. Brochard L, Rauss A, Benito S, et al. 1994 Comparison of three methods of gradual withdrawal of ventilatory support during weaning from mechanical ventilation. Am J Respir Crit Care Med 150:896-903. A careful analysis of the impact of different modes of weaning in a group of patients over a period of 21 days. Esteban A, Frutos F, Tobin M J et al. 1995 A comparison of four methods of weaning patients from mechanical ventilation. Spanish Lung Failure Collaborative Group. New England Journal of Medicine 332(6):345-350. A short term assessment of different weaning techniques. Boyd O, Grounds R M, Bennett F D 1993 A randomised clinical trial of the effect of deliberate peri-operative increase of oxygen delivery on mortality in high risk surgical patients. Journal of the American Medical Association: 2699-2707. A prospective analysis of optimizing the cardiovascular system and its impact on peri-operative mortality. Hayes M, Timmins A C, Yau E H S, Palazzo M, Hinds C J, Watson J D 1994 Evaluation of systemic oxygen delivery in the treatment of critically ill patients. New England Journal of Medicine 330:1717-1722. A seminar paper describing the problems of applying optimization theory to patients on general Intensive Care Units. Lang C H, Abumrad N N 1995 Nutrition in the critically ill patient, vol 11, no. 3. W B Saunders, Philadelphia. A good review of the subject. Palmer R M J, Ferrige A G, Moncada S 1987 Nitric oxide release accounts for the biological activity of endolthelium-derived relaxing factor. Nature 327:524-526. The original paper describing nitric oxide and its vasodilating properties. A good review of the problems of systemic circulation. Rachon E C, Astiz M E 1993 Mechanisms and management of septic shock. Quote cares clinics vol 9:219-237. W B Saunders, Philadelphia. The Acute Respiratory Distress Syndrome Network 2000 Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 342: 1301-1308. Tobin M J 1988 State of the art: Respiration monitoring in the intensive care unit. American Review of Respiratory Disease 138:1625-1634. An excellent review of all aspects of respiratory monitoring. Yang K L, Tobin M J 1991 A prospective study of indexes predicting the outcome of trials of weaning from mechanical ventilation. New England Journal of Medicine 324:1445-1450. An assessment of weaning from mechanical ventilation.

15 Gastrointestinal Disease Richard G Long and Brian T Cooper

Introduction 751

Colorectal disease 797

Clinical features of gastrointestinal disease 751

Inflammatory bowel disease 804

Investigation of gastrointestinal disease 751 The mouth 760 The oesophagus 762 Gastrointestinal bleeding 772

Infective, microscopic, ischaemic and radiation colitides 815 Functional bowel disorders 817 Self-induced bowel disease 823

describe a variety of different symptoms related to eating. It is important that patients define precisely what they mean. Proctalgia fugax is a transient shooting pain in the rectum. Tenesmus, an uncomfortable desire to defecate, often when there is no stool in the rectum, may be associated with rectal inflammation and the irritable bowel syndrome. Abnormal signs in patients with gastrointestinal disease are summarized in Table 15.2. Diseases of the gut can result in malnutrition; failure of fluid intake or increased loss due to vomiting or diarrhoea results in dehydration. Peripheral oedema and ascites occur because of hypoproteinaemia due to protein loss, malabsorption or malnutrition. Vitamin D and calcium malabsorption result in hypocalcaemia and osteomalacia (Ch. 18). Finger clubbing can be associated with inflammatory bowel disease and coeliac disease, leukonychia (white nails) with hypoalbuminaemia, and koilonychia (spoonshaped nails) with chronic iron deficiency anaemia. Aphthous ulcers occur in patients with coeliac disease, and more indolent mouth ulcers in Crohn's disease. Angular cheilitis, stomatitis and glossitis occur in patients with iron and water-soluble vitamin deficiencies. Facial and oral telangiectasia are associated with telangiectasia lower in the gut, and these may bleed.

Diseases of the peritoneum 823 The pancreas 824

The stomach and duodenum 775 The small intestine 786

INTRODUCTION Despite important developments in our understanding of the causes of many important gastrointestinal diseases, coupled with major improvements in therapy, such diseases remain a major cause of morbidity and mortality throughout the world. In the UK gastrointestinal disorders are responsible for one in 10 of all deaths, one in six of all admissions to general hospitals, and two in 10 of all visits to a general practitioner. The gastrointestinal tract is the commonest site for cancer. Worldwide, gastrointestinal disease is a vast cause of ill health and it has been estimated that on any one day, 200000000 people suffer from diarrhoea.

CLINICAL FEATURES OF GASTROINTESTINAL DISEASE Typical symptoms of gastrointestinal disease are shown in Table 15.1. General symptoms are useful in assessing the state of health of many patients with gastrointestinal symptoms. Patients commonly complain of non-specific symptoms, but these are rarely associated with serious disease. Dyspepsia and indigestion are non-specific terms which

INVESTIGATION OF GASTROINTESTINAL DISEASE 1 HAEMATOLOGICAL AND BIOCHEMICAL ABNORMALITIES Iron and folate are absorbed from the upper small intestine, and malabsorption of either results in characteristic haematological changes (Table 15.3). In some patients with upper intestinal malabsorption (especially coeliac disease), a mixed folate and iron deficiency may be seen (dimorphic anaemia). Dietary iron deficiency can occur in patients with diets low in meat, and is particularly common in the elderly and in vegans. Blood loss can occur throughout the gut, with consequent iron deficiency and anaemia. Gastric surgery can cause iron deficiency by reduced acid secretion impairing absorption, by bypassing the upper small intestinal mucosa, and by intestinal hurry. Folate is present in large amounts in green vegetables and liver, but dietary deficiency may occur if the diet is inadequate. Vitamin B12 is bound to intrinsic factor in the stomach and is absorbed specifically in the terminal ileum (as are bile acids). Intrinsic factor deficiency can be congenital or secondary to pernicious anaemia or gastrectomy. Malabsorption of the B12-intrinsic factor complex is generally due to terminal ileal resection or disease (usually Crohn's disease). However, it can also occur (rarely) secondary to infestation with the fish tapeworm (Diphyllobothrium latum), to intestinal fistulae and blind loops, or with chronic pancreatitis (p. 829). The double-isotope Schilling test

751

CASE STUDY 15.1 CHRONIC IRON DEFICIENCY ANAEMIA IN A 48-YEAR-OLD NURSING SISTER History A 48-year-old white Caucasian nursing sister presented with a 3month history of increasing malaise and breathlessness. Her GP had checked her full blood count and found that her haemoglobin was 7.6, with a blood film showing red cell microcytosis and hypochromia. Haematinic estimation showed a low serum iron and ferritin, a high total iron binding capacity and normal folate and B12 levels. By the time that she was seen in the clinic, she had been taking ferrous sulphate 200 mg b.d. for 3 weeks and her symptoms had improved. Question 1. What are the most important questions to ask with regard to the predisposing cause? She had light periods for 3 days in every 28. She ate meat regularly and appeared to have a good oral iron intake. She was on no medication apart from occasional paracetamol. In particular, she did not use aspirin or non-steroidal anti-inflammatory drugs. She had no upper or lower gastrointestinal symptoms. There was no relevant family history of upper or lower gastrointestinal diseases causing bleeding or malabsorption. She drank five units of alcohol a week and was a nonsmoker. There was no relevant family history. Examination She was clinically anaemic. The chest, heart and abdomen were

752

1

Figs 15.1,15.2

4

Case 15.1

2 Fig. 15.3

unremarkable. Sigmoidoscopy showed no polyps or inflammation. Investigations Upper gastrointestinal endoscopy showed the oesophagus, stomach and first part of the duodenum to be normal. The circular folds in the second part of the duodenum were reduced in number and height, which is a marker of coeliac disease. 1 She was also IgA endomysial and gliadin antibody positive. Question 2. What histological changes would you expect in the second part of the duodenum in a patient with coeliac disease? Biopsies of the second part of the duodenum showed total villous atrophy with increased intraepithelial lymphocytes. 2

Question 3. Does she need further investigation? She should have imaging of her large bowel, as right-sided colonic tumours are relatively common in patients with iron deficiency anaemia. The large bowel could be visualized by colonoscopy, which has the advantage of a direct view, the ability to take histology and good views of mucosal lesions such as ulcerative colitis or angiodysplasia, or by a barium enema which has a high chance of obtaining good images of the right colon. She

3 Fig. 15.4

had a colonoscopy, which showed a large carcinoma of the ascending colon. 3 Treatment The patient was started on a glutenfree diet. She had a right hemicolectomy for her carcinoma, which was a Duke's stage B. She made a good recovery from the operation. She was treated with adjuvant chemotherapy in a randomized placebo-controlled clinical trial. Discussion Iron deficiency anaemia is a very important clinical finding which, unless there is severe menorrhagia, should always lead to investigation of the gastrointestinal tract. In about 10% of cases lesions are found in both the upper and the lower gastrointestinal tracts. In general, most patients consequently require imaging of both ends of the gut. It is unusual for patients presenting with iron deficiency anaemia to have lesions between the second part of the duodenum and the distal terminal ileum. Small bowel radiology and enteroscopy therefore have only a small role. This patient should have a longterm cure rate of about 50% from her Duke's B tumour. Coeliac disease is associated with an increased risk of certain rare gastrointestinal malignancies. It is generally accepted that this risk returns to that of the normal population after 5 years on a gluten-free diet. Four years after the initial presentation, the patient remains well.

(p. 1210) is used to determine whether B12 deficiency is due to lack of intrinsic factor or to malabsorption of the intrinsic factor-Bi2 complex. Vegans have a diet very low in B12 and thus may become deficient. Hyposplenism is a feature of some bowel disorders, particularly coeliac disease. The typical finding on a blood film is of Howell-Jolly bodies (red cells containing small

TABLE 15.1 Symptoms of gut disease Small intestinal symptoms Diarrhoea Steatorrhoea Symptoms of anaemia Abdominal colic Abdominal distension

General symptoms Anorexia Weight loss Nausea and vomiting Malaise and lethargy Non-specific symptoms Flatulence Abdominal bloating Poorly or variably localized pain Oesophageal symptoms Dysphagia Heartburn Regurgitation Retrosternal pain Painful swallowing (odynophagia)

4

Large intestinal symptoms Symptoms of anaemia Diarrhoea Constipation Abdominal colic Blood and mucus per rectum Tenesmus Proctalgia Incontinence

Gastroduodenal symptoms Upper abdominal pain/discomfort Early satiety Haematemesis and melaena

nuclear fragments). It also occurs in some patients with inflammatory bowel disease. Neutrophil leukocytosis and raised acute-phase reactants (e.g. ESR, serum orosomucoids or C-reactive protein) are useful non-specific markers of intra-abdominal inflammation, particularly inflammatory bowel disease, and sepsis. A low serum albumin is commonly seen in patients with bowel disease (Table 15.4) and reflects protein leakage into the gut, decreased dietary intake, reduced absorption and, rarely, decreased synthesis. Albumin has a plasma half-life of about 23 days. Proteins such as transferrin and prealbumin have a shorter half-life and plasma levels consequently fall earlier. Electrolytes may be deranged. In patients with gastric outflow obstruction (pyloric stenosis), a hypochloraemic hypokalaemic alkalosis may occur. Diarrhoea may result in dehydration and electrolyte loss, particularly potassium. Postgastrectomy patients and patients with severe small bowel disease causing malabsorption may have a low serum calcium and osteomalacia. These patients develop secondary hyperparathyroidism, hyperphosphaturia and hence hypophosphataemia. Patients, especially postmenopausal females, with small bowel disease and/or costicosteroid therapy can develop osteoporosis.

ENDOSCOPY The development in the 1970s and 1980s of flexible fibreoptic and video endoscopes revolutionized the gastro-

15

TABLE 15.2 Signs of gut disease General Malnutrition Dehydration Anaemia Oedema Bone tenderness Chvostek's sign Trousseau's sign Finger clubbing Leukonychia Koilonychia Oral Aphthous ulcers Other oral ulcers Angular cheilitis Stomatitis Glossitis Telangiectasia

Gastroduodenal Epigastric tenderness Epigastric mass Succussion splash | 1 Hypocalcaemia

Small intestinal Diarrhoea steatorrhoea Mass Abdominal distension Intestinal obstruction Colonic Mass Obstruction Distension Anorectal Anal fissures, tags, haemorrhoids, sinuses, fistulae Rectal examination: mass, blood, mucopus Sigmoidoscopy: proctitis, ulcers, tumours, melanosis, haemorrhoids

enterologist's view of the gut. Endoscopy is generally superior to radiology as an investigative technique, as it provides a direct mucosal view, enables biopsies to be taken for histological and biochemical analysis, and permits therapeutic manoeuvres (Table 15.5). Endoscopes are used in the upper gut to examine the pharynx, oesophagus, stomach and duodenum. Oesophageal strictures can be dilated using balloons or bougies, and metal or plastic stents can be put through strictures (usually malignant). Bleeding lesions can be cauterized by electric current or lasers. The ampulla of Vater can be cannulated and contrast medium injected to reveal the anatomy of the pancreatic ducts and biliary tree (endoscopic retrograde cholangiopancreatography, ERCP). If there are stones in the biliary tree, a sphincterotomy can be performed to allow them to fall out or be extracted by baskets or balloons. Malignant strictures in the biliary tree can be bypassed by Teflon or metal stents passed endoscopically or percutaneously through the liver. The anorectum can be examined by a rigid proctoscope, which is particularly useful for demonstrating haemorrhoids, and the rectum and lower sigmoid colon by a 25 cm rigid sigmoidoscope. A flexible fibreoptic sigmoidoscope can be used for the distal colon and rectum, and may be advanced to the splenic flexure (Fig. 15.1). Colonoscopes can be passed to the caecum and thence, via the ileocaecal valve, into the terminal ileum. Suspicious lesions can thus be biopsied, polypoid lesions snared by excision diathermy, and abnormal colonic blood vessels (angiodysplasia)

753

TABLE 15.3 Haematological abnormalities in gut disease Abnormality

Laboratory changes

Causes

Iron deficiency anaemia

Blood film: hypochromic microcytic red cells Low mean cell volume (MCV) Low serum iron and high total iron-binding capacity

Blood loss Malabsorption

Dietary

Postgastrectomy Upper small bowel disease Folate deficiency anaemia

B12 deficiency anaemia

Blood film: macrocytic red cells High MCV

Dietary Upper small bowel disease

Low red cell folate

Drugs (e.g. sulphasalazine) Pernicious anaemia

Blood film: macrocytic red cells High MCV

Terminal ileal disease Postgastrectomy Chronic pancreatic disease

Low serum B12

Bacterial overgrowth of the small intestine

Fish tapeworm (Diphyllobothrium latum) Hyposplenism

Coeliac disease Tropical sprue Crohn's disease and ulcerative colitis

Howell-Jolly bodies Giant platelets

Neutrophil leukocytosis and high ESR

cauterized. The small bowel mucosa from the end of the second part of the duodenum to the terminal ileum cannot normally be viewed endoscopically, except by the littleused enteroscope; at laparotomy, however, it is possible to make an enterotomy and inspect the whole small bowel endoscopically.

Chronic inflammation, e.g. inflammatory bowel disease, abscess etc.

TABLE 15.4 Gastrointestinal causes of a low serum albumin Increased loss (protein-losing enteropathy) Stomach Menetriere's disease (giant rugae) Multiple gastric ulcers Gastric tumours

BIOPSY Diffuse mucosal disorders and focal lesions can be diagnosed by histological examination of tissue biopsies. Biopsies of the stomach, duodenum and colon are taken via the appropriate endoscope. A proximal small bowel biopsy is essential in suspected small bowel disease (e.g. coeliac disease). In the past proximal small bowel biopsies were taken from the jejunum with a swallowed capsule, but now such biopsies are usually taken from the distal duodenum via a fibreoptic endoscope. All gastrointestinal biopsies should be correctly orientated on card. This is especially important with small bowel biopsies so that the villous pattern can be assessed with a hand lens or dissecting microscope before being processed for histological examination. Some disorders (e.g. Crohn's disease) show a

Small intestine Coeliac disease Crohn's disease Tropical sprue Lymphoma

Decreased intake Anorexia Protein-deficient diet Decreased synthesis Associated liver disease (as can occur in inflammatory bowel disease) Decreased absorption Coeliac disease Crohn's disease Tropical malabsorption .

Lymphangiectasia Large intestine Crohn's disease Ulcerative colitis Colonic tumours

predilection for the distal ileum, and this can be biopsied via a colonoscope passed through the ileocaecal valve.

BARIUM STUDIES 1

754

MCQ15.3

If a patient has difficulty with swallowing (dysphagia) a barium swallow is often recommended before

15

TABLE 15.5 Endoscopic procedures Procedure

Range of examination

Therapeutic procedure

Oesophagogastroduodenoscopy (upper Gl endoscopy)

Mouth to third part of duodenum

Oesophageal dilatation and stent insertion Oesophageal varices sclerosis or banding Tumour destruction by laser or alcohol injection Injection or laser of bleeding sites Balloon dilatation of pyloric stenosis Percutaneous endoscopic gastrostomy (PEG)

Endoscopic retrograde cholangiopancreatography (ERCP)

Ampulla and second part of duodenum (visual) Biliary tree and pancreatic duct (radiological)

Sphincterotomy Stone extraction Stent insertion Stricture dilatation Nasobiliary drainage

Enteroscopy Colonoscopy

Duodenum to ileum Anus to caecum/terminal ileum

Flexible sigmoidoscopy

Anus to descending colon (sometimes to splenic flexure)

endoscopy to outline the anatomy and to reduce the risk of endoscopic perforation. However, many endoscopists would not agree and use endoscopy as their first investigation for dysphagia. A barium swallow on video is very helpful in elucidating motor problems of the oesophagus, such as bulbar palsy, spasm and achalasia. In the investigation of gastroduodenal symptoms upper gastrointestinal endoscopy is preferred to a barium meal, as it is more accurate and allows biopsy and cytology specimens to be obtained. The small intestine can be examined by a small bowel meal, in which barium is swallowed and its progress followed to the terminal ileum. For better views of the distal small bowel and focal lesions a small bowel enema, in which contrast medium is placed directly into the upper jejunum via a nasal tube, is performed (Fig. 15.2). Double-contrast enemas with air and barium are useful in demonstrating colonic lesions such as carcinomas, polyps and inflammatory bowel disease. Barium enema should be delayed for at least 1 week after rectal biopsy to avoid the risk of perforation. Colonoscopy should be performed if there is doubt about the barium appearances, if angiodysplasia is suspected and if lesions requiring polypectomy are present.

Polypectomy Stricture dilatation Tumour destruction by laser Injection or laser treatment of bleeding sites As for colonoscopy

liver, biliary tree, spleen, pancreas, pelvis and retroperitoneal structures. They are helpful in demonstrating tumours, abscesses and stones, but are of little help in outlining the bowel. Isotopic techniques can be used to investigate gut disease. Technetium-99m (99mTc) sulphur colloid can be used to monitor the liquid and solid phases of gastric emptying (in general, liquids empty faster) and to measure the time taken by nutrients to pass through the small bowel (small bowel transit). Red cells can be radiolabelled with chromium-51 (51Cr) to demonstrate active gastrointestinal blood loss, especially when this cannot be demonstrated by conventional endoscopic and radiological means. White cells can be radiolabelled to demonstrate abscesses and areas of active inflammatory bowel disease. Meckel's diverticula often contain gastric epithelium and can be localized by radiolabelled pertechnetate, which is secreted by the parietal cells. Arteriography outlines the coeliac axis and superior and inferior mesenteric arteries, and can define stenosis or occlusion of these arteries in patients with pain after eating (mesenteric angina) and sites of active bleeding (e.g. tumours and angiodysplasia).

GUT REGULATORY PEPTIDES 1 IMAGING TECHNIQUES Ultrasound, computed tomography (CT) and magnetic resonance imaging (MRI) are all useful in imaging the

Three classes of regulatory peptides have been defined: • Hormones, such as secretin and gastrin, are released as a result of food in the gut and have effects at distant sites.

755

FIG. 15.2 A normal small bowel enema This shows the jejunum and ileum in the upper and lower parts of the picture, respectively.

RECENT ADVANCES IN ENDOSCOPY, IMAGING AND ENDOSCOPIC THERAPY

FIG. 15.1 Endoscopic instruments A A rigid proctoscope and Sigmoidoscope with their respective trocars removed are seen centrally. They are surrounded by a diagram of a typical fibreoptic gastroscope. B Diagram of the distal end of a typical video or fibreoptic gastroscope, Light is provided by the guides and the image is transmitted via the objective lens to the operator. The air and water outlets allow insufflation of the gut with air and cleansing of the lens with water. Gastric contents and secretions can be removed via the suction channel and biopsy forceps, and cytology brushes can also be passed by this route. C Diagram of the movements which the tip of a typical fibreoptic endoscope can achieve.

756

Videoendoscopy has now largely replaced the older optical instruments. Videoendoscopy is more comfortable for the endoscopist and allows assistants and trainees to see the same image as the endoscopist. A permanent record (polaroid photograph or videofilm) is much easier to obtain. Ultrasound scanning via special probes passed through the endoscope is a new and rapidly developing technique. It seems to have significant advantages over external ultrasound, CT and MRI in the diagnosis of pancreatobiliary disorders and in staging and assessing the operability of GI tumours, especially of the oesophagus and rectosigmoid colon. The development of spiral (helical) CT scanning raises the prospect of such superior and detailed images of the biliary tree and pancreas that diagnostic ERCP will be unnecessary. Spiral CT scanning of the colon (colography, virtual colonoscopy) gives images approaching those of endoscopic colonoscopy. The endoscopic stenting of oesophageal and biliary malignant strictures as part of palliative therapy has been greatly simplified and made much less dangerous by the use of expanding metal stents. Paracrine regulatory peptides, discovered in the gut mucosa, are released locally and have profound effects on neighbouring cells. Neurotransmitters. Immunocytochemistry has demonstrated some peptides in the nerves of the gut,

e.g. somatostatin, vasoactive intestinal polypeptide (VIP) and substance P, which are thought to act as non-adrenergic non-cholinergic neurotransmitters (neuropeptides). The site of origin, mechanism of action and main physiological effects of the best-documented regulatory peptides are summarized in Table 15.6. Clinically, the most important aspects of these hormones are their use in diagnostic tests (e.g. the pentagastrin test for gastric acid secretion and the secretin-cholecystokinin test for pancreatic function) and their role in tumour syndromes (p. 833). Measurement of the peptides in plasma or tissue can be performed by radioimmunoassay. The following abnormal results may have a role in diagnosis: • In functioning tumours, plasma levels of peptides are high (e.g. gastrin in gastrinoma and vasoactive intestinal polypeptide in VIPoma). • In untreated coeliac disease, which is a proximal enteropathy, meal-stimulated plasma levels of the upper small intestinal hormones (secretin, cholecystokinin and gastric inhibitory polypeptide) are low. The terminal ileal hormones (neurotensin and enteroglucagon) are increased. • In chronic pancreatitis with steatorrhoea, levels of mealstimulated plasma pancreatic polypeptide are reduced. • In neuropathic diseases of the gut, such as Hirschsprung's disease and Chagas' disease, tissue neuropeptide levels are reduced.

TESTS OF FUNCTION

pylorus, the ileocaecal valve and the rectum. Control of gut motility is thought to be mainly neural but hormones, in particular motilin, may also play a role. The stomach plays a major role in grinding food to small particles and in controlling the exit of contents into the duodenum, which may take up to 5 hours after a large meal. Oesophageal motility can be investigated by a barium swallow on video, by static or 24-hour ambulatory manometric pressure studies, and by radionuclide transit studies. A pH probe in the lower oesophagus can assess gastrooesophageal reflux. Gastric emptying can be investigated by labelling the solid or the liquid phase with an isotope. Small intestinal motility can be assessed by the time taken for isotopes to reach the caecum, by demonstrating a peak of breath hydrogen excretion liberated from bacterial breakdown of a carbohydrate load when a test meal reaches the distal ileum and caecum, or by radiotelemetry, where a radiotransmitter capsule is swallowed and signals are received, allowing direct measurement of gut motility patterns. Whole-gut motility can be assessed by measuring mouth-to-anus transit time using swallowed radiolucent beads or non-absorbed radionucleotides. Disruption of anal function causing incontinence can be assessed by ultrasound. Obstructive defecation, where patients either contract rather than relax their anal muscles at defecation, or have other problems with rectal evacuation, can be assessed by putting barium paste in the rectum and then monitoring the speed of evacuation and the appropriateness of muscle action (proctography). A variety of clinical syndromes give rise to abnormal motility (Table 15.7). This usually causes a delay in transit, but gastric and other surgery can result in excessively rapid transit.

15

Motility Normal gut motility promotes the passage of food and food residue along the lumen at a rate that allows the appropriate physiological processes to take place. Well-regulated 'brakes' occur at the gastro-oesophageal junction, the

Gastrointestinal secretion The salivary glands, stomach, pancreas, biliary tree and intestine all produce a large volume of secretions; these change the pH and produce enzymes that allow

TABLE 15.6 Origin and action of some gastrointestinal regulatory peptides Regulatory peptide

Site of origin

Mechanism of action

Main physiological effect

Gastrin Secretin Cholecystokinin

Gastric antrum Duodenum Duodenum

Hormone Hormone Hormone

Motilin Gastric inhibitory polypeptide (GIP) Enteroglucagon Neurotensin Pancreatic polypeptide Somatostatin

Upper small bowel Upper small bowel Terminal ileum and colon Terminal ileum Pancreas Pancreas and all the gut

Vasoactive intestinal polypeptide (VIP)

Pancreas and all the gut

Hormone Hormone Hormone Hormone Hormone Hormone Paracrine Neu retransmitter Neurotransmitter

Stimulates gastric acid secretion Stimulates pancreatic bicarbonate secretion Stimulates pancreatic enzyme secretion and gallbladder contraction Increases gastric emptying and gut motility Stimulates insulin secretion Delays gastric emptying; trophic effects on gut mucosa Delays gastric emptying Decreases pancreatic and biliary secretion Reduces hormonal and intestinal secretions Increases intestinal secretion

757

TABLE 15.7 Disorders of gut motility Oesophagus

Stomach

Small bowel

Colon

Anatomical Tumours Inflammatory stricture (e.g. acid and acid/alkali burns) I

Tumours Pyloric stenosis Postgastrectomy

Tumours Inflammatory stricture (e.g. Crohn's disease and tuberculosis) Postsurgical resection

Tumours Inflammatory strictures (e.g. Crohn's disease) and ischaemic colon Postsurgical resection

Postvagotomy Diabetes mellitus

Diabetes mellitus Systemic sclerosis 4

Hirschsprung's and Chagas' diseases Diabetes mellitus

Non-ulcer dyspepsia Vomiting Aerophagy

Irritable bowel syndrome

Irritable bowel syndrome Slow-transit constipation

Neuropathic Achalasia Systemic sclerosis Functional Globus hystericus

increased motility

decreased motility

the digestion of food and thus absorption of nutrients (Fig. 15.3). The secretions are controlled by vagal nerves, including non-cholinergic non-adrenergic fibres, by local paracrine substances and by classic hormones. Modes of investigation and abnormalities of secretion are shown in Table 15.8. In general, meal tests are more physiological than hormonal tests, e.g. the pentagastrin and secretin-cholecystokinin tests, where pharmacological doses of hormones are given intravenously. The triplelumen tube can be passed perorally into the intestine and various solutions containing a non-absorbable marker may then be perfused. By sampling at the end of the tube, the amount of secretion or absorption can be assessed.

Gastric acid studies Gastric acid secretion can be assessed by collecting gastric juice via a nasogastric tube in the fasting state, measuring basal acid output and, after stimulation with a gastrin analogue, pentagastrin, assessing maximal acid output. Very high levels of basal acid outputs are seen in patients with gastrin-secreting tumours (gastrinomas), and high maximal acid outputs often occur in patients with duodenal ulcers. Twenty-four-hour acid secretion can also be monitored, to assess the effects of drugs that reduce acid secretion. Insulin hypoglycaemia stimulates gastric acid secretion via the vagus and can therefore be used postoperatively to assess the efficacy of surgical vagotomy. pH probes can also be placed in the oesophagus to monitor acid reflux, and in the stomach to monitor the effects of treatment.

Intestinal digestion and absorption 758

The stomach secretes acid and pepsin and, by physical churning and mixing, breaks down food into proportions that allow pancreatic and small intestinal brush border

FIG. 15.3 Typical daily fluid intake, secretions and absorption in a healthy person

enzymes to digest it ready for absorption. Fat and fatsoluble vitamins require bile acids, lipase and micelle formation to permit absorption. Proteins require gastric pepsin, pancreatic trypsin and brush border peptidases to break down proteins to absorbable single amino acids or 2-A amino acid peptides. Carbohydrates require pancreatic amylase, and then brush border disaccharidases, to form monosaccharides which can be absorbed (Fig. 15.4). Calcium, iron and folic acid are absorbed in the duodenum and upper jejunum, whereas vitamin Bi2 and bile acids are

15

TABLE 15.8 Investigation and examples of important hyper* and hyposecretion syndromes Type of secretion

Modes of investigation

Examples of increased secretion

Examples of reduced secretion

Salivary glands

Volume, pH and amylase levels, with or without local oral stimulant

Sialorrhoea

Sjogren's disease

Stomach

Pentagastrin and meal-stimulated acid output

Duodenal ulcer Gastrinoma

Carcinoma of stomach Atrophic gastritis

Pancreas

Lundh test meal and secretion-cholecystokinin stimulation

Small intestine

Triple lumen tube and assessment of secretion/ absorption over given length of intestine

VIP-secreting tumours Cholera Phenolphthalein ingestion

Large intestine

Triple-lumen tube

VIP-secreting tumours Dihydroxy bile acids Long-chain fatty acids

Chronic pancreatitis with exocrine insufficiency

Lactase

FIG. 15.4 Intestinal digestion and absorption A Enzyme activity allowing the conversion of major unabsorbable disaccharides to monosaccharides, which the enterocyte then can absorb. B Glucose is transported by an active carrier with sodium at the brush border membrane. The gradient of sodium across the enterocyte is maintained by the Na+-K+-ATPase at the basolateral membrane, which allows sodium to pass out of the cell in exchange for potassium. Glucose leaves the cell at the basolateral membrane by a Na+-independent carrier.

absorbed in the terminal ileum; all other nutrients are absorbed throughout the small intestine. Small intestinal absorption requires active or passive transport across the brush border membrane, transport through the cell (often in Golgi or mitochondria), and then extrusion through the cell's basolateral membrane. The most important test of small bowel function is the measurement of fat absorption, because most diffuse disorders of the small intestine can cause fat malabsorption. This can be performed by measuring faecal fat excretion or the 14C-triolein breath test, whereby fat digestion is measured by 14CO2 excretion. Specific tests of carbohydrate digestion and absorption are available; brush border disaccharidase activity (e.g. lactase) can be measured in small bowel biopsies, and digestion and absorption of specific disaccharides can be assessed by measuring blood sugar levels after oral ingestion of the test sugar or by specific breath hydrogen tests. Protein absorption is assessed rarely, although faecal nitrogen excretion and faecal protein loss can be measured. Terminal ileal function is assessed by B12 absorption (e.g. Schilling) tests or bile salt absorption tests (e.g. 75Selabelled homocholic acid conjugated with taurine (75SeHCAT)). The integrity of the small bowel mucosa can be assessed by measuring the permeability of the mucosa to small inert hydrophilic non-metabolized molecules of varying molecular sizes (e.g. L-rhamnose, mannitol, lactulose, cellobiose, EDTA). Exocrine pancreatic function can be assessed by faecal fat measurement and specific pancreatic function tests (p. 824). The colon is primarily involved in the absorption of water and electrolytes. Normally, up to 1500mL of fluid passes through the ileocaecal valve daily, and this is concentrated to a 100 g formed stool. Specific colonic function tests are of research interest only.

759

THE MOUTH NORMAL AND ABNORMAL FUNCTION The importance of the mouth, tongue, teeth and salivary gland secretions in starting the digestive process is often overlooked. Mastication and lubrication of the food bolus is desirable to facilitate its painless progression. Once swallowing has been voluntarily initiated, the rest of the act is automatic. Coordination of the soft palate rising to occlude the nasopharynx, the epiglottis tilting backwards to occlude the larynx, and the temporary cessation of respiration, is under complex neurological control, principally by the brainstem. Afferent pathways from taste and olfactory receptors initiate the cephalic phase of vagally mediated gastric secretion. Saliva contains two digestive enzymes, an amylase and a lipase. Lingual lipase may be important in patients with chronic pancreatic insufficiency. Xerostomia (dry mouth) can be severe enough to cause problems in initiating swallowing. Lack of saliva may predispose to ascending bacterial infections of the salivary glands (sialadenitis), especially the parotids (parotitis). Saliva also contributes to the prevention of dental caries, the pathogenesis of which involves the facultative anaerobe Streptococcus mutans, together with carbohydrate (mainly sucrose) and the formation of dental plaque. Both the edentulous and patients with ill-fitting or painful dentures may limit their food intake, often unconsciously, and thus lose weight. Other painful lesions of the mouth or tongue may have a similar effect but, being more obvious, come readily to medical attention.

DISEASES OF THE MOUTH Disease of the mouth may be important because: • The disease, although confined to the mouth, may give rise to unpleasant symptoms and disordered function (Table 15.9). • It may represent mucous membrane involvement of a generalized dermatological condition (Table 15.10). • Involvement of the mouth may be part of a systemic disease (Table 15.11). • The condition may be iatrogenic (Table 15.12).

ORAL ULCERATION 1 Recurrent aphthous ulcers There are many possible causes of oral ulceration (Tables 15.9-15.12) but the commonest is idiopathic aphthous 1

760

MCQ 15.4

TABLE 15.9 Diseases confined to the mouth Ulceration

Plaques

Herpes simplex (HSV1) Recurrent herpes labial is Primary gingivostomatitis

Candida Leukoplakia Carcinoma

Acute ulcerative gingivitis (Vincent's angina, caused by Fusobacterium fusiformis and Borrelia vincenti) Cancrum oris Coxsackie A Herpangina Hand, foot and mouth disease Recurrent aphthous ulcers Minor Major Denture gingivitis

TABLE 15.10 Dermatological conditions involving the mouth Ulceration (burst bullae)

Plaques

Pemphigus vulgaris Pemphigoid Erythema multiforme Lichen planus

Lichen planus Psoriasis

ulceration. This occurs in up to 20% of patients seen in hospital and 10% in general practice. It often begins in late childhood or the early teens and frequently recurs. In about 80% of cases the condition (though still unpleasant) is regarded as minor, with up to five ulcers, less than 1 cm in diameter, affecting the lips, cheeks and tongue and healing within a week. Major ulcers are larger, often more numerous, and may involve the pharynx and palate. Cervical lymph node enlargement may occur and healing may be delayed for several weeks, leaving scarring. The aetiology of aphthous ulcers is unknown. The preponderance of the disease in females, its onset at the menarche, relationship to menstruation and improvement in pregnancy have led to the postulation of hormonal factors. A proportion of patients with coeliac disease present in this way and appropriate tests, including small bowel biopsy, are indicated. Treatment is with tetracycline mouthwash to prevent secondary infection, local corticosteroids and, in severe cases, systemic corticosteroids.

Herpes simplex The commonest manifestation of herpes simplex (HSV 1)

TABLE 15.11 Mouth involvement in systemic disorders Mouth involvement

Systemic disorder

Mucocutaneous pigmentation of lips

Peutz-Jeghers syndrome

Telangiectasia (lips)

Scleroderma (CRST syndrome) Hereditary haemorrhagic telangiectasia (Osler-Weber-Rendu syndrome)

Buccal pigmentation

Addison's disease ACTH-producing tumour

Macroglossia

Hypothyroidism Acromegaly Amyloid

Glossitis

Iron deficiency (also angular cheilitis) Folate and vitamin B12 deficiency

Gingival hypertrophy

Acute leukaemia Amyloid

Gingivitis (with or without bleeding)

Acute leukaemia Scurvy Thrombocytopenia Agranulocytosis

Oral ulceration

Behget's disease Reiter's syndrome Crohn's disease Coeliac disease Wegener's granulomatosis Systemic lupus erythematosus Tuberculosis Syphilis Glandular fever Agranulocytosis

TABLE 15.12 latrogenic disorders of the mouth latrogenic disorder

Cause

Xerostomia

Atropine-like drugs Tricyclic antidepressants

Gingival hyperplasia

Phenytoin, primidone, phenobarbitone, ciclosporin

Gingival pigmentation

Bismuth, mercury, lead, arsenic, silver

Ulceration

Sulphonamides, barbiturates, penicillin (cause erythema multiforme) Gold Penicillamine Cytotoxics Drugs causing agranulocytosis, e.g. carbimazole, sulphasalazine

is of recurrent herpes labialis (cold sore). The virus lies dormant in the trigeminal ganglion. Antiviral therapy with local idoxuridine or acycloguanosine (aciclovir) is probably only necessary in the immunocompromised patient. A rarer, more severe and usually non-recurrent infection is primary herpetic gingivostomatitis. It occurs primarily in childhood and the sore mouth and gums, with vesicles preceding ulceration, is accompanied by fever and regional lymphadenitis. Treatment is analgesia and antibiotics to prevent secondary bacterial infection.

15

Acute ulcerative gingivitis Acute ulcerative gingivitis (Vincent's angina) is thought to be due to bacterial infection; Fusobacterium fusiformis and Borrelia vincenti are the currently incriminated organisms. This disease affects young adults and the immunosuppressed, and is again accompanied by fever and lymphadenitis. The painful gums may bleed. Vincent's angina may be distinguished from herpes by the foul fetor and the absence of vesicles. Penicillin or metronidazole are usually effective treatments.

Crohn's disease Oral lesions are relatively common, particularly if there is colonic involvement (p. 808). The lesions include aphthous ulceration and cobblestoned or fissured epithelium. They can be the sole presenting feature of the disease and may be painful. Biopsies of the lesions may show typical noncaseating granulomata.

Other causes of ulceration Oral ulceration may predominate in Behcet's disease, Reiter's syndrome and erythema multiforme. In all three conditions ocular and genital lesions may be present. Although many cases of erythema multiforme have no apparent cause, some are induced by herpes simplex and drugs such as sulphonamides, barbiturates and penicillin. Bullous disorders of mucous membranes result in their rupture to form ulcers in the mouth. As at other sites, differentiating the intraepithelial bullae of pemphigus from the subepithelial bullae of pemphigoid may be difficult (p. 410). Lichen planus may also present with erosion of vesicles causing ulceration, and in a significant number of patients the mouth may be the only site of the disease (p. 413).

PLAQUE LESIONS Candida Candidiasis usually arises as a result of a defect in immunity, as the causative fungus is a common commensal. Candidiasis may complicate HIV infection and is common in poorly controlled diabetics and asthmatic patients on corticosteroid inhalers (see Ch. 9, p. 332 and Ch. 10, p. 391).

761

Leukoplakia Leukoplakia is a serious condition because of its propensity for malignant transformation in about 5% of cases. The white patches can occur anywhere within the oral cavity. Biopsy, which must be repeated regularly to detect atypia or early cancer, confirms the diagnosis by showing epithelial keratosis and hyperplasia. The aetiology is multifactorial, but idiopathic cases do occur. There is no specific treatment other than excision and avoiding the presumed causative factors. These include smoking, local friction and chronic infection, e.g. chronic candidiasis or tertiary syphilis; the latter has a worse prognosis regarding malignant change. Leukoplakia is an important manifestation of HIV infection (p. 444).

Carcinoma Squamous carcinoma of the oral cavity may affect the lips, tongue or any other part of the mouth, and may present as a nodular or ulcerating lesion. As with leukoplakia, its incidence appears to be declining, emphasizing the important association between the two conditions. Pipe and cigar smoking is strongly associated with oral cavity carcinoma, as is the chewing of tobacco or betel nut.

OTHER LESIONS Other disorders of the mouth are listed in Tables 15.9-15.12. Minor abnormalities of the tongue may, because they are easily seen by the patient, give rise to considerable anxiety. Georgraphic tongue (benign migratory glossitis) is caused by shifting areas of desquamation of short duration in any one location. It is invariably symptomless and the patient should be reassured of this. Fissure tongue may be associated with symptomatic glossitis but, if symptomless, the patient can be reassured. Black hairy tongue is asymptomatic but may cause anxiety. Tripotassium dicitrate bismuthate (Denol) in liquid form is an obvious cause and the disorder may also follow antibiotics, but the aetiology is usually unknown. Reassurance is required. Brushing the tongue surface may help.

THE OESOPHAGUS Normal swallowing involves coordinated action of the oropharynx, relaxation of the cricopharyngeus, peristalsis down the oesophageal body (upper one-third striated muscle, lower two-thirds non-striated muscle) and relaxation of the lower oesophageal sphincter.

1

762

Case 15.2

CLINICAL FEATURES OF OESOPHAGEAL PATHOLOGY Dysphagia 1

Difficulty with swallowing (dysphagia, or 'food sticking') is a most important symptom. Some of the causes are shown in Table 15.13. When there are obstructive lesions, such as carcinoma, dysphagia is initially for large boluses of food but then progresses to less solid food and finally to liquids ('total dysphagia'). Most of the causes can be elucidated by a good history, examination, barium swallow and endoscopy. Patients with central and peripheral nervous diseases and primary muscle disorders usually have other neuromuscular symptoms and signs. Hysterical dysphagia is a diagnosis of exclusion and usually occurs in psychiatrically disturbed patients.

Other symptoms Painful swallowing and retrosternal pain are important symptoms. On occasions it may be difficult to differentiate oesophageal and cardiac pain, and there is evidence that gastro-oesophageal reflux may lower the threshold for angina. Regurgitation of gastric juice into the lower oesophagus and the back of the mouth may cause heartburn and is due to an incompetent lower oesophageal sphincter; this is commonly precipitated by raising the intra-abdominal pressure (lifting, pregnancy). If food accumulates in the oesophagus nausea, vomiting and weight loss may occur. The oesophagus is also a site of bleeding from varices, ulcers and inflammatory lesions (oesophagitis).

Signs Abnormal physical signs are unusual in patients with oesophageal disease, but epigastric tenderness may indicate oesophagitis, peptic ulceration or a carcinoma. Neurological examination of palatal movement and the gag reflex are important when dysphagia could be due to a

TABLE 15.13 Causes of dysphagia • Local oropharyngeal problems, e.g. infections, tumours • Central nervous system disorders affecting oropharyngeal function, e.g. brainstem lesions associated with multiple sclerosis, cerebrovascular disease, Parkinson's disease, Huntington's chorea and motor neuron disease • Malignant and benign oesophageal strictures • Other local oesophageal problems, e.g. reflux oesophagitis, infections, web, diverticulum • Disorders affecting oesophageal unstriated muscle, e.g. achalasia, chronic alcoholism, diabetes mellitus • Generalized striated muscle diseases, e.g. myasthenia gravis, dystrophia myotonica • Extrinsic compression (e.g. mediastinal tumour) • Psychiatric, i.e. globus hystericus

neurological cause. An epigastric mass or palpable supraclavicular lymph nodes suggests a neoplasm. Decreased oesophageal clearance of solids and fluids (especially in patients with achalasia and carcinoma) may cause pulmonary aspiration.

INVESTIGATION OF OESOPHAGEAL PATHOLOGY A barium swallow outlines the oropharyngeal and oesophageal anatomy, and a video recording can be taken to look for disordered motility. A chest X-ray helps to assess oesophageal dilatation. If a patient has dysphagia a barium swallow can be carried out before endoscopy; this is because symptoms alone cannot always localize the site of the lesion, and if the endoscopist is unprepared for the abnormal anatomy, perforation of an oesophageal carcinoma or diverticulum may occur. However, experienced endoscopists would perform endoscopy as the first investigation. Manometry is useful to assess the oesophagus and lower oesophageal sphincter in achalasia and other motility disorders. It can be performed on one occasion or dynamically over a 24-hour period. Reflux can be monitored by a pH-sensitive probe in the distal oesophagus (24-hour ambulatory oesophageal monitoring).

patients with reflux there are repeated episodes of significant quantities of acid, bile and food regurgitating into the lower oesophagus. This can lead to inflammatory change (reflux oesophagitis), which is associated with decreased oesophageal clearance of liquids and solids. Gastro-oesophageal reflux causes a spectrum of oesophageal abnormalities (Table 15.14a)

15

Aetiology The main cause of gastro-oesophageal reflux is the failure

TABLE 15.14a The spectrum of gastro-oesophageal reflux disease Reflux symptoms without oesophageal damage Reflux oesophagitis Oesophageal ulcer Benign oesophageal stricture Barrett's oesophagus Oesophageal adenocarcinoma

TABLE 15.14b Factors predisposing to gastro-oesophageal reflux GASTRO-OESOPHAGEAL REFLUX Normally, gastro-oesophageal reflux is prevented by the lower oesophageal sphincter, aided by the diaphragm. In

SUMMARY 1 The main symptoms, signs and investigations of oesophageal disease Symptoms Difficulty with swallowing (dysphagia) Painful swallowing (odynophagia) Refrosternal pain Heartburn Regurgitation Vomiting Weight loss Gastrointestinal bleeding Signs Epigastric tenderness or mass Abnormal palatal movement Supraclavicular lymphadenopathy Chest signs due to aspiration Investigations Full blood count Barium swallow Chest X-ray Endoscopy Manometry pH monitoring

Cause

Example

Disordered anatomy

Hiatus hernia Obesity Post-surgical treatment of achalasia Presence of tube for treating carcinoma

Myogenic

Systemic sclerosis Reflux oesophagitis

Hormonal

Pregnancy Menstruation

External influences

Smoking Eating fat/chocolate Coffee/tea Large meals Alcohol Drugs

TABLE 15.14c Non-oesophageal manifestations of gastro-oesophageal reflux Pulmonary aspiration -> wheeze and cough (especially nocturnal); recurrent chest infections Reflux laryngitis/pharyngitis -> sore throat, hoarseness Regurgitation into the mouth -> sore tongue, glossitis, dental enamel damage

763

CASE STUDY 15.2 A 60-YEAR-OLD MAN WITH A 3-MONTH HISTORY OF PROGRESSIVE DIFFICULTY IN SWALLOWING History A 60-year-old white Caucasian male clerical officer presented with a 3month history of progressive dysphagia. Initially the dysphagia was intermittent but it had become steadily worse, so that all solids 'stuck'. He had no difficulty in swallowing liquids. He felt food sticking in the low retrosternal area and was aware of discomfort in this area after swallowing. Over the previous 2 weeks he had regurgitated solid food after swallowing. He had become anorexic and had lost 1 stone in weight. On direct questioning he gave a history of intermittent heartburn and regurgitation for more than 10 years. These symptoms were controlled with the alginate/antacid mixture Gaviscon. He was a lifelong smoker, currently smoking 25 cigarettes a day. He had mild shortness of breath on exertion and coughed mucoid sputum first thing in the morning on most days, and had done so for several years. He was often wheezy. There was no other previous medical history and he had had no operations. He was taking a (32 agonist and beclomethasone by inhaler daily, and Gaviscon as required. He drank three pints of beer a day. There was no relevant family history. Examination He looked thin but not ill. He had a barrel chest with loss of cardiac dullness to percussion, and prolonged expiration over both lung fields on auscultation. He had no lymphadenopathy and his abdomen was normal. Investigations Haemoglobin 10.5 g/dL, MCV 70 fl, MCH 24 pg; white cell count and platelets, normal; plasma viscosity 2.00; biochemical profile normal except for serum alkaline phosphatase 150iu/L and y-glutamyl transferase 90iu/L;

764

chest X ray, emphysematous changes. Question 1. What is the most likely cause of the dysphagia, and why? Is the history of heartburn arid regurgitation significant? The patient had an upper GI endoscopy which showed that the last 10cm of the oesophageal body were replaced with Barrett's oesophagus. There was an ulcerated stricture extending 2 cm above the oesophagogastric junction. The scope passed with some difficulty through the stricture. There was a hiatal hernia and the stomach and duodenum were normal. Question 2, What histological changes would you expect in the distal oesophagus and stricture? Biopsies of the distal oesophagus showed specialized columnar epithelium (intestinal metaplasia) with areas of low-grade dysplasia, and biopsies of the stricture showed highgrade dysplasia and invasive adenocarcinoma. Brush cytology of the stricture revealed adenocarcinoma cells. Questions 3. What is the treatment of choice? What further investigations are needed? Abdominal and chest CT scans showed a concentric bulky tumour above the oesophagogastric junction. There was no involvement of the aorta and no lymphadenopathy, but a single metastasis was seen in the left lobe of the liver.

Treatment The results of the CT scan showed that a curative resection was not an option; therefore, he was considered for palliation. Radiotherapy was not considered as adenocarcinomas are poorly radiosensitive. The patient did not want to be considered for chemotherapy because he felt that as the results were so poor, he wanted to enjoy his remaining life without the side-effects of chemotherapy drugs. Question 4. What are the options for palliation? Endoscopic palliation was instituted. His malignant stricture was dilated and any exophytic tumour was injected with absolute alcohol to cause necrosis. He was started on omeprazole 40 mg daily to control symptoms of reflux, a potential problem after stricture dilatation. His swallowing was satisfactory for 5 weeks, but troublesome dysphagia returned. At a second endoscopy the stricture was again dilated and the tumour injected. However, he had only 10 days of improvement in his swallowing. Therefore, an expanding covered metal stent was placed through the stricture endoscopically. Two weeks later, he had sudden onset of complete dysphagia.

Question 5. What is the most likely reason for the sudden onset of dysphagia in this situation? Urgent endoscopy showed food bolus obstruction of the stent, which was easily relieved. He was discharged the same day after seeing the dietitian for further advice. He died 4 months later without having had any more troublesome dysphagia.

15

CASE STUDY 15.2 CONTINUED Latterly he had required opiate analgesia to control retrosternal chest pain and right upper quadrant abdominal pain. Discussion Dysphagia is a very important symptom which always requires investigation. In this case the dysphagia was progressive, which was suspicious of malignancy. However, other causes must be considered (see Table 15.13). There is debate as to whether the first investigation should be barium swallow or upper GI endoscopy. The patient can usually (but not always) localize the level of the obstruction accurately. Occasionally a high stricture may be associated with low dysphagia. Instrumentation of any stricture, especially a malignant one, is associated with an increased risk of perforation. Thus there is a risk that the endoscopist might perforate an unsuspected high oesophageal malignant stricture. For these reasons some specialists still insist that a barium swallow precede endoscopy. However, most endoscopists would not agree and regard a careful endoscopy by an experienced endoscopist as the appropriate first investigation of the patient with dysphagia.

This patient had Barrett's oesophagus (BO) complicated by an oesophageal adenocarcinoma. BO is a complication of long-standing gastro-oesophageal reflux (GOR): 10% of patients with endoscopic GOR have associated BO. Our patient had symptoms of GOR, namely heartburn and regurgitation, predating the development of dysphagia. Moreover, 2 agonist therapy taken for COPD will have reduced lower oesophageal sphincter tone and predisposed to GOR. BO can progress to oesophageal adenocarcinoma and 10% of BO patients have associated oesophageal adenocarcinoma at diagnosis. BO patients without carcinoma at diagnosis are offered endoscopic surveillance every 1-2 years, with multiple biopsies looking for premalignant dysplastic changes. BO is commoner in males; smoking and alcohol may be risk factors. Oesophageal adenocarcinoma has a much poorer prognosis than squamous carcinoma of the oesophagus, even if it appears operable initially. Less than 20% are deemed operable at diagnosis. The treatment options are more limited in that adenocarcinomas are less radiosensitive and respond less well to chemotherapy than do

of lower oesophageal sphincter tone and of secondary peristalsis in the oesophageal body. The initial pathological events are transient losses of lower oesophageal sphincter tone. Various other factors may predispose to reflux (Table 15.14b). Pregnancy is commonly associated with reflux, owing partly to sex hormones lowering the gastrooesophageal sphincter pressure, and partly to pressure from the intra-abdominal mass. Smoking, eating fat and chocolate, drinking coffee, tea and alcohol, and certain drugs (e.g. calcium antagonists and 2 agonists) act by reducing lower oesophageal sphincter pressure. The presence of a hiatus hernia is not necessary for reflux but is often an associated finding.

Clinical features The typical symptoms of gastro-oesophageal reflux are

squamous carcinomas. If surgery is to be considered, accurate preoperative staging with endoscopic ultrasound is increasingly thought to be essential. CT can miss local tumour infiltration of adjacent structures and local lymph node metastases. The present case was deemed inoperable because of the presence of a hepatic metastasis. If he had been a candidate for surgery the risks of the surgery would have been increased by his COPD. He would have needed preoperative lung function tests and assessment by a respiratory physician and anaesthetist. Most oesophageal adenocarcinomas are treated by palliation only. Simple dilatation with tumour destruction by laser or alcohol injection can give symptomatic relief for weeks or even months, but most patients will require stenting sooner or later. Covered metal stents require less oesophageal dilatation than do plastic stents, with a lower risk of perforation, and have a lower risk of tumour ingrowth than uncovered ones. With all stents, food bolus obstruction is a problem and all patients must receive appropriate dietary advice to minimize its occurrence.

heartburn and regurgitation; chest pain may also occur. They are often aggravated by lying down, bending or raising intra-abdominal pressure (e.g. by lifting). If reflux is complicated by oesophagitis or stricture formation, vomiting, haemorrhage, anaemia, dysphagia and pulmonary aspiration may occur. Gastro-oesophageal reflux may present with nonoesophageal symptoms (Table 15.14c).

Investigation Gastro-oesophageal reflux with typical symptoms requires little investigation. However, if the symptoms are less typical, severe, or include gastrointestinal bleeding or dysphagia, further investigation is indicated. Endoscopy is preferred, as this allows more accurate visualization of

765

associated lesions (oesophagitis, ulcers, stricture formation etc.) and permits biopsies of inflamed, ulcerated or questionably malignant areas. However, endoscopy does not permit satisfactory demonstration of reflux itself; this can sometimes be achieved by a barium swallow or the ingestion of a radioisotope. Objective assessment of reflux can be obtained by placing a pH probe in the lower oesophagus and measuring falls in pH due to acid regurgitation over a 24-hour period. Manometry demonstrates the loss of lower oesophageal sphincter tone but is not necessary as a routine test.

REFLUX OESOPHAGITIS 1 Reflux oesophagitis is an inflammatory response secondary to gastro-oesophageal reflux. It can be graded into four stages: 1. Mild. Single or non-confluent areas of superficial mucosal erosions and erythema. 2. Moderate. Confluent erosions not including the whole circumference of the oesophagus. 3. Severe. Confluent erosions including the whole oesophageal circumference. 4. Complicated. Oesophagitis with additional complications, e.g. peptic ulcers, strictures.

Management Medical Obesity is frequently a major causal factor. Raising the head of the bed on blocks by 20cm or more discourages nocturnal acid regurgitation and is more satisfactory than increasing the number of pillows. The avoidance of big meals before going to bed and precipitating foods and drinks also often helps. In mild cases a mixture of alginic acid and a simple antacid is effective. In patients with persisting symptoms (e.g. Gaviscon), a 6-week course of fulldose H2-receptor antagonists (taken twice daily or all at night) or a prokinetic drug (e.g. metoclopramide or domperidone) is often effective. Prokinetic drugs increase lower oesophageal sphincter tone and gastric emptying, and H2-receptor antagonists reduce gastric acid secretion. However, there is increasing use of proton pump inhibitors as first-line therapy in all patients with persisting reflux symptoms. Proton pump inhibitors profoundly reduce hydrogen ion secretion from gastric parietal cells. In severe erosive oesophagitis or cases resistant to other therapy, proton pump inhibitors for at least 8 weeks are the treatment of choice. Relapse after stopping treatment is a problem in 30-80% of patients, who may require maintenance antacid treatment. 1 Fig. 15.5

766

4 Fig. 15.8

2

Fig. 15.6

3 Fig. 15.7

SUMMARY 2 Treatment of gastro-oesophageal reflux and oesophagitis • Simple measures - Lose weight - Raise head of bed on blocks - Avoid bending • Avoid possible aetiological factors which reduce lower oesophageal sphincter pressure - Smoking - Eating fat/chocolate - Coffee/tea - Alcohol • Drugs to reduce gastric acid secretion - H 2 -receptor antagonists (cimetidine, ranitidine, famotidine, nizatidine) - Proton pump inhibitors (omeprazole, lansoprazole, pantoprazole, rabeprazoie) • Drugs to increase gastric emptying and increase lower oesophageal sphincter tone - Metoclopramide - Domperidone • Drugs to protect mucosal lining - Alginic acid containing antacid - Carbenoxolone • Drugs to neutralize gastric acid - Simple antacids • Surgery

Surgical Surgery may be indicated for the very few patients with severe symptoms of reflux unresponsive to all medical treatment. It should be considered in young patients who require maintenance acid suppression treatment to prevent relapse. Selected patients should preferably be non-obese, free of serious medical problems and under 70 years of age. Surgery involves an antireflux procedure (e.g. fundoplication) and can now be done laparoscopically. The short- and long-term results of antireflux surgery are sometimes disappointing.

HIATUS HERNIA Hiatus hernias can be congenital, in which case they usually present and are treated during childhood. Acquired hiatus hernias are often of the sliding type and result in varying lengths of the stomach being in the chest (Fig. 15.5). The rolling or paraoesophageal type is less common, and occurs when a portion of the fundus of the stomach lies anterior to the oesophagus.

Clinical features Sliding hiatus hernias may be asymptomatic or associated with symptoms of gastro-oesophageal reflux. Difficulty with swallowing suggests the presence of a complicating stricture.

symptoms of reflux as appropriate and are offered surveillance endoscopy, with multiple biopsies every 1-2 years looking for dysplastic or frank malignant changes. Patients with high-grade dysplasia or carcinoma are offered oesophagectomy. An alternative treatment for patients unfit for surgery is laser or photodynamic therapy. PPI therapy or antireflux surgery do not cause complete regression of Barrett's oesophagus, but it is possible that they prevent progression or the development of dysplasia.

15

OESOPHAGEAL ULCERATION FIG. 15.5 Diagram of A normal anatomy, B sliding hiatus hernia and C rolling hiatus hernia

The symptoms of rolling hiatus hernia do not include reflux and are commonly very non-specific, such as vague upper abdominal discomfort, nausea, and fullness after eating. Complications include bleeding from gastritis or a gastric ulcer (sometimes presenting as an iron deficiency anaemia), gastric volvulus and hernial strangulation.

Peptic ulcer in the oesophagus is caused by acid reflux and may be located in squamous mucosa above the cardia or within a segment of Barrett's oesophagus in the distal oesophagus. Oesophageal ulcer may present with retrosternal pain, odynophagia, dysphagia, and overt or occult bleeding in patients who often have a history of heartburn. It is diagnosed endoscopically and must be biopsied to exclude malignancy. It is treated with a PPI for 8 weeks and healing is confirmed by repeat endoscopy. Most patients will require maintenance acid suppression after healing. Healing may be associated with stricture formation.

Investigation If a sliding hiatus hernia is large it may be seen on a chest X-ray as a retrocardiac shadow with a fluid level. Endoscopy is the diagnostic procedure of choice, allowing direct assessment of hernial size, oesophagitis and stricture formation, the taking of diagnostic biopsies and the passage of dilating bougies to treat a stricture. Alternatively, a barium swallow can be used to assess hernia size and the development of complications.

Management A sliding hiatus hernia does not require treatment but its complications, such as reflux, oesophagitis and stricture formation, do. Surgery is sometimes indicated to reduce a large rolling hernia in view of the risk of strangulation. However, as rolling hernias are often found in the very elderly, surgery may be postponed until a major complication occurs.

PEPTIC OESOPHAGEAL STRICTURE 4 Peptic oesophageal strictures are fibrous strictures which occur at the lower end of the oesophagus as a result of persistent gastro-oesophageal reflux. The main symptom is intermittent and then progressive dysphagia, first for solids and then for liquids. Patients become frightened to eat, regurgitate meals, develop increasing pain and sometimes aspiration, and as a result become anorexic and lose weight. Treatment is by the passage of dilators with the aid of an endoscope. Treatment can be repeated if symptoms recur, but many patients only require it once or every few years. Oesophageal perforation is rare. Maintenance proton pump inhibitor therapy reduces the risk of restricturing. Surgery may be indicated in younger individuals requiring frequent dilatation.

DISSECTING INTRAMURAL HAEMATOMA BARRETT'S OESOPHAGUS 2 Barrett's oesophagus is a complication of long-standing gastro-oesophageal reflux. In response to the reflux, the squamous epithelium in the distal oesophagus is replaced by a columnar epithelium which usually includes areas of intestinal metaplastic epithelium. The latter is unstable and can become dysplastic, with eventual progression to adenocarcinoma. 3 Barrett's oesophagus is discovered during the investigation of patients with reflux symptoms or complications such as ulcer or stricture. Patients are treated for

This condition typically presents with chest pain, odynophagia, dysphagia particularly for solids, and minor haematemesis. The onset of pain is usually sudden and may follow retching or a Valsalva manoeuvre. The diagnosis is made by endoscopy, when a long dark submucosal haematoma projecting into one side of the oesophageal lumen is seen. Management is conservative, with the patient being kept nil by mouth until the pain and dysphagia have settled. Conservative treatment results in complete resolution of symptoms after 1-2 weeks.

767

CORROSIVE AND FOREIGN BODY INGESTION

In certain parts of the world (e.g. India and South America), strong acid or alkali solutions are commonly taken with suicidal intent. These cause severe inflammation, most commonly affecting the oesophagus, although the stomach and duodenum can also be damaged. In some patients oesophageal perforation requires urgent surgery. In patients without perforation, conservative treatment with neutralizing lavage can be successful without long-term complications, but in others a stricture develops. Gastric or duodenal strictures usually require surgical intervention. Foreign bodies (e.g. fish bones, coins) may stick in the oesophagus and cause dysphagia. Small articles can be extracted endoscopically. A bolus of food often obstructs the oesophagus of patients with benign or malignant strictures, and this can usually be treated by using the endoscope to push the bolus distal to the obstruction. When larger solid articles are swallowed and are thought likely to obstruct the bowel or cause perforation, a laparotomy should be considered.

The incidence of squamous oesophageal cancer varies greatly in different regions of the world, with a high incidence in central Asia, the Far East and Transkei (Fig. 15.6). The age-specific incidence is shown in Figure 15.7. There are also racial differences: squamous carcinoma is commoner in North American blacks than in whites.

Aetiology Possible aetiological factors in squamous cell carcinomas of the oesophagus are cigarette smoking, high alcohol intake, coeliac disease, radiation, achalasia, ingestion of strong acids and alkalis, environmental carcinogens, hot drinks, nutritional deficiencies and human papilloma virus. Squamous carcinoma is rarely preceded by histological dysplasia, which can be detected by endoscopic biopsies. Barrett's oesophagus predisposes to the development of adenocarcinomas and premalignant dysplasia may be seen. The Plummer-Vinson syndrome, which occurs in

OESOPHAGEAL TUMOURS 1 A basic classification of oesophageal tumours is shown in Table 15.15. Benign tumours are rare. Malignant tumours occur in the upper third of the oesophagus (about 15%) but are more common in the middle (about 40%) and lower thirds (45%). Half of oesophageal tumours are squamous cell in origin and half are adenocarcinomas. The incidence of adenocarcinoma is increasing, probably secondary to the increasing incidence of reflux and Barrett's oesophagus. There were 6240 deaths from carcinoma of the oesophagus in the UK in 1992. The tumours spread early by local invasion and lymphatic and bloodborne metastases.

FIG. 15.6 Areas of high incidence of oesophageal squamous carcinoma worldwide

TABLE 15.15 Basic classification of oesophageal tumours

768

Class

Example

Malignant

Squamous cell carcinoma Adenocarcinoma Adenosquamous carcinoma Melanoma Sarcoma (usually leiomyosarcoma)

Benign

Leiomyoma Adenoma Cysts

1 MCQ 15.5

2 Fig. 15.9

4 Fig. 15.11

5 MCQ 15.6

3 Fig. 15.10 FIG. 15.7 Age-specific incidence of squamous carcinoma of the oesophagus in the UK

15

SUMMARY 3 Symptoms and signs of oesophageal carcinoma Symptoms Dysphagia (initially solids, then liquids, then complete) Pain Weight loss Cough Acute gastrointestinal bleeding

Signs Weight loss Anaemia Enlarged cervical and supraclavicular lymph nodes Hepatomegaly

long-standing iron deficiency anaemia is, rarely, associated with postcricoid squamous carcinoma. A rare inherited syndrome of palmar hyperkeratosis (tylosis palmaris) is associated with squamous oesophageal cancer.

Clinical features Dysphagia is the main symptom of oesophageal carcinomas, and is rapidly progressive. Pain is usually retrosternal and is due to invasion into the mediastinum. A productive cough may reflect aspiration or a broncho-oesophageal fistula. Patients with adenocarcinoma may give a history suggestive of gastro-oesophageal reflux. Paraneoplastic syndromes, e.g. hypercalcaemia, may be seen in patients with squamous carcinomas. On examination, weight loss is often obvious. Anaemia usually reflects bleeding. Lymph node enlargement and hepatomegaly due to distant spread are late signs.

Investigation Patients who present with dysphagia often localize the point where food sticks, but this does not always correlate with the anatomical site of obstruction. The first investigation is a barium swallow (Fig. 15.8) 2 or an endoscopy 3 with biopsy to ascertain cell type. A chest X-ray is performed to assess mediastinal spread, and liver function tests and ultrasound are performed to detect metastases. Distant metastases, usually hepatic, are a contraindication to surgery. If curative surgery is contemplated imaging of the chest and abdomen with CT or MRI is necessary to stage the tumour. 4 Endoscopic ultrasound gives useful information about local spread.

Management The management of squamous oesophageal carcinoma (Table 15.16) is controversial because the tumours are radiosensitive and the best results after radical radiotherapy may be as good as after surgery. Surgery or radical radiotherapy may be effective in patients who are fit, with no involved nodes or metastases, but these patients make up only 30-40% of those presenting with the disease. In the upper third radiotherapy may be the preferred treatment, as radical surgical treatments are technically difficult. Surgery involves total oesophagectomy and colonic

FIG. 15.8 Barium swallow examinations A A smooth benign oesophageal stricture (white arrow) above a hiatus hernia B A typical carcinoma with a long irregular stricture narrowing the oesophageal lumen.

interposition. In the middle and lower thirds surgery is preferred to radiotherapy if the patient is fit enough. The stomach is mobilized and anastomosed to the oesophagus in the chest. Radiation may be complicated by oesophagitis and stricture, and surgery by stricture, mediastinitis and anastomotic leaks. Overall survival with either treatment is probably no better than 5%, but some surgical series report 30% survival at 5 years. Malignant stricture formation can be palliated by repeated endoscopic dilatations; when these are needed more than once every 6 weeks it is useful to insert an expanding metal stent to keep the lumen open and to allow liquidized food to pass. The stent can be placed using a fibreoptic endoscope under local or general anaesthesia, and is also useful for sealing off fistulae with bronchi or the trachea. Stents can become blocked by food and fizzy drinks, and endoscopic procedures can be used to clear the blockage. Repeated endoscopic laser treatments or absolute alcohol injections cause tumour destruction and are alternatives to stents, and can be used to deal with tumour overgrowth of a stent. Adenocarcinomas are not radiosensitive and few are amenable to curative surgery, with less than 5% 5-year survival.

ACHALASIA 5 Achalasia is a functional obstruction at the level of the lower oesophageal sphincter caused by a failure of relaxation. It results in progressive dilatation of the oesophagus proximal to the 'obstruction'. Achalasia is caused by the

769

TABLE 15.16 Treatment of squamous carcinoma of the oesophagus Aim

Treatment modality

Criteria

Result

'Curative'

Radical surgery (with or without pre- or postoperative radiotherapy) Radical radiotherapy

Tumour localized to oesophagus

Less than 40% 5-year survival for squamous cell carcinoma Less good than surgery, usually reserved for elderly and infirm

Palliative

Radiotherapy Endoscopic dilatation Endoscopic intubation Endoscopic laser therapy Pain relief and general nutritional support Percutaneous endoscopic gastrostomy

Tumour localized to oesophagus Widespread local tumour; distant metastases; other medical conditions preclude surgery or radical radiotherapy

Good palliation, but few survive more than 6 months

degeneration of ganglion cells in the muscle of the mid and lower oesophagus. In the west the aetiology is obscure, although the disease is sometimes familial. In South America a similar problem occurs as a late result of destruction of the ganglion cells by the protozoon Trypanosoma cruzi (Chagas' disease; see p. 354). Pseudoachalasia may be seen in malignant disease. The symptoms and manometric characteristics are similar to those of conventional achalasia.

Clinical features and investigation Presentation is usually with intermittent but slowly progressive dysphagia for both solids and liquids. There may be associated pain, and regurgitation of food swallowed some hours before may occur; heartburn is unusual. Anorexia and weight loss may also occur and some patients develop symptoms of pulmonary aspiration. Carcinoma of the oesophagus is a rare complication of achalasia. The diagnosis is often made radiologically. A barium swallow reveals a failure of the lower oesophageal sphincter to relax, and sometimes proximal dilatation with food and fluid in the lumen (Fig. 15.9). However, the barium swallow can be normal, and therefore patients with dysphagia who have a normal swallow and endoscopy should have oesophageal manometry. Manometry confirms the diagnosis by demonstrating high lower oesophageal resting tone with poor relaxation to swallowing, associated with weak or absent peristalsis in the oesophageal body. Upper gastrointestinal endoscopy excludes mechanical obstruction at the cardia, especially malignancy, and often reveals food in the oesophagus despite a prolonged fast.

Management

FIG. 15.9 Barium swallow examination showing oesophageal achalasia There is hold-up of barium at the level of the gastro-oesophageal junction (arrow) and oesophageal dilatation and lack of contractions above. Advanced cases also have a fluid level which may be seen on a chest X-ray. 1

Achalasia may be treated with:

1

770

Fig. 15.12

• pneumatic balloon dilatation under endoscopic control; • oesophageal myotomy (Heller's operation); • intrasphincteric botulinum toxin. All these treatments can be effective but can lead to troublesome reflux. Botulinum toxin injection has the

poorest medium-term results. Balloon dilatation has good medium- to long-term results, but if more than three dilatations are required surgery should be considered.

OTHER OESOPHAGEAL DYSMOTILITY SYNDROMES Diffuse oesophageal spasm is usually seen in the elderly and presents with retrosternal chest pain and/or dysphagia. A barium swallow shows a normal oesophagus, but the oesophagus may undergo diffuse spasm during the investigation to produce a corkscrew configuration. Manometry shows intermediate bursts of high-amplitude contractions, often unrelated to swallowing. Angina is a major differential diagnosis, but oesophageal spasm is not precipitated by exertion and stress. Treatment is also similar to that for angina, as glyceryl trinitrate sublingually (as necessary) and regular isosorbide dinitrate or nifedipine often help. In resistant cases balloon dilatation or myotomy can be considered, but the results are not as good as for achalasia. Other motility disorders. Nutcracker oesophagus and hypertensive lower oesophageal sphincter are defined manometrically, present with chest pain and dysphagia, and respond variably to nitrates, calcium channel blockers and antidepressants. Irritable oesophagus (p. 820) is seen in association with irritable bowel syndrome and manometry is usually normal.

SYSTEMIC SCLEROSIS Systemic sclerosis causes loss of oesophageal contractions (which may also occur in patients with Raynaud's phenomenon, systemic lupus erythematosus and polymyositis). Dysphagia and heartburn may occur. A barium swallow shows lack of peristalsis and hence failure of the barium to move if the patient lies flat. The gastrooesophageal sphincter may become incompetent, with consequent risk of reflux, oesophagitis and peptic stricture.

ANATOMICAL ABNORMALITIES Oesophageal diverticula are outpouchings of the oesophageal wall that can occur at three main locations: immediately above the upper oesophageal sphincter (Zenker's diverticulum, or pharyngeal pouch), midoesophagus (traction diverticulum) and immediately above the lower oesophageal sphincter (epiphrenic diverticulum). Dysphagia and regurgitation are typical symptoms of pharyngeal pouch. If symptoms become severe, corrective surgery may be necessary. Traction and epiphrenic diverticula rarely cause symptoms. Oesophageal webs are thin structures impinging a few millimetres into the lumen and consisting of mucosa and

submucosa. They occur throughout the oesophagus and can be multiple. An association with chronic iron deficiency is now rare; treatment of the latter may prevent an associated postcricoid carcinoma. Dilatation helps if dysphagia occurs. Oesophageal rings occur at the lower end of the oesophagus and contain muscle as well as mucosa and submucosa. They can be asymptomatic or present with intermittent dysphagia, which may become more common as the years pass. If symptoms are serious, endoscopic dilatation usually relieves them.

15

OESOPHAGEAL INFECTIONS The most important oesophageal infections are candidiasis and herpes simplex. Both occur in patients who are immunosuppressed, e.g. patients who are receiving corticosteroids or cytotoxic drugs, or who have HIV infection or diabetes. Antibiotic treatment predisposes to candidal infections. Patients with oesophageal infections usually present with dysphagia and odynophagia. Diagnosis of candidiasis is by endoscopic demonstration of the creamy exudates on the oesophageal wall (the mouth and pharynx are often also involved) and by demonstration of the typical yeast hyphae on biopsy. Herpes simplex appears similar at endoscopy, but can be confirmed by histological demonstration of inclusion bodies in epithelial cells and by viral culture. The treatment of oesophageal candidal infection is with nystatin or amphotericin suspension, but severe or recurring attacks are treated with a systemic antifungal drug, e.g. fluconazole. Herpes simplex is treated symptomatically with local anaesthetic drinks, although acyclovir is necessary in severaly immunosuppressed patients. Herpes usually settles over 1-2 weeks.

FURTHER READING ON THE OESOPHAGUS Barham C P, Jones R L, Biddlestone L R, Hardwick R H, Shepherd N A, Barr H 1997 Photothermal ablation of Barrett's oesophagus. Endoscopic and histological evidence of squamous reepithelialisation. Gut 41:281-284. Pera M, Cameron A J, Trastek V F, Carpenter H A, Zinsmeister A R 1993 Increasing evidence of adenocarcinoma of the esophagus and esophago-gastric junction. Gastroenterology 104:510-513. El-Serag H B, Sonnenberg A 1997 Comorbid occurrence of laryngeal or pulmonary disease with esophagitis in US military veterans. Gastroenterology 113:755-760. Klinkenberg Knol E C, Festen H P M, Jansen J B M J, Lamers C B H W, Nelis F, Snel P 1994 Long term treatment for erosive esophagitis with omeprazole; efficacy and safety. Ann Intern Med 121:161-167. McDougall N I, Johnson B T, Kee F, Collins J S A, McFarland R J, Love A H G 1996 Natural history of reflux oesophagitis: a ten year follow up of its effect on patient symptomatology and quality of life. Gut 38:481-486. Katz P O, Gilbert J, Castell D O 1998 Pneumatic dilatation is effective in long term treatment for achalasia. Dig Dis Sci 43:1973-1977.

771

Knyrim K, Wagner H-J, Bethge N, Keymling M, Vakil N 1993 A controlled trial of an expansile metal stent for palliation of oesophageal obstruction due to inoperable cancer. N Engl J Med 329:1302-1307. Trimble K C, Pryde A, Heading R C 1995 Lowered oesophageal sensory thresholds in patients with symptomatic but not excessive gastro-oesophageal reflux: evidence for a spectrum of visceral sensitivity in gastro-oesophageal reflux disease. Gut 37:7-12.

GASTROINTESTINAL BLEEDING ACUTE GASTROINTESTINAL BLEEDING

1

Upper gastrointestinal bleeding presenting with haematemesis (vomiting blood or altered 'coffee ground' material) and/or melaena (the passage of black tarry faeces) is a common medical emergency. In the UK it has an annual prevalence of about 50 cases per 100000 population. The different sites of gastrointestinal bleeding and approximate mortality rates in the UK are shown in Figure 15.10.

Aetiology The causes of gastrointestinal bleeding are shown in Table 15.17. Peptic ulcers, gastritis, oesophagitis Mallory-Weiss tears, varices and tumours account for the majority of cases. Bleeding may be provoked by the ingestion of alcohol, aspirin and non-steroidal anti-inflammatory drugs (NSAIDs). Corticosteroids may cause existing peptic ulcers to bleed. A Mallory-Weiss tear is a small lesion at the gastro-oesophageal junction, which usually follows vomiting and causes an episode of self-limiting bleeding. Systemic and vascular causes are rare. Occasionally patients invent the symptoms, ingest blood from outside (e.g. from animals) or deliberately break their mucosa to cause bleeding. Haematemesis is usually from a bleeding site proximal to the distal duodenum. If there is rapid intestinal transit and very severe haemorrhage, fresh blood from upper gastrointestinal lesions may be passed per rectum. Melaena is altered, partially digested blood passed per rectum. Meckel's diverticula are uncommon developmental abnormalities in the distal ileum; these frequently contain gastric acid-secreting cells, and usually present with recurrent melaena before adolescence. Angiodysplasia is increasingly recognized as a cause of melaena in the elderly. In this condition there are leashes of dilated mucosal vessels which usually occur in the ileum and right side of the colon, although any part of the gastrointestinal tract may be involved.

1 MCQ15.7 772

FIG. 15.10 Gastrointestinal bleeding: sites of origin and mortality rates in the UK A Typical values for the site of origin of gastrointestinal bleeding. B Approximate mortality rates for different causes of gastrointestinal bleeding. C Mortality rates related to age showing a rise from 60 years of age onwards.

Clinical features The history is important in diagnosis, particularly with regard to previous symptoms (e.g. dyspepsia, pain, vomiting, weight loss), past medical history (e.g. previous episodes, ulcers, liver disease) and ingestion of alcohol, aspirin, NSAIDs, warfarin and corticosteroids. Overt gastrointestinal bleeding may present with haematemesis, melaena or rectal bleeding. With severe bleeding there will be signs of shock, with a tachycardia and hypotension (particularly a postural drop in blood pressure). Anaemia suggests either acute bleeding over several days with time for haemodilution, or an underlying chronic disease which has been causing blood loss over weeks or months. Spooning of the nails (koilonychia) implies long-standing iron deficiency. Signs of chronic liver disease suggest varices as a possible aetiology. Vascular

CASE STUDY 15.3 HAEMATEMESIS AND MELAENA IN A 66-YEAR-OLD MAN History A 66-year-old white Caucasian retired postman was admitted with a 5-day history of passing black tarry stools. On the day of admission he had vomited a teacupful of dark blood. For more than 20 years he had had occasional episodes of mild epigastric pain and indigestion. The symptoms tended to occur 1 hour or so after eating and in bed at night, and were relieved with antacids. Two days before the onset of black stools he had taken ibuprofen for a sore left knee. Since then he had had severe epigastric pain. There was no other relevant history. Examination He was pale, sweaty and agitated, with a pulse of 110pm and a blood pressure of 90/50. He had epigastric tenderness to light palpation. Rectal examination revealed black stool.

Questions 1. What is the likely diagnosis? 2. What is your immediate management? 3. What tests would you arrange? Investigations Haemoglobin 9.8g/dL; red cell indices, white blood cell count and platelet count, normal. Routine biochemistry, normal except for blood urea 15.5mmol/L. Prothrombin time, normal. Upper GI endoscopy, no blood in upper gastrointestinal tract; chronic duodenal ulcer (DU) on anterior wall of duodenal cap with fresh clot; rapid urease test for H. pylori, positive at 2 hours.

Questions 4. How severe was this bleed? 5, What factors are present in this case which increase the risk of a poor outcome?

Management After initial assessment he was given oxygen and a large-bore i.v. cannula was inserted immediately, through which crystalloids were infused. A CVP line and urinary catheter were then inserted. His blood pressure improved and his urine output was good. He was transferred to a highdependency area for monitoring. As soon as it was available, he was transfused four units of blood.

Question 6. At what stage is acid suppression therapy given in acute upper GI haemorrhage?

Six hours after admission he was endoscoped, which showed a chronic DU with a stigma of recent haemorrhage. The clot was gently washed off the ulcer and the bleeding point injected with 1:1000 adrenaline (epinephrine). Afterwards he was started on omeprazole 20 mg b.d. There was no rebleeding and he was able to go home after 48 hours. He was discharged on omeprazole 20 mg b.d. for 1 month and amoxycillin 1 g b.d. and clarithromycin 500mg b.d. for 1 week; he was told to avoid aspirin and NSAIDs. On review 6 weeks later he was well. A urea breath test confirmed the eradication of H. pylori. Discussion This man gave a typical history of duodenal ulcer. He bought ibuprofen over the counter for a sore knee, with the result that the NSAID induced the ulcer to bleed. The bleed was a severe one, as the presence of tachycardia, hypotension and raised blood urea at presentation indicated. Risk factors for a possible poor outcome were haematemesis, anaemia, shock and raised blood urea at presentation, age, bleed as a result of chronic peptic ulcer and NSAID

15

ingestion. In view of the shock, oxygen was given and a CVP line and a urinary catheter were inserted. All patients who have had an acute upper GI bleed are potentially unstable, and must be monitored closely for continued bleeding and rebleeding in a high-dependency area. Upper GI endoscopy is mandatory in all cases of acute upper GI haemorrhage and should be performed after the patient has stabilized, if that occurs, and no later than 18 hours after the bleed. In practice, this means the morning after admission in most cases. Endoscopy allows identification of the source of the bleed and permits topical therapy where appropriate. In this case the ulcer showed one of the stigmata of recent haemorrhage, fresh clot, which indicates a high risk of rebleeding. This risk is reduced by injecting the bleeding point in the ulcer with adrenaline (epinephrine). Other therapies available are injection with sclerosant, heater probe coagulation and laser coagulation. There is no evidence that immediate treatment with acid suppression improves outcome, and so this can be delayed until the cause of the bleed is known. In peptic ulcer H. pylori infection is highly likely, therefore gastric antral biopsies can be taken. However, biopsies should probably not be taken if the patient is actively bleeding, and the presence of blood in the stomach can cause antral biopsies to be falsely negative for H. pylori. Although NSAID ingestion caused the duodenal ulcer to bleed, the patient clearly had a long-standing DU as indicated by the history. Therefore, the presence of H. pylori would be suspected and was confirmed at endoscopy. However, H. pylori is present in 75% of patients with NSAID-induced peptic ulcers. Because of the role of H. pylori in causing PU relapse, eradication therapy is essential and particularly so in a patient who has

773

CASE STUDY 15.3 CONTINUED bled from a PU. Moreover, eradication must be confirmed by the most accurate method, a urea breath test. Failure to eradicate first time round occurs in up to 20% of well motivated patients whatever the regimen, and therefore a further

course of eradication therapy and a repeat urea breath test are required. There is no need to repeat the endoscopy in a DU patient who is asymptomatic after treatment and has had H. pylori eradication confirmed. NSAIDs should be stopped if

malformations on the face and mucous membranes of the lips or mouth suggest hereditary telangiectasia.

Investigation The most precise investigation for finding the cause of upper gastrointestinal bleeding is endoscopy. The patient should be resuscitated and endoscoped as soon as is practicable. At this stage most Mallory-Weiss tears, acute ulcers and areas of acute inflammation are obvious and the bleeding point may be identified. The presence of stigmata of recent bleeding implies a bad prognosis for ulcer rebleeding; in order of importance these include: • • • • •

a spurting artery active oozing fresh blood clot organized blood clot black spots.

The small intestine, beyond the distal duodenum and proximal to the distal terminal ileum, can be reached by enteroscopy or endoscopy during a laparotomy. Small intestinal bleeding may, however, be identified preoperatively by arteriography. Colonic bleeding may be identified by sigmoidoscopy or colonoscopy, or by arteriography if the bleeding is sufficiently profuse.

Management

774

A joint medical and surgical approach to acute gastrointestinal bleeding is important. Before investigation, if there are signs of shock, a central venous pressure line and urinary catheter are essential. Blood should be given if the haemoglobin is less than lOg/dL or if there is a tachycardia or hypotension. Antisecretory drugs should be avoided in upper gastrointestinal haemorrhage until a diagnosis has been made, as there is little evidence that early use reduces operation rates or mortality. Tranexamic acid (which inhibits fibrinolysis) and octreotide (which reduces acid secretion and splanchnic blood flow) may have a role in controlling bleeding. Vasopressin is usually reserved for variceal bleeding (Ch. 16, p. 873). A management summary of acute upper gastrointestinal bleeding is shown in Table 15.18. Perendoscopic thermal coagulation, laser therapy or injection with adrenaline and

possible, and this was not a problem in the present case. However, in cases where continued NSAID therapy is necessary to control symptoms, e.g. in severe arthritis, PPIs are as effective in healing PUs as when the patient is not taking an NSAID.

sclerosants may be used on any bleeding vessels seen. Surgery is usually limited to peptic ulcer bleeding and is indicated if the bleeding is life-threatening, restarts after apparently stopping, or is persistent despite receiving 3 or more litres of blood, especially if the patient is over 60 years of age. Early surgery is indicated if there is a visible vessel at the base of an ulcer, or if an operable carcinoma is present. Surgery is best avoided in self-limiting conditions, such as Mallory-Weiss tear, and in clotting disorders or varices (Ch. 16, p. 873). The overall mortality of acute upper gastrointestinal bleeding is 5-10% with each episode. However, this relatively high mortality is mostly due to deaths in patients aged over 65 (Fig. 15.10) who have associated disorders (such as cardiovascular, pulmonary or renal diseases) rather than to exsanguination. Lower gastrointestinal haemorrhage rarely requires emergency surgery, but surgery may be necessary for the cause of the bleed (e.g. cancer or diverticular disease).

CHRONIC GASTROINTESTINAL BLEEDING Aetiology and clinical features A microcytic hypochromic anaemia is a common presentation of chronic gut blood loss. The commonest causes are shown in Table 15.19. Patients usually present with symptoms of anaemia - malaise, dyspnoea and angina. Clinical examination usually shows signs of anaemia, but the classic signs of iron deficiency (glossitis, cheilitis and koilonychia) are rare.

Investigation The purpose of investigations is to identify the site of the bleeding. Positive faecal occult blood testing indicates gastrointestinal blood loss, but bleeding may be intermittent and so false negatives can occur. False positives may occur with meat ingestion and bleeding gums. Other causes of blood loss (e.g. from the genital and urinary tracts) must be considered. In the presence of upper gastrointestinal symptoms, or if there are no specific symptoms and signs, upper gastrointestinal endoscopy is usually performed first. This may

TABLE 15.17 Causes of gastrointestinal bleeding (parentheses indicate a rare cause)

TABLE 15.18 Management guidelines for acute upper gastrointestinal haemorrhage

Causes of haematemesis Oesophagus Carcinoma Oesophagitis Peptic ulcer Varices

• Initial assessment Full history and examination after any appropriate initial resuscitation (oxygen; i.v. fluids to expand plasma volume) especially noting history of previous dyspepsia and Gl bleeds, NSAID, aspirin and corticosteroid therapy, alcohol abuse, signs of shock and liver disease. A rectal examination is essential. • High-risk patients Patients at higher risk of rebleeding or of bleed-related mortality include those aged more than 70, and those with other serious diseases, hypovolaemic shock, evidence of chronic liver disease (suggesting varices) or evidence of continued bleeding or rebleeding. • Intravenous therapy Insert a large bore i.v. cannula or, in cases of shock, a CVP cannula. Give plasma expanders until blood is ready in shocked patients. Give blood transfusion to shocked patients and to patients with a haemoglobin <10g/100mL In shocked patients insert a urinary catheter. • Blood tests Take blood for haemoglobin, prothrombin time, serum urea, electrolytes and liver function tests and for cross-match (at least 4 units of blood). • Admit to clinical area to monitor closely for continued bleeding or rebleeding. Notify on-call surgical team. • Upper Gl endoscopy is required within 18 hours of admission or earlier if the patient continues to bleed after admission. Local injection therapy is used for active bleeding sites. • Treatment depends on endoscopic findings; antacid drugs given on admission do not reduce mortality from bleeding. • Surgery is indicated in non-variceal patients if: - hypovolaemia cannot be controlled - there is evidence of rebleeding - transfusion requirements are more than 4 units in those aged over 60 and more than 8 units in those aged less than 60.

Stomach Peptic ulcer Acute erosions (often NSAID associated) Carcinoma Acute gastritis (often alcohol associated) Mallory-Weiss tear (Varices) (Other tumours) (Angiodysplasia) (Portal hypertensive gastropathy) Duodenum Peptic ulcer Duodenitis (Tumours)

Systemic causes Chronic renal failure (Thrombocytopenia) (Clotting factor deficiency) (Anticoagulant therapy) Vascular causes (Mesenteric ischaemia) (Hereditary telangiectasia) (Systemic vasculitis) Swallowed blood from oropharynx, nose and respiratory tract (Munchausen's syndrome) - symptom invented - ingested blood - self-induced

Causes of melaena Causes of haematemesis (see above) Jejunum and ileum (Meckel's diverticulum) (Angiodysplasia) (Tumours) (Crohn's disease) (Infarction of bowel) Colon Right-sided tumours Angiodysplasia (Inflammatory bowel disease) Causes of blood intermixed with stool Colon Left-sided colonic tumours (benign or malignant) Inflammatory bowel disease Diverticular disease Infective colitis Angiodysplasia Ischaemic colitis Causes of fresh blood per rectum Anus and rectum Haemorrhoids Proctitis of any cause Carcinoma of anus or rectum Crohn's disease of anus

15

Management Wherever possible, the primary cause of the bleeding is treated. The iron stores are replenished by giving oral iron in the form of ferrous sulphate, fumarate or gluconate. Oral iron can cause nausea, dyspepsia and altered bowel habit. If these side-effects develop, a lower dose or a different preparation should be tried. If the patient is unable to tolerate oral iron, a total dose infusion of intravenous iron dextran, or intramuscular doses of smaller amounts of iron sorbitol, can be used (p. 1207). Severe symptomatic anaemia may require blood transfusion; this is especially likely to occur if the primary cause cannot be found or treated successfully. If no cause is found, the patient is followed carefully and reinvestigated if the problem recurs.

THE STOMACH AND DUODENUM be combined with small bowel biopsy if there is a possibility of underlying iron malabsorption. If no abnormality is found in the upper gut, sigmoidoscopy with barium enema or colonoscopy are the next investigations. If still no cause is evident, colonoscopy (if not yet performed) and investigation of the small bowel - e.g. small bowel enema, Meckel scan (p. 755), arteriography or enteroscopy - must be considered.

Anatomy and physiology The lower end of the oesophagus and the fundus of the stomach are separated by the gastro-oesophageal sphincter at the cardia. which prevents reflux and allows the downward passage of food. The stomach is divided into three parts: fundus, body and antrum. The pyloric sphincter controls passage of gastric contents into the first part of

775

TABLE 15.19 Causes of chronic occult gastrointestinal bleeding in the UK (In approximate order of frequency) • Colonic neoplasm • • • • • • • •

Duodenal ulcer Reflux oesophagitis/oesophageal ulcer Gastric ulcer NSAID or aspirin-induced erosions and gastritis Gastric carcinoma Colonic diverticular disease Meckel's diverticulum Angiodysplasia

Note: Hookworm is the commonest cause worldwide.

protects the small intestine from bacteria. The pH is kept at around 2. In humans, HC1 secretion is mainly stimulated by the hormone gastrin, which is released in response to dietary protein, gastric distension and vagal stimulation. At the level of the parietal cell the final common pathway of HC1 synthesis is K+-H+-ATPase, which pumps H+ ions into the lumen of the stomach. The interrelationships of these stimuli are shown in Figure 15.12. • Acting as a reservoir. The stomach constantly churns food into small particles. Gastric emptying proceeds slowly so that the small bowel can digest and absorb maximally. Gastric emptying is a complex process, dependent on duodenal receptors; it is, for example, slower for fats than for protein. Control is partly vagal and partly hormonal (motilin accelerates the process, whereas enteroglucagon and neurotensin slow it down). • Secretion of intrinsic factor (see p. 751). • Secretion of pepsinogen to initiate proteolysis. The duodenum is largely retroperitoneal and is divided into first, second, third and fourth parts. The pH is lowest in the first part and nearly all duodenal ulcers are sited here. The medial border of the second part of the duodenum receives alkaline biliary and pancreatic secretions. The duodenum contains Brunner's glands, which secrete bicarbonate and help bring pH towards neutral. Villi are present throughout the duodenum and increase in size down its length. Both enzyme digestion and absorption of small molecules not needing digestion (e.g. calcium, iron, folic acid) occur in the duodenum.

FIG. 15.11 Diagram showing secretion of H+ by a gastric parietal cell There are receptors on the basolateral membrane for histamine (H2), gastrin and acetylcholine, all of which can mediate the activation of second messengers. A complex series of metabolic events then occurs, with the result that H+-K+-ATPase at the apical surface pumps H+ into the gastric lumen in exchange for K+. Antagonists exist for each of the three basolateral receptors (e.g. cimetidine A , proglumide C and pirenzipine D and for H+-K+-ATPase (omeprazole). The prostaglandin E2 receptor B inhibits the stimulatory effect of histamine on cAMP; misoprostol is a PGE2 analogue.

the duodenum. Parietal cells in the fundus and body of the stomach synthesize hydrochloric acid and secrete intrinsic factor; gastrin is synthesized in the antral G cells and chief cells in the crypts secrete pepsinogens. The stomach has four main functions: • The secretion of HCl, which converts pepsinogen to pepsin to initiate protein digestion into polypeptides and

1

776

MCQ15.8

Clinical features of gastroduodenal pathology The site of gastroduodenal pain is usually epigastric, but may radiate through to the back and to the right and left hypochondria. Duodenal ulcer pain is typically relieved by food and simple antacids. Functional (or non-ulcer) dyspepsia usually occurs at the time of maximal emotional stress and the site of pain is often poorly localized. Nausea, vomiting, anorexia and weight loss raise the possibility of luminal obstruction or a carcinoma. Bleeding can cause haematemesis and melaena. Epigastric tenderness may be present on superficial or deep palpation. An epigastric mass may be palpable, and can be difficult to distinguish from the liver. With gastric carcinoma, metastatic supraclavicular nodes and the liver may be palpable. A succussion splash, present 3 or more hours after the last drink or meal, suggests delayed gastric emptying, commonly due to pyloric obstruction. It is elicited by shaking the patient's upper abdomen while listening over the epigastrium.

Investigation of gastroduodenal pathology Upper gastrointestinal endoscopy is the preferred method

SUMMARY 4 The main symptoms, signs and investigations in the diagnosis of gastroduodenal disease

TABLE 15.20 Aetiological factors in peptic ulceration Causal factor

Gastric ulcer

Duodenal ulcer

Symptoms Pain or discomfort Site Relationship with food, drugs, stress Nausea and vomiting Haematemesis and melaena Anorexia Weight loss Heartburn Early satiety

Helicobacter pylori NSAIDs and aspirin Acid

Important important Normal or low acid output Important Important Evidence for head injuries and burns only Uncommon

Very important Important High or normal acid output Unimportant Important

Signs Tenderness Presence of an epigastric mass Succussion splash Supraclavicular lymphadenopathy Hepatomegaly

Bile retlux Smoking Stress Family history

15

Common

Investigations Endoscopy Barium meal Gastric acid secretory studies Plasma gastrin Full blood count Liver function tests

of investigation. A barium meal is reserved for patients who refuse endoscopy or who cannot swallow the endoscope. The clinical indications for measuring gastric acid secretion are to exclude a gastrinoma (when the basal acid secretion is very high) and to confirm that parietal cell innervation has been lost after a surgical vagotomy (if successful, insulin-induced hypoglycaemia no longer causes acid secretion).

PEPTIC ULCERS 1 Peptic ulcers can occur in the oesophagus (usually in association with gastro-oesophageal reflux), the stomach, the pyloric canal and the duodenum. An erosion is present when the mucosal surface is broken; ulcers are present when the muscularis mucosae is penetrated. In practice, it may be difficult to tell the difference at endoscopy. Lesions may vary in size from 1 mm to several centimetres in diameter.

Aetiology The aetiological factors in peptic ulceration are shown in Table 15.20. The aphorism 'no acid: no ulcer' is true for benign peptic ulcers. Gastric ulcers, however, are usually associated with normal or low acid outputs, although this may partly reflect duodenogastric reflux of alkaline juice. The bacterium Helicobacter pylori is present in the gastric antrum of more than 90% of duodenal ulcer patients and 75% of gastric ulcer patients, and its continued presence is related to a high rate of recurrence (90% of duodenal ulcer patients treated with H2-receptor antagonists or proton

FIG. 15.12 Incidence of peptic ulcer related to age, in the UK

pump inhibitors relapse in the first year after stopping treatment). H. pylori may also have an aetiological role in gastric ulcers. Duodenogastric reflux of alkaline juice, bile and pancreatic enzymes is thought to be important for gastric ulcers. NSAIDs and aspirin are associated with an increased incidence of peptic ulcers and of the complications of bleeding and perforation. These drugs are a particular problem in patients aged over 65. Smoking plays a role in pathogenesis and delays the healing of duodenal ulcers. The objective evidence that ordinary stress causes ulcers is poor, but severe head injuries (Cushing's ulcer) and burns (Curling's ulcer) are associated with the development of acute ulcers. Duodenal ulcers are often familial and are associated with blood group O. The factors involved in mucosal cell resistance or liability to ulceration are not understood. Other possibly important factors are gastric mucus, and gastric and duodenal bicarbonate secretion, which may have protective roles. An additional, and rare, cause of peptic ulceration is gastrinoma (p. 781). Duodenal ulcers are approximately four times as common as gastric ulcers. Whereas there appears to be a tendency for spontaneous remission with duodenal ulcers, the prevalence of gastric ulcer progressively increases with age (Fig. 15.12).

777

Clinical features The symptoms and signs of ulcer disease are non-specific for each type of ulcer and commonly overlap those found in patients with cholelithiasis, oesophagitis, pancreatitis, gastric cancer and no demonstrable organic disease (nonulcer dyspepsia). Duodenal ulcer patients tend to have epigastric pain, which is worse with fasting, eased by antacids and food, and wakes them at 2-4 am. Gastric ulcer patients' epigastric pain is sometimes worse with food and relieved by vomiting. Nausea and vomiting are relatively unusual and imply outlet obstruction. Anorexia and weight loss can occur with both gastric and duodenal ulcers. Ulcers may be asymptomatic. In an uncomplicated peptic ulcer patient, epigastric tenderness is usually the only abnormal sign, and even this is often absent. If there is gastric outflow obstruction, a succussion splash may be present.

Investigation (see Fig. 15.13) Endoscopy is the procedure of choice in the diagnosis of peptic ulcers (Table 15.21 and Fig. 15.14). 1 It also enables accurate histological diagnosis of malignancy, and local treatment with injection, electrocautery or lasers as necessary. Endoscopic complications are rare, but bleeding, perforation, pulmonary aspiration, transmission of infection and excessive sedation can occur. Table 15.22 shows the methods for the diagnosis of H. pylori, which contains the enzyme urease.

Complications There are three main complications of peptic ulcers: • Gastrointestinal bleeding (p. 772) is the first symptom of a peptic ulcer in a small proportion of patients, but usually there is a variable period of preceding dyspepsia. • Perforation of an ulcer usually presents with acute severe upper abdominal pain, and circulatory collapse may rapidly develop. Specific symptoms may be minimal or absent in the elderly and in patients on corticosteroid therapy. After resuscitation, surgery should be performed to clean the peritoneum and oversew the ulcer; a procedure may also be performed to prevent further ulcers, e.g. a vagotomy with or without a drainage procedure. Rarely, surgery may be contraindicated by high anaesthetic risks. Under these circumstances the perforation will sometimes heal with conservative management. • Gastric outflow obstruction (pyloric stenosis) usually

1

778

Figs 15.13-15.16

RECENT ADVANCES IN HELICOBACTER PYLORI The major advance in gastroenterology in recent years is the understanding of the role of H. pylori in gastroduodenal disease. More than 90% of chronic duodenal ulcers and 75% of chronic gastric ulcers are associated with H. pylori antral gastritis; healing of the ulcer and eradication of the bacterium leads to very prolonged remission, if not permanent cure. Three-quarters of NSAID-associated peptic ulcers have H. pylori gastritis, but the role of the bacterium in NSAID ulcers is unclear. There is no evidence that H. pylori promotes gastro-oesophageal reflux disease. Chronic H. pylori gastritis may lead to intestinal metaplastic change in some individuals. This epithelium has malignant potential and has led to the hypothesis that H. pylori is a causative factor for gastric carcinoma. Most patients with gastric carcinoma have evidence of previous exposure to H. pylori, but definite evidence of a link is so far lacking. More than 90% of patients with gastric MALT lymphoma (B-cell lymphoma) have H. pylori in their stomachs. It has been suggested that MALT (mucosa-associated lymphoid tissue) develops in the stomach in response to H. pylori infection. There are reports that low-grade MALT lymphomas regress after H. pylori eradication without any other therapy. The complete genome sequence of a strain of H. pylori was reported in 1997. It has a circular genome of 1667867 base pairs and 1590 predicted coding sequences. It has well-developed systems for motility, for scavenging iron and for DNA restriction and modification. Many adhesins, lipoproteins and other membrane proteins showed the potential complexity of host-pathogen interaction. It was thought that the bacterium used recombination and slipped strand mispairing within repeats as mechanisms for antigenic variation and adaptive evolution. Its survival in acid depended in part on its ability to establish a positive inside membrane potential in low pH. Major implications of this discovery for clinical medicine include the ability to predict many different functions, the potential to develop strategies for new antimicrobial agents, and an insight into a range of potential vaccine candidates. The vaccine and new agents will then be tested in animals and, if appropriate, assessed in clinical trials.

follows long-standing duodenal ulcer or pyloric channel ulcer, which leads to oedema and scarring. Patients complain of vomiting food eaten many hours before. The failure of gastric emptying results in postprandial fullness, anorexia and weight loss. A succussion splash may be heard many hours after the last meal. If there is an

15

FIG. 15.13 Management plan for ulcer-like dyspepsia

TABLE 15.21 Relative merits of endoscopy and barium studies in the diagnosis of peptic ulcer Merit

Endoscopy

Barium study

Accuracy

Almost 100%

Ability to obtain biopsies and brush cytology Detection of H. pylori (by urease test) Enables other treatments (e.g. diathermy, lasers) Detection of bleeding Assessment of healing Complications

Yes

Lower, especially for superficial lesions No

Yes

No

Yes

No

Good Good Very few

Poor Poor for duodenal ulcers Very few

FIG. 15.14 Barium studies and endoscopy in diagnosis of peptic ulcer H Barium meal demonstrating a large gastric ulcer filled with barium. B Typical duodenal ulcer with pyloric stenosis. C Endoscopic appearance of a chronic prepyloric ulcer.

779

active ulcer with oedema, pyloric stenosis may respond to medical treatment alone. Otherwise, endoscopic balloon dilatation followed by proton pump inhibitor therapy and H. pylori eradication, or surgery can be considered.

Management Medical Before a definite diagnosis is made, patients can be treated with simple antacids. In standard doses these are effective symptomatically, but rarely heal ulcers. When a diagnosis is made, patients should be encouraged to give up smoking as this will improve their chance of healing and reduce relapse rate. Minimization of alcohol intake should also be encouraged. NSAIDs and aspirin should be avoided. Many effective medical treatments are now available (Table 15.23), but proton pump inhibitors are the most efficaceous drugs in terms of healing rates, speed of healing, symptom relief and speed of symptom relief. They also have anti-Helicobacter efficacy. In H. pylori-positive duodenal ulcer and gastric ulcer patients the organism should be eradicated. It is not known which is the best regimen, but regimens eradicating H. pylori in 80-90% of patients are shown in Table 15.24. If the bacterium is eradicated, ulcer recurrence is rare. More effective treatments with higher eradication rates are being sought. In H. pyloriassociated peptic ulcers the success of eradication should be confirmed by a urea breath test. For H. pylori-negative ulcers, H2-receptor or proton pump inhibitors are used and may sometimes be required for maintenance treatment. Gastric ulcers should be followed up by endoscopy at 6 weeks to check healing, and by rebiopsy to exclude a tumour if unhealed; repeat endoscopy is not necessary in duodenal ulcers. Surgical Surgery is now rarely needed for uncomplicated peptic ulcers. The indications for surgery for peptic ulcers are shown in Table 15.25. Severe haemorrhage and perforation are indications for surgery unless the patient is thought unfit for a general anaesthetic. Gastric outflow obstruction also often requires surgery, although occasionally medical treatment of an active ulcer may improve symptoms. The best current operations may be highly selective vagotomy for duodenal ulcers and Billroth I partial gastrectomy for gastric ulcers.

Complications of gastric surgery

780

Surgery for gastric and duodenal ulcers is usually successful, but in 10% of patients who have a highly selective vagotomy for a duodenal ulcer the ulcer may recur. Postoperative weight loss is a common symptom but usually

stabilizes. If a partial gastrectomy is performed, bile reflux may result and cause gastritis in the remnant. Following gastric surgery, gastric emptying into the small intestine may become very rapid. Hyperosmolar contents cause fluid to be secreted into the jejunum, the plasma compartment contracts and the haematocrit rises. This results in 'early dumping', and about 30 minutes after a meal the patient complains of sweating, palpitations, faintness and abdominal distension. These symptoms may be mediated in part by the vasomotor peptides neurotensin and vasoactive intestinal polypeptide (VIP, p. 757). The rapid emptying also results in increased early insulin secretion, so that reactive hypoglycaemia may occur 3^4 hours after a meal ('late dumping'). As a result of the

RECENT ADVANCES IN NSAIDs AND THE STOMACH AND DUODENUM

NSAIDs may cause gastric and duodenal erosions, haemorrhage and peptic ulcers. Healing is best achieved by proton pump inhibitors. The rate and success of healing of NSAID ulcers is not impaired by continuing NSAID treatment. This is not true for H2receptor antagonist treatment. Prophylactic treatment is indicated for patients who require NSAIDs but have a history of peptic ulcer, severe dyspepsia on upper GI haemorrhage. Trial data now suggest that PPIs are the most effective treatment for primary and secondary prophylaxis. The widely used misoprostil is less effective in preventing duodenal injury and frequently causes diarrhoea. There are two isoforms of the enzyme cyclooxygenase (COX). COX 1 is beneficial because it promotes the production of cytoprotective prostaglandin E2, thromboxane and prostacyclin from arachidonic acid in response to cell injury. COX 2 is harmful because it promotes the production from arachidonic acid in response to cell injury of mediators of inflammation, e.g. certain prostaglandins which are harmful to the gastroduodenum. NSAIDs block both COX 1 and COX 2. It is now realized that selective inhibition of COX 2 leads to reduced NSAID-induced gastroduodenal damage. Some NSAIDs (e.g. etodolac and meloxicam) are relatively COX 2 specific, but COX 2 inhibitors with no COX 1 effects are now available (e.g. refecoxib). Initial studies are encouraging in terms of their lack of GI sideeffects. Three-quarters of patients with NSAID-associated peptic ulcers have H. pylori in their stomachs. The relative roles of the NSAID and the bacterium remain unclear, but H. pylori should probably always be eradicated in NSAID ulcers. Nevetheless, it should be assumed that continued NSAID treatment in spite of H. pylori eradication may lead to further ulceration.

15

TABLE 15.22 Tests for diagnosis of Helicobacter pylori Test

Advantages

Disadvantages

Accuracy

Indirect tests (endoscopy not required) Helicobacter antibodies (IgG)

Single blood sample

++

14/13C urea breath test

Best test for checking eradication

Laboratory assay only as commercial bedside kits are inaccurate Remain positive at least 6 months after eradication Expensive technology

Direct (invasive) tests on gastric antral mucosa Rapid urease (CLO) test

Result in 2 hours

Can miss patchy infection High false negative rate 15% False positives due to other bacteria Can miss patchy infection Relies on pathologisfs skill and interest Can miss patchy infection Slow result

Histology Culture of bacteria in urea containing broth

+++

Antibiotic sensitivities can be determined

TABLE 15.23 Important drugs used In the treatment of peptic ulcer Healing rates Drug Antisecretory drugs H2-receptor antagonists (e.g. ranitidine 150mgb.d.: cimetidine400mgb.d.) Proton pump inhibitors e.g. omeprazole 20mgb.d., lansoprazoie30mgb.d.) Antimuscarinics* (e.g. pirenzepine 150mgb.d.) Prostaglandin analogues* (also cytoprotective) (e.g. misoprostil400ngb.d.)

DU

GU

75% (4 weeks) 60% 90% (8 weeks) 80% 95% (4 weeks) 80% 90%

Main side-effects (4 weeks) (8 weeks) (8 weeks) (8 weeks)

similar to H2-receptor antagonists

Rash, enzyme induction (cimetidine), antiandrogen effects (cimetidine) Diarrhoea, rash, headache Dry mouth Diarrhoea, abdominal pain, risk of abortion

Acid-independent therapies (?mucosal protective) Sucralfate+(lgq.d.s.) Colloidal bismuth* (e.g. Denol 240mgb.d.) Carbenoxolone* (20mgq.d.s.)

similar to H2-receptor antagonists

Constipation Black tongue and stools, bad taste in mouth Sodium retention causing hypertension and oedema

* Largely obsolete. tUsed to protect stomach in ITU. *Can be used as second-line H. pylori therapy.

rapid gastric emptying, there may be small bowel malabsorption, causing diarrhoea associated with raised faecal fat excretion (steatorrhoea). All these problems can be improved by small, regular and relatively iso-osmolar meals. After both partial and total gastrectomy some patients become deficient in iron, vitamin B12, calcium and vitamin D. A full blood count and plasma calcium, phosphate and alkaline phosphatase should be checked regularly, and lifelong vitamin supplements initiated as necessary. Adeno-carcinoma of the stomach is a risk after partial gastrectomy, particularly after 20 years.

GASTRINOMA Rarely, tumours secreting gastrin stimulate excessive gastric acid secretion, resulting in severe duodenal and upper small bowel ulceration. The primary tumours are usually in the pancreas or duodenum. Fifty per cent of tumours may be malignant, and have metastasized by the time of diagnosis. The gastrinoma syndrome (originally described by Zollinger and Ellison in 1955) is characterized by severe peptic ulceration, but 40% of patients may have diarrhoea, occasionally associated with malabsorp-

781

TABLE 15.24 Eradication treatment for Helicobacter pylori (after British National Formulary (Mf) 2001) Clarithromycin regimens (all for 7 days) Omeprazole 20mgb.d., Clarithromycin 500 mg b.d., amoxycillin 1 gb.d. (UK price £38.24) Lansoprazole 30mgb.d., Clarithromycin 500 mg b.d., amoxycillin 1 gb.d. (£35.84) Omeprazole 20mgb.d., Clarithromycin 500 mg b.d., metronidazole 400mgb.d. (£37.35) Lansoprazole 30mgb.d., Clarithromycin 500mgb.d., metronidazole 400mgb.d. (£34.95) Non-clarilhromycin regimens (all for 7 days) Omeprazole 20 mg b.d., amoxycillin 500 mg t.d.s., metronidazole 400 mg t.d.s. (£16.26) Lansoprazole SOmgb.d., amoxycillin 1 gb.d., metronidazole 400mg b.d. (£13.93) Ranitidine bismuth citrate 400 mg b.d., amoxycillin 1 gb.d., metronidazole 400 mg b.d. (£15.05) Failure of first treatment: Further course with different drugs; consider quadruple therapy for 2 weeks in resistant cases, e.g. omeprazole 40mgb.d. tetracycline 500mgq.d.s., metronidazole 400 mg t.d.s., colloidal bismuth (Denol) 1 tab q.d.s. Notes: • Highest eradication (90%) is achieved with PPI and Clarithromycin regimens. • All PPIs are probably equally good. • Metronidazole resistance is a problem in some areas. Tinidazole is an alternative. • Dual therapy is less effective than triple therapy. Triple therapy for 2 weeks may be better than for 1 week but side-effects and compliance become problems. • Tetracycline or erythromycin are alternatives for penicillin-hypersensitive patients.

782

tion of the tumour(s) is by CT scanning, arteriography and, if there is still doubt, portal venous sampling for gastrin. The tumours are usually slowly progressive, and should only be resected if they are thought to be single and nonmetastatic. The acid hypersecretion is best controlled by proton pump inhibitors (see Table 15.23).

GASTRITIS

In gastritis there are areas of inflamed red oedematous mucosa in the stomach and the diagnosis can be confirmed by biopsy. Acute Acute gastritis is commonly associated with infections, NSAIDs, alcohol or iron intake, and renal failure. It may cause dyspepsia and can present with gastric bleeding. It can be diagnosed by endoscopy, but not by a barium meal. Treatment involves removing the cause and giving simple antacids. Patients with renal disease have a poor prognosis if they bleed: prophylactic H2-receptor antagonists or proton pump inhibitors are therefore used in those with dyspeptic symptoms and/or inflammatory changes on endoscopy. Chronic Chronic gastritis may not give rise to symptoms, although occasionally it may be associated with dyspepsia. There are two types: chronic superficial and chronic atrophic. The former is usually associated with H. pylori colonization of the stomach and the latter may lead to vitamin B12 malabsorption. There is an increased incidence of gastric adenocarcinoma if atrophy or metaplasia is present.

TABLE 15.25 Indications for surgery for peptic ulcers

MENETRIERE'S DISEASE

• • • • • •

Menetriere's disease is a rare condition in which the normal gastric rugae are increased in height to 1-2 cm. The aetiology is unknown. The main problem is gastric protein loss leading to peripheral oedema (see Table 15.4). Reduced acid secretion and iron and vitamin B12 deficiency occur; occasionally a gastric ulcer develops. In some patients the condition may improve with a highprotein diet, antacid drugs or fibrinolysis inhibitors. There is a tendency to spontaneous remission. In some patients, only total gastrectomy relieves the problem. There is an increased risk of developing gastric adenocarcinoma.

Severe haemorrhage Perforation Gastric outflow obstruction Failure of healing on medical treatment (now rare) Repeated relapses despite treatment (rare if H. pylori is eradicated) Non-compliance with medical treatment

tion. The duodenal ulcers are often multiple and may be associated with perforation, bleeding and pyloric stenosis. The acid may also reflux into the oesophagus and cause oesophagitis and oesophageal strictures. About a third of the patients have a functioning parathyroid tumour causing hypercalcaemia, and adenomas also occur in the pituitary, adrenal and thyroid (multiple endocrine neoplasia type 1, see p. 833). Diagnosis is by demonstrating gastric hyperacidity and a high fasting plasma gastrin. Localiza-

DUODENITIS In duodenitis the mucosa appears inflamed, reddened and sometimes haemorrhagic. It may be seen in isolation or around ulcer craters and may follow or precede the devel-

opment of duodenal ulcers. When no break in the mucosa (i.e. ulcer) is present, histology shows acute and chronic inflammation. The relationship of duodenitis to symptoms is variable: sometimes patients are asymptomatic and sometimes they have typical ulcer symptoms. Treatment is by eradication of H. pylori, if present, and by avoiding other potential aetiological factors such as smoking, alcohol and NSAIDs. If symptoms are severe, H2receptor antagonists or proton pump inhibitors may help.

NON-ULCER DYSPEPSIA The differential diagnosis of epigastric pain is wide and is shown in Table 15.26. Non-ulcer dyspepsia is a functional disorder and is sometimes related to the irritable bowel syndrome. It is characterized by the symptoms of dyspepsia associated with negative investigations, especially endoscopy. Possible causes of non-ulcer dyspepsia can be divided into non-motility and motility associated (Table 15.27). Some patients have a history of peptic ulcer or duodenitis. Others may be found to have Helicobacter gastritis with a positive urease test. Many stressed patients develop dyspeptic symptoms (as others develop headache or diarrhoea), and hints to a psychosomatic aetiology are usually obvious from the history and examination. Non-erosive gastro-oesophageal reflux can be confirmed by the correlation between symptoms and episodes of acid or bile reflux during 24-hour oesophageal pH monitoring. Idiopathic gastroparesis is a rare cause of upper abdominal bloating and early satiety. Small intestinal and biliary tract motility disorders are causes of abdominal pain. Non-ulcer dyspepsia should be used as a positive diagnosis and an attempt should be made to define its cause. All patients should be reassured that there is no serious alternative pathology. Gastric acid suppression drugs have been shown to be effective in patients with ulcer-type reflux problems. H. pylori eradication helps a large proportion of patients with ulcer-type symptoms associated with the organism. Prokinetic agents such as metoclopramide and domperidone are particularly helpful in patients with motility symptoms, such as bloating and early satiety. Many patients with psychosomatic problems settle with reassurance and modification of lifestyle, but others require psychotropic medication such as tricyclic antidepressants and selective serotonin reuptake inhibitors.

MALIGNANT TUMOURS OF THE STOMACH: ADENOCARCINOMA There are marked geographic variations in the incidence of adenocarcinoma (Fig. 15.15). In the UK the incidence is 20 per 100000 for men and 10 per 100000 for women. The incidence of body and antral carcinomas is declining, but that for adenocarcinomas of the cardia is increasing in

TABLE 15.26 Differential diagnosis of recurring epigastric pain

15

Gastro-oesophageal reflux Peptic ulcer disease Gallstones Gastric cancer Pancreatitis (acute and chronic) Other cancer (e.g. pancreas, metastases) Drugs (e.g. aspirin, NSAIDs) Systemic disorders (e.g. hypercalcaemia) Gastroparesis Carbohydrate malabsorption (e.g. lactose) Parasites (e.g. Giardia) Non-ulcer dyspepsia

TABLE 11,27 Potential causes of non-ulcer dyspepsia Non-motility Peptic ulcer diathesis Gastritis Helicobacter pylori Bile reflux Duodenitis Augmented perception of visceral pain Psychosomatic Motility Non-erosive gastro-oesophageal reflux Idiopathic gastroparesis Small intestinal dysmotility Biliary tract dyskinesia Psychosomatic

western countries, in parallel with oesophageal adenocarcinoma. In England and Wales there were 5300 deaths each year in 1995-1997.

Aetiology and pathology Gastric carcinoma is five times more common in patients with pernicious anaemia. H. pylori is also found more frequently in patients with body and antral tumours. A common alternative is a polypoidal tumour with areas of necrosis. Other risk factors include intestinal metaplasia in the stomach, achlorhydria, previous partial gastrectomy, gastric adenomatous polyps and blood group A. It has been postulated that gastric atrophy leads to a fall in acid production, which allows colonization of the stomach with bacteria. The result of bacterial growth may be to convert dietary nitrates to nitrites, which then react with amino

783

acids in food to form 7V-nitroso compounds. These have been shown to be carcinogens in animals. The typical tumour is a diffuse infiltrating carcinoma of variable size with a rolled edge and a central ulcer. It spreads through the gastric wall, invades lymph nodes and metastasizes to the liver via the bloodstream. It may also seed into the peritoneal cavity, giving rise to abdominal and pelvic tumours and ascites. Less commonly, the tumour infiltrates diffusely through the stomach wall (linitis plastica, or 'leather bottle stomach'). Microscopically the carcinomas may resemble intestinal cells or may be diffusely infiltrating tumours. The 'intestinal' type of tumour has a better prognosis and is more frequent in areas of high incidence, such as Japan. In Japan, screening for gastric carcinoma has led to the detection of early-stage tumours limited to the mucosa or submucosa; such tumours have a good prognosis. Adenocarcinoma of the gastric cardia may be related to gastro-oesophageal reflux and the development of short segments of Barrett's oesophagus, or to H. pylori'-induced carditis.

Clinical features Gastric carcinoma is commonest in 55-65 year olds and is rare at less than 30 years of age. Typically, the patient presents with anorexia, nausea, early satiety and vague abdominal discomfort. More severe epigastric pain may develop later. Bleeding is common, leading to iron deficiency and, occasionally, frank haematemesis or melaena. There may be dysphagia if the tumour involves the gastrooesophageal junction, and vomiting if there is outflow obstruction. Non-metastatic paraneoplastic manifestations (p. 152) are unusual and include acanthosis nigricans (of which it is the commonest cause; p. 420), venous thrombosis and dermatomyositis. Physical examination may reveal evidence of weight loss, anaemia and an epigastric mass. Enlarged lymph nodes may be palpable in the supraclavicular region.

Investigation

FIG. 15.15 Gastric adenocarcinoma: incidence and mortality A Worldwide incidence per 100000. B Gastric adenocarcinoma mortality related to social class. Figures are for men aged 15-64 in England and Wales. Values for each social class are in proportion to 100, which represents all social classes combined. C Age-adjusted incidence of gastric carcinoma in England.

1

784

Fig. 15.17

The diagnosis should be suspected when dyspepsia appears for the first time in later life, and is usually made at endoscopy when an ulcer is found. It is safest to assume that every gastric ulcer is malignant until proved otherwise. Diffusely infiltrating tumours of the linitis plastica type may be missed at endoscopy and can be more easily diagnosed on a barium meal, when the stomach appears contracted and non-distensible. On barium meal the ulcerative neoplasms appear as an ulcer with irregular borders and disruption of the normal mucosal folds (Fig. 15.16). If there are no clinically evident metastases and surgery is a possible treatment, it is essential to stage the tumour further. This will include chest X-ray, liver function tests and abdominal imaging with liver ultrasound, CT or MRI and possibly endoscopic ultrasound. Tumour markers are not helpful in diagnosis 1 or management.

abdominal discomfort, nausea or gastrointestinal bleeding, and an ulcer is found on endoscopy. The histological appearances may be misinterpreted as inflammatory infiltrate and the diagnosis established only later when the disease is more florid. Gastric lymphoma is a tumour of B lymphocytes (centrocytes and centroblasts, see Ch. 23). Surgical resection may be curative in the early stages, but the prognosis with surgery alone is worse if local nodes are involved. These patients should receive combination chemotherapy following surgery. B-cell lymphomas in mucosa-associated lymphoid tissue (MALToma) may be caused by H. pylori infection and successfully treated by Helicobacter eradication alone.

15

BENIGN TUMOURS OF THE STOMACH Gastric leiomyomas Gastric leiomyomas are benign submucosal tumours which rarely become sarcomatous. They arise from the muscularis mucosae. The aetiology is unknown. In many patients they are asymptomatic, but in others may present with nonspecific features such as epigastric pain. The centre of the tumour often ulcerates and patients can present with haematemesis and melaena. In this situation, surgical resection is indicated. FIG. 15.16 Barium meal examination showing typical gastric carcinoma

Management Surgical resection is the only curative treatment for adenocarcinoma, and a total or partial gastrectomy is performed. The prognosis is directly related to stage, being good (5-year survival 55%) when the tumour is confined to the mucosa or submucosa. Only 2.5% of tumours are in this category. If the tumour has spread through the stomach wall (but no further) the survival falls to 25%; if the lymph nodes are involved few patients will survive 5 years. Overall, only 60% of patients undergoing surgery have respectable tumours, and of these only 20% will live 5 years. Chemotherapy may sometimes be helpful in palliating symptoms of metastatic disease. Radiotherapy may be helpful in the palliation of symptoms resulting from local recurrence.

MALIGNANT TUMOURS OF THE STOMACH: GASTRIC LYMPHOMA Although uncommon (less than 3% of neoplasms), gastric lymphoma is of interest because it remains localized for a considerable period and the prognosis is much better than that of gastric carcinoma. It typically presents with upper

Gastric polyps Gastric polyps are a common incidental finding at endoscopy and should always be biopsied. Hyperplastic polyps are not associated with specific symptoms, have no malignant potential, and can therefore be ignored. Adenomatous polyps may be premalignant and should be removed endoscopically or surgically. They are usually asymptomatic but can be associated with dyspepsia, gastrointestinal bleeding or gastric outflow obstruction.

FURTHER READING ON THE STOMACH AND DUODENUM Fisher R S, Parkman H P 1998 Management of nonulcer dyspepsia. N Engl J Med 339:1376-1381. Langman M J S, Weil J, Wainwright P et al 1994 Risk of bleeding peptic ulcer associated with individual non-steroidal antiinflammatory drugs. Lancet 343:1075-1078. Lau J Y W et al 1999 Endoscopic re treatment compared with surgery in patients with recurrent bleeding after initial endoscopic control of bleeding ulcers. N Engl J Med 340:751-756. Palmer K R, Choudari C P 1995 Endoscopic intervention in bleeding peptic ulcer. Gut 37:161-164. Rockall T A, Logan R F A, Devlin H B, Northfield T C 1995 Incidence of and mortality from upper gastrointestinal haemorrhage in the United Kingdom. Br Med J 311:222-226. Sayer J M, Long R G 1993 A perspective on iron deficiency anaemia. Gut 34:1297-1299. Thijs J C, Kuipers E J, van Zwet A A, Pena A S, de Graaff J 1995 Treatment of Helicobacter pylori infections. Q J Med 88:369-389. Tomb J-F et al 1997 The complete genome sequence of the gastric pathogen Helicobacter pylori. Nature 388:539-547.

785

Wolfe M M, Lichtenstein D R, Singh G 1999 Gastrointestinal toxicity of nonsteroidal anti-inflammatory drugs. N Engl J Med 340:1888-1899.

THE SMALL INTESTINE The small intestine is the prime site of nutrient digestion and absorption. It is also a site of water and electrolyte secretion and consequent marked fluid fluxes.

Clinical features of small bowel disease The main symptoms and signs of small bowel diseases are shown in Table 15.28. The faeces must be examined for colour, consistency and amount, and a sigmoidoscopy performed to exclude rectal disease.

Investigation of small bowel disease Investigation of systemic effects The haemoglobin, blood film and serum vitamin B!2, folate and iron assess haematinic deficiency and hyposplenism. Raised white cell count, C-reactive protein or ESR are non-specific indicators of underlying inflammation. The albumin is measured to look for protein-losing enteropathy and nitrogen deficiency. The plasma urea and electrolytes are measured to assess renal function and hypokalaemia. A low plasma calcium usually indicates calcium deficiency, which is due in part to malabsorption of vitamin D, which is fat-soluble. This may be associated with low plasma 25-hydroxy and 1,25-dihydroxy vitamin D levels. As a consequence, there may be increased parathyroid hormone secretion (secondary hyperparathyroidism) resulting in hypophosphataemia and a rise in the plasma bone alkaline phosphatase. Abnormal liver function tests are associated with the hepatic complications of Crohn's disease. With steatorrhoea, malabsorption of the other fatsoluble vitamins (A, E and K) may lead to a low plasma vitamin A, and the prothrombin time may be prolonged owing to vitamin K deficiency. Water-soluble vitamin deficiencies are rare; thiamine can be assessed by the red cell transketolase level. Faecal weight and fat can be measured in a 3-day collection. Small intestinal function tests are discussed on page 758. Structure and function Investigations are listed in Summary Box 5. Endoscopy Upper gastrointestinal endoscopy allows visualization of the duodenum and colonoscopy permits inspection of the terminal ileum. The rest of the small bowel can be seen at

1

786

MCQ 15.9

2

Case 15.3

TABLE 15.28 Clinical features of small bowel disease Symptoms Diarrhoea - watery (due to osmotic effects of malabsorbed solutes) or to electrolyte secretion or steatorrhoeic (due to fat malabsorption) Central abdominal pain Abdominal bloating Anorexia; weight loss Sore tongue; sore mouth; mouth ulcers Bone pain Easy bruising Malaise; tiredness Ankle swelling Signs Weight loss; muscle wasting (due to anorexia or malabsorption) Abdominal tenderness; abdominal distension Anaemia (due to iron, folate or B12 deficiency) Aphthous ulceration (seen in coeliac disease, Crohn's disease and tropical sprue) Glossitis, angular stomatitis (seen in iron or B vitamin deficiencies) Finger clubbing (seen in coeliac disease and inflammatory bowel disease) Ankle oedema (due to hypoalbuminaemia as a result of anorexia, malabsorption of protein or Gl protein loss) Dehydration (due to severe watery diarrhoea) Ecchymoses (due to vitamin K malabsorption) Tetany (due to hypocalcaemia) and tender bones (as a result of vitamin D malabsorption)

SUMMARY 5 Investigation of small bowel disease Investigation of systemic effects Haemoglobin and blood film White cell count and acute phase reactants Serum vitamin B12, folate and iron status Plasma proteins Urea and electrolytes Calcium, phosphate and alkaline phosphatase Liver function tests Fat-soluble vitamin deficiency (vitamin A, D, E, K) Test of small bowel structure and function Small bowel endoscopy, histology and brush border enzyme analysis Faecal weight and fat content Absorption tests Triple-lumen tube tests for absorption and secretion Breath tests Radiology

Small bowel enema Small bowel meal Arteriography

Radioisotopes

Radiolabelled albumin for protein loss Radiolabelled white cells for sites of inflammation Radiolabelled red cells for sites of bleeding

enteroscopy, which is, however, time-consuming and rarely indicated. Radiology The small bowel can be visualized by a small bowel meal (patient drinks barium) or a tube small bowel enema (dilute barium is infused directly into the upper jejunum via a peroral tube). The former is adequate for diagnosing most diffuse mucosal disorders, but the latter is superior for demonstrating focal lesions and examining the terminal ileum. Arteriography is used to assess ischaemia and demonstrate bleeding sites. The small bowel cannot be satisfactorily imaged by ultrasound, CT or MRI. Radioisotopes Radiolabelled albumin is used to assess protein-losing enteropathy, and radiolabelled (e.g. indium) white cells are used to indicate the site of active inflammation in patients with Crohn's disease. Chromium-51 (51Cr)-labelled red cells can be used to indicate the site of small intestinal bleeding, if this is more than 1 mL blood per minute. Bile acid and vitamin B12 absorption can also be measured by the use of radioisotope (p. 759).

MALABSORPTION A simple classification of small intestinal diarrhoea is shown in Table 15.29. Osmotic diarrhoea occurs when the

luminal contents are osmotically active and draw excessive amounts of water into the small bowel. Secretory diarrhoea occurs when small intestinal water and electrolyte secretion is increased, leading to fluid loss into the gut lumen, which overcomes absorptive capacity. A number of drugs can cause malabsorption. The most severe steatorrhoea (more than 100 g fat/24 h) is seen in patients with very short small intestines or in pancreatic exocrine failure. The amount of fat malabsorption in most patients with small intestinal malabsorption is less than 40g/24h. Steatorrhoea is minimal or absent in patients with osmotic or secretory diarrhoea. Large volumes of stool (more than 1 kg/24 h) are usually seen in patients with secretory diarrhoea, and occasionally in patients with osmotic diarrhoea and in those with substantially reduced lengths of functional small intestine.

COELIAC DISEASE

15

12

Coeliac disease is also known as gluten-sensitive enteropathy or coeliac sprue. It is a genetically determined disease which requires exposure to an environmental factor (dietary gluten) for its clinical expression. The disorder results in abnormal small bowel mucosa, with loss of villi, crypt hyperplasia and infiltration with chronic inflammatory cells. The clinical syndrome and small bowel

TABLE 15,29 Classification of small intestinal diarrhoea Abnormality

Causes

Steatorrhoea

Watery stool

Other results

Upper small bowel disease

Coeliac disease Giardia etc.

+

+

Iron Folate Calcium

Osmotic diarrhoea

Disaccharidase deficiencies Magnesium salts

-

++

Secretory diarrhoea

Cholera VIP Phenolphthalein

Rapid transit

Postgastrectomy Thyrotoxicosis

+

+

Short gut

Surgery Fistula

++

++

llal dysfunction and resection

Crohn's disease Terminal ileal resection Tuberculosis

+

+

Drug-induced

Neomycin Cholestyramine Alcohol

+

±

+++

Dehydration Hypokalaemia

Vitamin B12 Bile acids

787

histology improve with the removal of dietary gluten, and return if gluten is eaten again.

Aetiology The aetiology of coeliac disease became clear at the end of World War II, when it was realized that Dutch coeliac children improved when wheat was unavailable (they subsisted on tulips and sugar-beet) but deteriorated when wheat was reintroduced into their diets. Wheat, rye, and barley glutens are toxic. During milling, bran is separated from flour. Gluten is contained in the water-insoluble portion of flour; the toxic moiety is present in an alcoholsoluble portion of gluten known as gliadin. Digestion of gliadin with pepsin and trypsin results in smaller toxic subunits. The mechanism of gluten toxicity is uncertain but past theories, such as missing intestinal peptidase, an enterocyte membrane defect and cross-reactivity with a protein produced by cells infected with adenovirus 12, are untenable. Evidence suggests that coeliac disease is a T-cell mediated enteropathy in which gluten (gliadin) activates CD 4+ Tcell lymphocytes in the lamina propria of the small intestinal mucosa. These lymphocytes damage the small intestinal mucosa by immunogenetic mechanisms related to the expression of HLA DQ2 or DQ8 molecules on immune cells. Moreover, there is evidence to suggest abnormal expression of these molecules on coeliac enterocytes, converting them into antigen-presenting cells. Ninety per cent of coeliacs have the haplotype of a variant of HLA DQ2, which is encoded by allelles HLA DQ A 1*0501 and HLA DQ B 1*0201. The remaining 10% have the haplotype HLA DQ8, which is encoded by allelles HLA DQ Al*0301 and HLA DQ Bl *0302. These allelles are located in the MHC locus on the short arm of chromosome 6. There is non-concordance for coeliac disease in some non-identical twins with identical HLA DQ haplotypes, suggesting that a non-HLA gene for susceptibility to coeliac disease must exist. The highest recorded incidence of coeliac disease is in the west of Ireland, where it is 1 in 300. In the UK the incidence is about 1 in 2000. The disease is very unusual in nonCaucasians. There is evidence that the incidence in children is falling in the UK, and this may reflect the fact that most infant feeds are now gluten free. There is a modest female preponderance. About 10% of asymptomatic first-degree relatives have the disorder.

Clinical features

788

Coeliac disease can present at any age, but there are peaks between 1 and 5 years of age and in the third and fourth decades. Typically, the disease is now diagnosed in patients found to be iron or folate deficient during investigations for non-specific symptoms. The extreme case of a cachectic, oedematous, clubbed, pigmented, dying patient with multiple vitamin deficiencies is now very rare.

The main symptoms and signs are shown in Table 15.30. In children the symptoms are often non-specific, with failure to thrive in infancy or growth retardation. Presentation in adolescence is unusual. Diarrhoea is more common in adults than in children, but is not universal and coeliacs may rarely present with constipation. Abdominal pain is usually mild. About 30% of patients have aphthous ulcers, which are usually severe and recurrent but respond to a gluten-free diet. Dermatitis herpetiformis is an itchy vesicular rash which is usually predominant on extensor surfaces such as the elbows and knees (p. 790). Coeliac disease is associated with other immune-mediated disorders such as insulin-dependent diabetes mellitus and thyroid disease. Examination is often non-contributory, with mild anaemia, weight loss, abdominal distension and diarrhoea

RECENT ADVANCES IN COELIAC DISEASE IgA class endomysial antibodies are present in the serum of 95% of adults and 100% of children with untreated coeliac disease. They are highly specific for coeliac disease and are useful for screening patients. On a strict gluten-free diet they disappear, and can be used to monitor response to the diet. The presence of these antibodies in individuals with minor small bowel mucosal abnormalities or with normal small bowel mucosa has led to the view that coeliac patients do not necessarily have severe small bowel mucosal damage. Endomysial antibodies react against tissue transglutaminase (TGase). Gliadin ingestion in coeliac patients causes gut inflammation, which leads to increased activity of TGase in small bowel mucosa. TGase effects deamidization of gliadins, which results in an epitope that binds to DQ2 and is recognized by gut-derived T cells. Therefore, TGase makes gliadin more efficiently recognized by DQ2-restricted gliadin-specific T cells, thereby modulating the activity of these cells. It is of interest that gut infections cause increased TGase activity, so could infection be a predisposing factor to the onset of coeliac disease in a genetically predisposed individual? The question arises, are endomysial antibodies involved in the aetiology of coeliac disease, or are they epiphenomena? They might appear in response to increased activity of TGase, itself a response to gliadininduced gut inflammation, and therefore be an epiphenomenon. However, the interaction between mesenchymal cells and gut epithelium is mediated by TGase and is important for maintaining normal small intestinal mucosal architecture. It is possible that antibodies to TGase interfere with this process and could therefore be a factor in the mucosal damage seen in untreated coeliac disease.

TABLE 15.30 Clinical manifestations of coeliac disease Symptoms/signs History Non-specific, e.g. malaise, anorexia Diarrhoea and steatorrhoea Weight loss Abdominal pain and bloating Aphthous ulcers Nausea and vomiting Itchy spots on extensor surfaces (dermatitis herpetiformis) Positive family history Irish background Oedema Pigmentation Tetany Bone pains; muscle weakness Bleeding tendency Night blindness In children Failure to thrive Irritability Growth failure Delayed puberty Examination Anaemia Weight loss Pigmentation Glossitis Angular cheilitis Aphthous ulcers Koilonychia Clubbing Oedema Bruising Abdominal distension Positive Chvostek's and Trousseau's signs Bone tenderness; proximal myopathy Reduced light adaptation Dermatitis herpetiformis

Mechanism Iron, folate or calcium deficiency Water, electrolyte and fat malabsorption Fat and carbohydrate malabsorption Distended bowel loops

Hypoproteinaemia

Vitamin D and calcium deficiency Vitamin K deficiency Vitamin A deficiency

Malabsorption

Iron and folate (and, rarely, vitamin B12) deficiency Malabsorption Unknown Iron and folate deficiency Gluten sensitivity and/or vitamin deficiency Iron deficiency Unknown Hypoproteinaemia Vitamin K deficiency Distended bowel loops Hypocalcaemia Osteomalacia Vitamin A deficiency

or steatorrhoea on rectal examination being the most common physical signs. Hypocalcaemia may develop relatively quickly, but bone tenderness and proximal myopathy imply long-standing hypocalcaemia with osteomalacia.

Investigations Small bowel histology The diagnosis is made by demonstrating the characteristic small bowel biopsy changes which improve on gluten withdrawal. The villi may be totally or, less commonly, partially lost (subtotal or partial villous atrophy, see Fig. 15.17). The duodenum and jejunum are predominantly affected, but in severe cases villous abnormalities may also be seen in the ileum. Subtotal villous atrophy usually indicates coeliac disease. However, in the absence of a response to a glutenfree diet and with lesser degrees of villous atrophy, conditions such as giardiasis, tropical sprue, small bowel ischaemia and gastrinoma must also be considered. Transient gluten intolerance with subtotal villous atrophy can be seen in children.

15

Haematology The most common haematological abnormality is anaemia due to iron and/or folate deficiency as a consequence of malabsorption. Folate deficiency is most reliably characterized by a low red blood cell folate, which reflects longstanding folate deficiency. Serum B12 levels are quite often low, often secondary to a defect in mucosal absorption as a result of folate deficiency. Hyposplenism is present in many patients, and Howell-Jolly bodies may be seen on the blood film. The ESR is raised in a minority of patients. Sideroblastic anaemia may occur as a result of vitamin B6 deficiency. If there is severe malabsorption of vitamin K the prothrombin time may be prolonged. Biochemistry Hypoalbuminaemia can occur and is due to protein loss from the small intestine and peptide malabsorption. Faecal fat levels are usually raised. Hypocalcaemia due to malabsorption of calcium is common, with consequent secondary hyperparathyroidism. In chronic hypocalcaemia the serum bone alkaline phosphatase rises. Some patients are vitamin D deficient, and this is best demonstrated by a low serum 25-hydroxy vitamin D. Serum vitamin A and magnesium levels may also be low. Zinc deficiency is rare. Absorption tests Absorption tests are not specific for coeliac disease. Radioisotopic fat tests can be used to screen for fat absorption and are less unpleasant for laboratory personnel, but less accurate than faecal fat estimation. Small bowel permeability tests are abnormal (p. 758). Radiology Small bowel radiology may be normal or may show nonspecific jejunal dilatation and flocculation of barium; its main value is to exclude anatomical abnormalities and inflammatory disorders (e.g. Crohn's disease). Bone radiology may show signs of osteomalacia (bone thinning and pseudofractures) and secondary hyperparathyroidism (subperiosteal erosions; see Ch. 18).

789

FIG. 15.17 Comparison of normal small bowel histology with that of a patient with coeliac disease A Photomicrograph of normal small intestinal finger-like villi with columnar surface cells. B Photomicrograph of small bowel biopsy in a patient with coeliac disease. There is total villous atrophy with complete loss of the villi, cuboidal surface cells, crypt hyperplasia and an increased number of lymphocytes.

Screening tests Whereas the diagnosis of coeliac disease requires the demonstration of villous atrophy at small bowel biopsy, some non-invasive tests can be helpful in screening patients for small bowel biopsy and in monitoring the response of coeliac patients to gluten withdrawal. IgA endomysial antibodies are present in the serum of most coeliacs (nearly 100% children, about 95% of adults) and false positives are very rare. Tests of small intestinal permeability are highly sensitive for severe villous atrophy but are not specific for coeliac disease. Antibody and permeability tests normalize as the coeliac intestine recovers on a gluten-free diet.

tional supplements. However, osteoporosis is a risk in all coeliac patients, even when there is a good response to the diet. Therefore, patients need DEXA scanning and appropriate treatment. Rarely, corticosteroids are indicated in patients who remain symptomatic and have a poor response to a strict gluten-free diet. Small bowel biopsies are repeated after 6-12 months of a gluten-free diet in all patients. Rebiopsy after a gluten challenge is required to confirm the diagnosis in children, but may not be essential in adults. For adults who are stable on a gluten-free diet an annual review is useful to check their weight and exclude deficiency states. In children, more frequent follow-up is necessary to monitor growth and development.

Management Management is with a strict gluten-free diet, avoiding flour from wheat, rye, barley and, occasionally, oats. This means the avoidance of most bread, cakes and biscuits, as well as foods where flour has been used as a thickening agent (e.g. gravy). Rice and maize can be used as substitutes. In the UK, gluten-free foods can be obtained on prescription. Such a diet allows the small bowel mucosa to return almost to normal in all children and most adults. A strict diet is essential in most cases, but in the newly diagnosed elderly patient with minor symptoms strict adherence to diet is less essential. In the UK, membership of the Coeliac Society should be encouraged. Supplements of folic acid, iron, vitamin B12, calcium, vitamin D etc. should be given to correct deficiency states. Most patients make a good response to the diet and nutri-

1

790

Figs 15.18,15.19

Development of malignancy 1 There is a greatly increased risk of developing nonHodgkin's lymphoma, which is of T-lymphocyte origin in most cases. It is usually multifocal throughout the small bowel and responds poorly to chemotherapy, with a very poor prognosis. There is also a higher incidence of adenocarcinoma of the small bowel and squamous cell carcinomas of the oesophagus and pharynx. Recent evidence suggests that a gluten-free diet may reduce the risk of lymphoma.

DERMATITIS HERPETIFORMIS Dermatitis herpetiformis is a blistering and pruritic skin eruption which particularly affects the extensor surface of the knees and elbows (p. 412). Most patients have associated small intestinal villous atrophy identical to that of coeliac disease, and may also develop other complications of coeliac disease. Histology of the skin lesions shows

blister cavities, and immunofluorescence shows IgA deposits at and just below the dermoepidermal junction (p. 411, Fig. 10.34). The rash and intestinal lesion improve with gluten withdrawal. The rash is also helped by treatment with oral dapsone.

15

BACTERIAL OVERGROWTH Aetiology Under normal circumstances the stomach and upper intestine have a low concentration of bacteria, owing to gastric acid secretion and normal small bowel motility. Greater numbers of bacteria, intermediate in amount between the jejunum and the colon, are found in the ileum. Bacterial overgrowth of the upper small bowel is associated with jejunal diverticular disease (Fig. 15.18), reduced or absent gastric acid secretion, partial gastrectomy, small bowel stasis due to strictures, systemic sclerosis, diabetic neuropathy, adhesions and fistulae from the small to the large bowel. It can also occur spontaneously in the elderly. Bacterial deconjugation and dehydroxylation of bile acids is thought to be the main mechanism of the steatorrhoea, but there is also some damage to the small intestinal mucosa.

Oral 14C glycocholate (Cholic acid + 14C glycinemam

Clinical features and investigation Patients present with weight loss, anorexia, nausea, diarrhoea (but sometimes constipation) and increased flatulence. Steatorrhoea and B12 malabsorption (resulting in red cell macrocytosis) are the commonest laboratory abnormalities. Deficiencies of iron, protein and calcium may occur. The red cell folate levels can be high because of bacterial synthesis of folic acid. The diagnosis can be confirmed by demonstrating increased numbers of bacteria in upper small bowel aspirates. This is technically difficult, however, and instead the 14 C-glycocholate breath test is often used to demonstrate early deconjugation of bile acids by bacteria (Fig. 15.18B). The diagnosis is supported by finding a high fasting breath hydrogen, which rises further after an oral glucose load. It is also important to demonstrate the cause of the overgrowth. This may be obvious from the history (such as a partial gastrectomy) or may require a small bowel enema (e.g. to show diverticular disease, systemic sclerosis or fistula formation).

Management If feasible, the underlying cause is treated, e.g. surgical correction of adhesions, fistulae and blind loops. Noncorrectable upper small bowel abnormalities, such as multiple diverticula and systemic sclerosis, have to be treated medically with broad-spectrum or anaerobe-directed antibiotics. Tetracycline, ampicillin, ciprofloxacin or metronidazole may be effective. There is a tendency for the problem to recur and cyclical antibiotic therapy can be

Disease or resection of terminal ileunm

Bacterial overgrowth in small bowel

Bile salts in colon

"C glycocholate deconjugated by bacteria in small or large bowel

"C glycine

B

"CO2in breath

FIG. 15.18 Diagnostic findings in bacterial overgrowth A Small bowel enema showing multiple small bowel diverticula. B In the normal 14C glycocholate breath test little 14C02 appears in the breath until 4-5 hours after ingestion. In the presence of small bowel bacterial overgrowth, there is early cleavage of glycine and thus an early peak of 14C02 at 1-2 hours.

indicated. If one antibiotic is ineffective another should be tried, but there is little evidence that bacterial aspirate sensitivities help in the choice of antibiotics. Deficiencies, particularly of vitamin B12, should be corrected. In severe cases there must be careful attention to nutrition, with the avoidance of dietary fats and their substitution with midchain triglycerides.

791

GIARDIASIS Giardia lamblia is a protozoon which colonizes the bowel, especially the upper small intestine, causing diarrhoea and malabsorption; it is discussed in Chapter 9 (p. 335).

TROPICAL SPRUE (POSTINFECTIVE TROPICAL MALABSORPTION) Tropical sprue is a chronic malabsorption disorder which occurs in patients who live, or have lived, in the tropics (in particular India and southeast Asia). The aetiology is uncertain, but viruses or toxin-forming coliform bacteria have been implicated in some patients. Tropical sprue results in small intestinal partial villous atrophy.

chest pain and lymphadenopathy. Intestinal symptoms such as diarrhoea, steatorrhoea, weight loss and abdominal pain occur 1-10 years later. Small bowel biopsy is diagnostic even before intestinal symptoms develop. Associated features are finger clubbing, cutaneous pigmentation, cardiac complications (such as pericarditis, myocarditis and left-sided valvular lesions) and central nervous system complications, such as depression, dementia and fits. Broad-spectrum antibiotics are rapidly effective in abolishing all the problems and are life-saving. It is generally suggested that intramuscular penicillin and streptomycin are used for 2 weeks and then replaced by tetracycline, continued for 1 year. About 30% of patients subsequently relapse, and so long-term follow-up is recommended.

LYMPHANGIECTASIA

Clinical features, investigation and management Symptoms include diarrhoea, steatorrhoea, weight loss, abdominal distension, oral ulceration with the clinical features of nutritional deficiency, and hypoproteinaemia. The diagnosis is made by demonstrating villous atrophy and malabsorption of more than one nutrient in a patient from an endemic area. If subtotal villous atrophy is present, coeliac disease must be considered. Many patients are cured by a 1-month course of oxytetracycline with folic acid and vitamin B12 supplements; any other deficiencies should be treated as necessary. The diarrhoea may be treated symptomatically with codeine phosphate or loperamide. There is a tendency to spontaneous remissions and relapses.

WHIPPLE'S DISEASE Whipple's disease is a rare multisystem disease caused by Trophyrema whippelli. The small intestine is characterized by large, foamy periodic acid-Schiff (PAS)-positive macrophages containing glycoprotein within the lamina propria. The overlying villi are distorted and upper small bowel radiology shows thickened and oedematous mucosa. The patients are often iron or folate deficient.

Primary lymphangiectasia is a rare idiopathic condition of children and young adults, associated with dilated lacteals in the villi due to hypoplasia of the lymphatics. Secondary lymphangiectasia may result from intra-abdominal malignancy or constrictive pericarditis.

Clinical features, investigation and management The most common symptom is peripheral oedema secondary to hypoproteinaemia due to protein-losing enteropathy. Diarrhoea, steatorrhoea, abdominal pain, hypocalcaemia and growth retardation occur. Protein loss leads to immunodeficiency. Diagnosis is by small bowel biopsy, which shows dilated lymphatics throughout the villi. A small bowel enema is characteristic and shows large intestinal folds, mucosal nodularity and contrast flocculation. The idiopathic variant should be treated with a low-fat diet and possibly substitution of dietary fat by mid-chain triglycerides. Occasionally only a small segment of bowel is involved, and this can be excised. Pericardectomy may result in complete resolution of the syndrome if constrictive pericarditis is the problem.

ABETALIPOPROTEINAEMIA

Clinical features, investigation and management The disease usually affects middle-aged men. Early symptoms are arthritis, fever, malaise, recurrent cough, pleuritic

1

792

Fig. 15.20

Abetalipoproteinaemia is a rare disorder, inherited as an autosomal recessive and usually presenting in childhood. Plasma betalipoproteins are absent and plasma cholesterol, triglycerides and phospholipid are low. The defect appears to be a failure of synthesis of apoprotein B, with defective transport of triglyceride from the liver and gut. The red cells are spiked acanthocytes. The patients initially have mild diarrhoea and steatorrhoea, but by 20 years of age most have developed progressive neurological deficits with peripheral neuropathy and cerebellar ataxia. A small

bowel biopsy shows cytoplasmic fatty vacuolation of the enterocytes at the apex of the villi. No treatment is known to affect the underlying problem, but a low-fat or mid-chain triglyceride diet is usually given and fat-soluble vitamin deficiencies are corrected.

SYSTEMIC SCLEROSIS 1 Systemic sclerosis (p. 1177) may affect the small bowel, causing a pseudo-obstruction syndrome with poor motility, duodenal dilatation and a distended jejunum with prominent folds. Histology shows submucosal collagen and muscle atrophy. Bacterial overgrowth often occurs. Weight loss, nausea, abdominal distension, diarrhoea and steatorrhoea occur, and colonic involvement may result in constipation. Treatment is of the nutritional deficiencies and bacterial overgrowth with antibiotics. No treatment influences the underlying disease process.

RADIATION ENTERITIS

in late childhood. Lactose is not split into its constituent monosaccharides, galactose and glucose, is not absorbed, and becomes osmotically active. Clinical features Eating dairy products (which are rich in lactose) results in loose stools, abdominal distension, borborygmi, colic and increased rectal flatus. The distension and flatus are due to colonic bacterial digestion of lactose. Symptoms may only become apparent if a colonic disorder such as ulcerative colitis develops, when there is decreased colonic fluid absorption. Most non-Caucasians have primary lactase deficiency and may suffer diarrhoea, bloating and flatulence if they eat lactose. The incidence of primary alactasia in the Caucasian population in Europe and North America is up to 20% (about 6% in the UK). Secondary lactase deficiency may also occur as a result of small bowel disease (such as coeliac disease and Crohn's disease) and during and after bowel infections. Investigation In many patients, especially where the symptoms occur following an acute infection, a response to a low-lactose diet allows a clinical diagnosis to be made. If definitive diagnosis is necessary, lactase activity is measured in a small bowel biopsy specimen. Bacterial breakdown of unabsorbed lactose in the colon results in abnormally high hydrogen excretion in the breath, so that measurement of exhaled hydrogen is an index of lactase deficiency.

Radiation enteritis usually occurs in the small intestine of patients who have undergone abdominal and pelvic irradiation. In the early phases of radiation treatment there is cell necrosis and shortening of the villi; with increasing dose, polymorphs become prominent. Oedema, collagen formation and telangiectatic blood vessels develop as longterm sequelae. During treatment, patients develop anorexia, nausea and diarrhoea; this usually settles when radiotherapy stops. Chronic complications are unusual and may follow years later. They include diarrhoea, steatorrhoea, intestinal obstruction from stricture formation, intestinal ulceration with haemorrhage or perforation, bleeding from telangiectatic vessels, fistula formation or infarction. Medical treatment of the chronic diarrhoea is unsatisfactory, but attention to diet, cholestyramine, corticosteroids and broad-spectrum antibiotics may help. Some patients require surgery for strictures but this can be hazardous.

Sucrase-isomaltase deficiency Sucrase-isomaltase deficiency is a rare enzyme deficiency which results in diarrhoea when sucrose is introduced into the diet in infancy. It responds to the elimination of sucrose, and tolerance may improve later in life. Very rarely the small intestinal glucose and galactose transport system itself is ineffective; this results in an infant with watery motions and glucose in the faeces. Treatment is by avoidance of glucose- and galactose-containing foods.

OSMOTIC DIARRHOEA

Purgative abuse

The two major causes of osmotic diarrhoea are brush border enzyme deficiencies and purgative abuse.

Brush border enzyme deficiencies Primary lactase deficiency Primary lactase deficiency is the commonest cause of osmotic diarrhoea and is an inherited deficiency of the brush border membrane enzyme, lactase. It develops

15

Management Lactose in the diet is reduced until symptoms resolve. If symptoms persist an underlying cause should be sought. The majority of patients with secondary hypolactasia can resume lactose some weeks after treatment.

The laxatives magnesium sulphate, magnesium hydroxide, magnesium carbonate and lactulose act by osmosis. Magnesium salts may be used for bowel preparation for surgery and colonoscopy, and lactulose (a synthetic disaccharide) is used for mild constipation and portosystemic encephalopathy. Patients may secretly take such drugs in excess (see also p. 823) to produce watery diarrhoea of unknown cause, but a diagnosis can be made by finding high faecal magnesium levels for the magnesium salts, and a low faecal pH for lactulose.

793

SUMMARY 6 Causes of secretory diarrhoea • Infection, e.g. cholera • Purgative abuse • Hormonal VIPoma Carcinoid syndrome Medullary carcinoma of the thyroid Gastrinoma • Bile salt malabsorption • Steatorrhoea

SECRETORY DIARRHOEA

Diarrhoea may be caused by the rapid transit of food and secretions through the small intestine. Causes include rapid gastric emptying, short bowel syndrome, enteroenteric or enterocolonic fistulae, thyrotoxicosis and the irritable bowel syndrome.

SHORT GUT

1

Under normal circumstances there is a daily flux of about 8L of water and electrolytes across the small intestinal mucosa, but only about 1L of this passes through the ileocaecal valve. Secretory diarrhoea is due to absorption being exceeded by active secretion, leading to profuse watery diarrhoea, often more than 1 kg/24 hours. The classic secretory diarrhoea is that caused by the enterotoxin of Vibrio cholerae (cholera), which stimulates adenylate cyclase. Oral rehydration solutions at up to 800mL/h are the mainstay of cholera treatment. Intestinal glucose and sodium absorption are interdependent and the solutions drive sodiumdependent glucose and water absorption. In the western world, secretory diarrhoea is usually due to a vasoactive intestinal polypeptide (VlP)-secreting tumour (p. 833) or stimulant abuse, especially phenolphthalein or castor oil (ricinoleic acid) (p. 823). Secretory diarrhoea may also occur with the carcinoid syndrome associated with small intestinal primary tumours, medullary carcinoma of the thyroid, and gastrinomas. Unabsorbed bile acids in patients with ileal resection and long-chain fatty acids in patients with Steatorrhoea can cause colonic secretion and diarrhoea.

Investigation The presence of secretory diarrhoea is suggested by the persistence of diarrhoea during fasting and a stool osmolarity similar to that of plasma. In cases of diagnostic difficulty confirmation of excess secretion is best obtained by placing a triple-lumen tube in the small intestine and measuring the absorptive and secretory response to various physiological solutions over a defined segment. Normally there is water and electrolyte absorption, but under these circumstances the solutions are further diluted by active intestinal secretion of water and electrolytes.

1

794

MCQ 15.10

DIARRHOEA DUE TO RAPID TRANSIT

Diarrhoea and malabsorption may follow small intestinal resection. This is most commonly due to Crohn's disease, ischaemia due to superior mesenteric artery occlusion, or radiation damage. The greatest problems with water and electrolyte handling occur with loss of the duodenum or upper small intestine. Heal resection leads to vitamin B12 and bile salt malabsorption (see below). Some patients can be managed by a low-fat diet with regular small meals, and isotonic glucose and electrolyte solutions to maintain hydration. If the patient continues to produce several litres of stool daily and to lose weight, long-term total parenteral nutrition must be considered.

ILEAL DYSFUNCTION AND RESECTION Heal dysfunction is most commonly seen as a result of Crohn's disease, but may also follow TB, tumours and radiation damage. These diseases often lead to ileal resection. Ileal resection following mesenteric vascular disease may be extensive.

Clinical features Ileal disease and/or resection result in vitamin B12 malabsorption and deficiency and malabsorption of bile acids, with disruption of their normal enterohepatic circulation. Bile acid malabsorption leads to: • Diarrhoea. This is the result of colonic secretion induced by the secretory effect of unabsorbed bile acids (e.g. chenodeoxycholic acid) on the colonic epithelium. • Depletion of the total body bile acid pool (if the disease or resection is extensive). This results in an increased cholesterol: bile acid ratio in the bile and the risk of gallstones. It also leads to reduced concentrations of bile acids in the jejunal lumen, with the consequence of steatorrhoea. In Steatorrhoea the unabsorbed long-chain fatty acids not only induce colonic secretion and diarrhoea, but also promote increased colonic absorption of dietary oxalate, leading to hyperoxaluria and oxalate renal stones.

Management Treatment is first of the causative disease process. Vitamin B12 deficiency is treated with an intramuscular loading dose

of 1 mg daily for 5 days; thereafter, 1 mg at 3-month intervals is adequate. Bile acid malabsorption can be assessed by giving a patient 75Se-labelled homocholic acid conjugated with taurine (75SeHCAT) and measuring whole-body retention at 7 days. Low values indicate bile acid malabsorption. Cholestyramine is an anion-exchange resin which can alleviate bile acid-induced diarrhoea by binding bile acids and reducing intracolonic free bile acid concentrations. Unfortunately, this often fails to cure the diarrhoea and results in steatorrhoea and further depletion of the bile acid pool, and hence a greater tendency to cholelithiasis. Gallstones should be treated surgically. Hyperoxaluria should be treated with a low-oxalate diet, which includes avoiding fruits such as strawberries; surgical treatment may be necessary for renal stones.

DRUG-INDUCED MALABSORPTION Numerous drugs have been implicated (Table 15.31); the effects are all dose related and reversible.

SMALL BOWEL ISCHAEMIA Aetiology

Aetiological factors in small intestinal ischaemia are shown in Table 15.32; experimentally, venous occlusion can cause infarction but it is not known whether this is the case in humans.

Clinical features and management Small bowel ischaemia may present acutely, with gut infarction following occlusion of the coeliac and superior mesenteric arteries, or chronically, with postprandial pain and weight loss.

15

Acute ischaemia Acute ischaemia presents with pain, diarrhoea, vomiting and an acute abdomen. The patient should be resuscitated with fluids and treated with antibiotics and urgent laparotomy performed. At operation, an attempt is made to restore patency to embolized or thrombosed major arteries, and all non-viable bowel is resected. The mortality is high and, if a large proportion of the upper jejunum and/or duodenum is removed, the patient has severe malabsorption, often with profuse diarrhoea. Such patients may require long-term total parenteral nutrition. Chronic ischaemia Chronic ischaemia is usually due to atheroma and, if undiagnosed, may progress to acute ischaemia. The patient is usually a middle-aged or elderly smoker with evidence of atheroma elsewhere, who complains of severe colicky abdominal pain after eating ('intestinal angina'). This may be relieved by analgesics or vasodilating drugs. Weight loss follows, due mainly to a fear of eating, but malabsorption may also occur. Examination is relatively unhelpful but intra-abdominal bruits may be audible. Investigation is by arteriography, which shows a localized narrowing at the origin of at least two of the major arteries off the aorta. Reconstruction of stenoses can help some patients.

SMALL BOWEL TUMOURS TABLE 15.31 Examples of drugs causing diarrhoea and malabsorption Drug

Mechanism of action

Cholestyramine

Binds bile acids

Steatorrhoea Gallstones Malabsorption of fat-soluble vitamins

Colchicine and cytotoxic drugs Neomycin

Arrest enterocyte mitosis Causes partial villous atrophy and enzyme inhibition

Diarrhoea

Methyldopa

Causes partial villous atrophy Binds with tetracycline

Ferrous sulphate Ethanol

Damage to enterocyte subcellular organelles

Result

Steatorrhoea Iron, vitamin B12 and glucose malabsorption Steatorrhoea Malabsorption of both drugs Folate deficiency and Steatorrhoea

Despite its long length, the small intestine is the site of less than 1% of all malignant tumours and less than 5% of gastrointestinal neoplasms. Malignant tumours occur more frequently in the duodenum and jejunum, and benign lesions, such as leiomyomas and adenomas, in the jejunum. Small bowel malignancy is more common in patients with coeliac disease, Crohn's disease and Gardner's syndrome (p. 800). The commonest tumours are adenocarcinomas (50%), carcinoid tumours (30%), lymphomas (15%) and sarcomas (5%).

TABLE 15.32 Aetiology of small • • • • • •

bowel ischaemia

Arterial atheroma Arterial thrombosis Arterial emboli (usually from the heart) Vasculitis due to collagen disorders External compression of arteries (e.g. tumours, aneurysms) ? Venous thrombosis

795

TABLE 15.33 Features of gut carci no ids according to embryological origin Embryological origin

Usual primary site

Principal active substance

Diagnosis

Treatment

Foregut

Stomach

Histamine 5-hydroxytryptophan

Normal or raised urinary 5-HIAA

Surgery H1 and H2-receptor antagonists

Mid-gut

Terminal ileum

5-hydroxytryptamine

Raised urinary 5-HIAA

Surgery Cyproheptadine, methysergide and somatostatin Embolization of hepatic metastases

Hind-gut

Colon and rectum

None

Histology

Surgery only

Adenocarcinomas Aetiology and pathology The tumours usually occur in the upper small bowel and are more common in patients with coeliac and Crohn's disease. They present as ulcerating obstructing neoplasms and metastasize to regional lymph nodes and the liver. Clinical features and investigation Patients usually present with abdominal pain, and intussusception may occur. Chronic blood loss is frequent and there may be melaena. The diagnosis can sometimes be made by small bowel enema and, occasionally, endoscopically (if in the duodenum) or by colonoscopy (if in the terminal ileum). Management Surgical excision is the treatment of choice but only 70% are resectable. The regional lymph nodes are resected where possible. Metastatic lesions respond poorly to cytotoxic chemotherapy.

Carcinoid tumours 1 Gut carcinoid tumours are rare; their behaviour is to some extent dependent on the part of the gut from which they are embryologically derived (Table 15.33). Histology shows sheets of neuroendocrine cells which are usually of low-grade malignancy. Immunocytochemistry shows positive silver staining for amines, and midgut tumours are often also positive for gut-regulatory peptides such as neurotensin. Foregut tumours usually originate in the stomach and predominantly secrete histamine. Appendiceal carcinoid tumours are commonly found in appendicectomy specimens but do not metastasize or cause clinical problems. Terminal ileal carcinoids are those most likely to metastasize. They

1 796

MCQ 15.11

2

MCQ 15.12

cause the carcinoid syndrome and predominantly secrete 5-hydroxytryptamine. Hindgut tumours do not produce the syndrome and behave similarly to colonic adenocarcinomas.

Carcinoid syndrome Aetiology The carcinoid syndrome only occurs when the liver is bypassed and cannot metabolize tumour products. Symptoms therefore arise when there are hepatic metastases which drain direct into the hepatic vein. Symptoms can also be seen in the absence of hepatic metastases with primary pulmonary, testicular and ovarian tumours whose venous drainage is directly into the systemic circulation. Clinical features The most characteristic symptom of the carcinoid syndrome is a brick-red flush, usually of the face and upper trunk. Provoking agents include stress (via noradrenaline), alcohol and food (possibly via gastrin). Other symptoms include secretory diarrhoea, weight loss, wheezing, dyspnoea and right hypochondrial pain. This is due to the marked hepatomegaly which is usually the main clinical sign. Fibrosis of right heart valves (giving both stenosis and regurgitation) may be present. Midgut carcinoid tumours produce large amounts of 5hydroxytryptamine (5-HT, serotonin; see Table 15.34). This substance is thought to play a part in the pathogenesis of carcinoid syndrome symptoms, such as flushing and diarrhoea. Foregut tumours may secrete histamine and 5-hydroxytryptophan, producing a pink, weal-like flush which may itch and migrate. Pellagra due to nicotinic acid deficiency (resulting from tryptophan consumption by the tumour) may occur. The cause of the right-sided cardiac fibrosis is unknown, but serotonin has been suggested as a possibility. Investigation Functioning carcinoid tumours can be diagnosed by demonstrating raised 24-hour urinary 5-HIAA levels or by whole blood 5-HT. If the diagnosis is confirmed by urine 5-HIAA, but no liver metastases are demonstrable, a bronchial carcinoid should be suspected.

TABLE 15.34 Biochemical pathway for the synthesis and degradation of 5-hydroxytryptamine Tryptophan 5-hydroxytryptophan 5-hydroxytryptamine (5-HT, serotonin) 5-hydroxyindole acetaldehyde 5-hydroxyindole acetic acid* (5-HIAA) * 24-hour urinary 5-HIAA levels are raised in midgut and foregut carcinoids

Management Surgical removal of primary tumours may be helpful, but is rarely so in patients with terminal ileal tumours, where the primaries are usually small and multiple. Symptoms of the carcinoid syndrome may be controlled by 5-HT antagonists such as cyproheptadine. Methysergide, also a 5-HT antagonist, has the disadvantage of causing retroperitoneal fibrosis. The symptoms of histamine-producing gastric carcinoids can be blocked by a combination of H1and H2-receptor antagonists. Somatostatin or its analogues prevents flushing and diarrhoea, presumably by inhibiting the release of provoking agents from carcinoid cells. Diarrhoea may be helped by loperamide and codeine phosphate. Pellagra responds to large doses of nicotinamide. There is no definite role for cytotoxic drugs or radiotherapy, but hepatic arterial catheterization and embolization or surgical tumour debulking may help. In a carcinoid crisis, when the flushing is so severe that arterial hypotension develops, corticosteroids and plasma volume expansion may be necessary.

Lymphomas Lymphomas are the commonest gut tumours in children and a rare small bowel tumour in adults. Aetiology In the west most cases of primary lymphoma arise without any known predisposing cause, but coeliac disease may be complicated by a T-cell lymphoma. In the Middle East, immune proliferative disease of the small intestine, of unknown aetiology, frequently develops into a nonHodgkin's B-cell lymphoma which secretes free -heavy chains (Mediterranean lymphoma). Most cases in the west are non-Hodgkin's B-cell lymphomas. The histology usually shows diffuse infiltration of the bowel wall; nodular forms are rare. The lower small bowel is most frequently involved. The disease spreads early to involve local nodes, but distant dissemination is late. In the late stages of the disease extraintestinal lymphoma can spread to involve the gut.

Clinical features Presentation is often with subacute intestinal obstruction. Bleeding may occur and is usually chronic. Perforation is not unusual. When extensive, the disease may present as a fever of unknown origin.

15

Management The main treatment of primary lymphoma in the west is surgical resection. Further treatment is not required if there is no evidence of spread, but involvement of nodes or distant spread are indications for intensive combination chemotherapy. Local recurrences may respond to radiotherapy. In childhood the risk of CNS involvement is small, in contrast to other childhood lymphomas. The prognosis of localized resected gut lymphoma is good, with over 60% of patients alive at 5 years. It is not yet clear what impact modern chemotherapy will have on survival in more advanced cases. Malignant lymphoma complicating coeliac disease has a poor prognosis.

FURTHER READING ON THE SMALL INTESTINE Davies G R, Benson M J, Gertner D J, Van Someren R M N, Rampton D S, Swain C P 1995 Diagnostic and therapeutic push type enteroscopy in clinical use. Gut 37:346-352. Dieterich W et al 1997 Identification of tissue transglutaminase as the autoantigen of coeliac disease. Nature Med 3:797-801. Fredricks D N, Relman D A 1997 Cultivation of Whipple bacillus: the irony and the ecstasy. Lancet 350:1262-1263. Mustalahti K, Collin P, Sievanen H, Salmi J, Maki M 1999 Osteopenia in patients with clinically silent coeliac disease warrants screening. Lancet 354:744-745. Sulkanen S et al 1998 Tissue transglutaminase autoantibody enzyme-linked immunosorbent assay in detecting coeliac disease. Gastroenterology 115:1322-1328.

COLORECTAL DISEASE The colon and rectum act to reduce up to 1500 mL of fluid material per day, which passes through the ileocaecal valve to create a formed stool of 50-200 g. Important symptoms of colorectal disease are constipation, diarrhoea, blood per rectum, tenesmus, colicky abdominal pain and the symptoms of iron deficiency anaemia. Important signs are abdominal masses and abnormalities on rectal examination. Radiologically, the rectum and colon can be visualized by plain abdominal X-ray and double-contrast barium enema. Endoscopically the rigid sigmoidoscope, flexible fibreoptic sigmoidoscope and flexible fibreoptic colonoscope are used to view the mucosa directly and to take biopsies.

CONSTIPATION 2 The normal person opens their bowels with a frequency between two to three times a day and once every 2-3 days.

797

The normal daily stool weight in the west is 50-200 g, but up to 500 g is common in countries where a much higher fibre intake is consumed. Constipation can be considered to be present when defecation is twice a week or less and is associated with straining, although some patients may be constipated but pass hard faecal pellets several times a day after straining. Constipation is a common complaint in the western world, and vast quantities of laxatives are purchased annually to treat the symptom.

Aetiology and diagnosis Constipation is commoner in women than men and is often due to slow transit through the caecum and ascending colon. Patients often complain of alternating diarrhoea and constipation, where the underlying problem is constipation. In patients under 30 with mild symptoms investigation beyond a sigmoidoscopy is usually not indicated. Some of the many causes of constipation are listed in Table 15.35. Slow-transit constipation is a diagnosis of

exclusion where there is no cause other than a low-fibre diet. Constipation is common in old age, when impaired colonic muscle function may contribute to the problem. Anorectal problems are associated with constipation because the patient delays defecating to avoid pain. Some patients with obstructive defecation contract their anal musculature when it should relax. This problem can be associated with a past history of child abuse. Inflammatory proctitis may result in faecal loading proximal to the inflamed mucosa (best seen on plain abdominal X-ray); the patient usually complains of passing small amounts of faeces frequently. Hypothyroidism, hypopituitarism and hypercalcaemia may have constipation as one of their earliest features. A careful drug history should be taken. Peripheral and central nervous system diseases (apart from Hirschsprung's, which usually presents in early childhood) commonly have obvious manifestations before constipation develops. Constipation may be the presenting symptom of depression. In difficult cases colonic transit studies, using the Shapes' test, and proctography, which assesses anorectal coordination with X-ray screening, can be helpful in determining the physiological abnormality.

TABLE 15.35 Causes of constipation Slow-transit constipation Colonic obstruction Tumours Encarceration in hernias Volvulus Abnormal muscle function Old age Irritable bowel syndrome Diverticular disease Crohn's and ulcerative colitis Ischaemic colitis Systemic sclerosis Hirschsprung's disease Anorectal problems Ulcerative proctitis Anal tissure Thrombosed haemorrhoid Obstructed defaecation Neuropsychiatric disorders Depression Spinal cord injury Cauda equina lesions Multiple sclerosis Parkinson's disease Cerebral tumours Cerebrovascular accidents Anorexia nervosa

1

798

Fig. 15.21

Drugs Anticholinergics Antidepressants Anticonvulsants Antacids (aluminium and calcium salts) Diuretics Opiate analgesics Iron Metabolic and endocrine problems Diabetes Hypothyroidism Hypercalcaemia Pregnancy Uraemia Fever Hypokalaemia Porphyria Lead poisoning Proximal gastrointestinal disease Gastric carcinoma Reduced food intake

Management Wherever possible, the underlying cause of constipation is treated. In slow-transit constipation and where a lowresidue diet is being consumed, a high-fibre diet should be recommended. Patients should be encouraged to drink plenty of fluids and to take regular exercise. Diet sheets and the advice of a dietitian can be useful, particularly in men (Table 15.36). Bran is the best source of fibre as it absorbs a large amount of water and allows the passage of larger amounts of soft stool. Beans, dried fruits, some nuts and pasta also contain large amounts of fibre, and fruit and vegetables contain non-absorbable polysaccharides, such as cellulose and pectins, which are useful. Patients with obstructive defecation can be treated with relaxation and biofeedback, when they are taught to relax their anus during defecation. An approach to the treatment of constipation is summarized in Figure 15.20. Severe constipation is defined as opening the bowels less than once a week. Pharmacological preparations of bulk-forming agents such as methylcellulose, isphaghula husk, sterculia and, possibly, guar are helpful if patients cannot remedy the situation by diet alone. Osmotic laxatives such as the disaccharide lactulose, or magnesium sulphate or hydroxide, can be useful. If possible, the stimulant laxatives which increase intestinal motility (e.g. senna, cascara, castor oil, bisacodyl, docusate sodium and danthron) should be avoided, as it is thought they can cause ganglion degeneration and an acquired megacolon, thereby perpetuating constipation. Stimulant laxatives also tend to cause abdominal colic and hypokalaemia. Faecal softeners (e.g. liquid paraffin and dioctyl sodium sulphosuccinate) have a short-term role in softening stool in the presence of anal disease.

TABLE 15.36 Dietary fibre content of foods Fibre content

g fibre/100 g food

High (more than 10g/100g food) Bran Partially purified bran Haricot and kidney beans Dried apricots Desiccated coconut Almonds Potato crisps Lentils Spaghetti, macaroni, lasagne

44.0 26.7 25.0 24.0 23.5 14.3 11.9 11.7 10.0

Medium (5-1 Og/100 g food) Wholemeal bread Peas Peanuts Raspberries Muesli Sultanas Chestnuts Spinach

8.5 7.8 7.6 7.4 7.4 7.0 6.8 6.0

Low (below 5g/100g food) Cornflakes White bread Boiled potato White rice Cauliflower Lettuce Jam Grapes

3.0 2.7 2.4 2.4 1.8 1.5 1.1 0.9

reflex on anorectal manometry, and a megacolon above the narrow aganglionic segment on barium enema. Treatment is surgical, with an extended internal sphincterotomy or myectomy for short-segment disease or a resection of the aganglionic segment with anastomosis of normal colon to the anorectum (e.g. Duhamel's procedure) for longersegment disease.

15

Idiopathic meg a rectum and colon Rarely, patients can present with long-standing severe constipation due to a grossly dilated rectum or colon. Rectal examination and plain abdominal X-ray show large bowel distended by large amounts of solid faeces. Treatment is with laxatives (see Fig. 15.19) and sometimes with surgical excision.

Chagas' disease Chagas' disease is endemic in South America and is due to infection with the protozoon Trypanosoma cruzi (p. 354). Chronic myocarditis, megaoesophagus and megacolon are late symptoms of the disease. Destruction of the ganglia produces a dilated and atonic colon. Treatment is with bulking agents and laxatives, but occasionally panproctocolectomy and ileostomy may be necessary.

BENIGN TUMOURS OF THE LARGE BOWEL Neuropathic colon

Large bowel polyps

Neuropathic colon occurs as a result of a loss of Auerbach's and Meissner's plexuses, along with the associated intrinsic neurons. This results in failure of colonic motility and constipation.

Some polyps may progress to carcinomas, and early diagnosis and removal therefore has the potential to prevent colonic carcinoma. Colonoscopes allow the passage of wire snares which can be pulled tight around the stalk of polyps; a diathermy current can then be passed through the polyp, enabling its removal. Many laparotomies have been avoided by the use of this technique.

Hirschsprung's disease (congenital colonic aganglionosis) In this familial disorder, which occurs in about 1 in 4500 live births, the ganglion cells of the neural plexus of the colonic wall are absent. The loss extends for a variable length from the anal margin. In two-thirds of patients the rectum and sigmoid colon are involved, but much shorter segments, the whole colon or, rarely, the whole intestinal tract may be involved. Most cases present in childhood, either with acute intestinal obstruction in the neonatal period or with chronic constipation in later childhood. Occasionally the disorder can present in adults - even in those past middle age - but all will have a history of constipation dating back to childhood. Some, but not all, of these adults will have short or ultrashort segments of aganglionosis. Diagnosis is made by the demonstration of aganglionosis and hypertrophic nerve fibres in the neural plexus on rectal biopsy, an absent rectosphincteric inhibitory

Classification A classification of colonic polyps is shown in Table 15.37. Colonic carcinoid tumours are relatively rare and do not produce biologically active amines or peptides. Metaplastic polyps are usually small (<1 cm), whereas adenomas may grow to several centimetres in diameter. Small adenomatous and metaplastic polyps cannot be differentiated visually and, as subsequent management is different, all polyps must be assessed histologically. Inflammatory polyps are islands of regenerating mucosa which occur with severe ulcerative or Crohn's colitis. Hamartomas are heaped-up areas of normal tissue arranged in a disorganized way; they are commonest in children but may be seen in patients of any age. Peutz-Jeghers syndrome is a dominantly inherited disease associated with pigmentation of the skin and mucous membranes, O and

799

FIG. 15.19 Treatment of constipation (After Gattuso & Kamm 1993 Alimentary Pharmacology and Therapeutics 7: 487-500)

hamartomatous polyps in the stomach and small and large intestines. These polyps only rarely undergo malignant change. Clinical features and investigation Many polyps are asymptomatic. When symptoms do occur, the most common is the passage of blood per rectum; altered bowel habit, colicky abdominal pain and, rarely, anorexia and weight loss may also be seen. For nonmalignant polyps the only physical signs are anaemia due to blood loss or mucocutaneous pigmentation (in PeutzJeghers syndrome). A double-contrast barium enema shows most polyps (Fig. 15.20). Colonoscopy is performed when the barium enema is equivocal, and is also essential for biopsy and treatment. If dysplastic or frankly malignant polyps are removed, long-term colonoscopic follow-up is required.

Familial adenpmatous polyposis (polyposis coli) Familial adenomatous polyposis is an autosomal dominantly inherited condition in which affected individuals develop multiple (usually more than 100) adenomatous colonic polyps. A variant of this condition, associated with epidermoid cysts and osteomas, is known as Gardner's syndrome. Most patients develop carcinomatous change in their polyps before the age of 40. Panproctocolectomy with ileostomy formation, or ileoanal anastomosis with a pelvic ileal pouch, is therefore recommended in affected individuals in their late teens. Family screening for affected siblings is essential and genetic markers will soon be available (see Recent Advances box, p. 802). Some patients also develop adenomas of their small intestine and stomach, and these may rarely become malignant. Periampullary carcinoma is an important complication.

CARCINOMA OF THE LARGE BOWEL O 1 800

MCQ 15.13

Colorectal carcinoma is primarily a disease of western Europe, Australasia and North America. Carcinoma of the

15

FIG. 15.21 Age-specific incidence of carcinoma of the large bowel in the western world

FIG. 15.20 Demonstration of colonic polyps Double-contrast barium enema showing many small colonic polyps.

syndrome, both of which are inherited as autosomal dominant traits, and in some other patients (e.g. familial adenomatous polyposis and hereditary non-polyposis colorectal cancer). Adenomatous polyps and ulcerative and Crohn's colitis are well established precursors.

Pathology TABLE 15.37 Histological classification of colonic polyps Potentially malignant

No malignant potential

Adenomas Tubular Villous Tubulovillous

Metaplastic (hyperplastic) polyps

Carcinoid Hamartomas Peutz-Jeghers Juvenile

Connective tissue polyps Lipomas Fibromas Leiomyomas Inflammatory Ulcerative colitis Crohn's disease

colon affects the sexes equally (Fig. 15.21) but carcinoma of the rectum has a higher incidence in men. There were 19625 deaths from carcinoma of the colon in the UK in 1992, and it is therefore second only to carcinoma of the bronchus as a cause of death from malignant disease.

Aetiology The aetiology is unknown, although clearly there are factors associated with a western lifestyle. For example, the incidence is low in Japan but higher in Japanese who have moved to the USA. Diet has been implicated, but its role is controversial. Increasing dietary calcium and the ingestion of non-steroidal anti-inflammatory drugs and aspirin, but not increasing dietary fibre, appear to have protective effects. Altered colonic bacterial flora and changes in bile acid secretion may be important in pathogenesis. Genetic factors are important in familial polyposis and Gardner's

The pathology is a malignant epithelial tumour of the rectum or colon. Macroscopically there are three main appearances: polypoid, ulcerative and annular lesions. The primary site is most commonly the rectum, followed by the sigmoid colon and the right side of the colon. These tumours then spread outwards through the wall of the bowel and invade local tissues, lymphatics and blood vessels. Staging of the tumour by Dukes' classification is important in predicting prognosis (Fig. 15.22). Microscopically, the tumour is nearly always an adenocarcinoma.

Clinical features Many tumours are initially asymptomatic or their symptoms (rectal blood loss, constipation) are misinterpreted by the patient, resulting in a delay in diagnosis. The change in bowel habit may be constipation, diarrhoea, or one alternating with the other. Colicky abdominal pain is usually associated with colonic obstruction, whereas a constant pain implies local invasion. Tenesmus suggests a rectal carcinoma, but anal pain suggests local invasion down towards the anus. Air and faeces being passed via the urethra or the vagina imply local invasion anteriorly from the rectum, with fistula formation; posterior invasion may result in sacral pain. Tumour bleeding may cause anaemia. Liver metastases may give rise to hepatomegaly, with jaundice as a late sign. Lymphatic spread is usually to para-aortic nodes and may cause extrahepatic biliary obstruction. Ascites may occur as a result of peritoneal carcinomatosis. The tumour may be palpable, and if it is obstructing the bowel lumen there will be proximal faecal loading. Rectal examination and sigmoidoscopy are essential as part of the initial

801

RECENT ADVANCES IN COLORECTAL CARCINOMA Genetics There have been great advances in the understanding of the genetic basis of colorectal carcinoma. There is a lot of evidence that colon cancers are associated with an accumulation of quantitative and qualitative alterations in gene expression. There are two main types of genetic change. First, alterations in pro-oncogene expression are described. Mostly these take the form of alterations of myc and ras oncogene products. Sometimes it is not clear if the changes are the cause of malignant transformation or the result of increased proliferative activity. Ras gene mutations are seen in 58% of adenomas more than 1 cm in diameter and in 47% of colorectal carcinomas. Chromosomal abnormalities are the second type of genetic change seen in colorectal cancers. These take the form of allellic deletions of putative tumour suppressor genes. Deletions at 17p and 18q are seen in more than 70% of colorectal carcinomas. The best worked-out example of allellic loss is that from 5p, which is seen in familial polyposis coli and some sporadic colon cancers. Screening Considerable research effort has been devoted to screening for colorectal carcinoma. Colorectal cancer fulfils two of the four main conditions for successful screening: it is a common major health problem and it can be treated successfully, at least in its early stages. It is less clear whether the other two conditions, namely a readily acceptable good screening test and costeffectiveness, can be achieved. In the absence of anything better so far, American data suggest that patients over the age of 50 with an average cancer risk should at least have an annual faecal occult blood test (with or without rectal examination) and a 3-5-yearly flexible sigmoidoscopy (to 60cm from the anal margin). Mathematical modelling shows that this screening programme would halve the probability of death from colorectal cancer in a 50-year-old man.

assessment. The fibreoptic sigmoidoscope can visualize the whole of the left side of the colon. Patients may present with large bowel obstruction or perforation, in which case urgent surgery is required.

Investigation If colorectal carcinoma is suspected, a colonoscopy or

1

802

Fig. 15.22

FIG. 15.22 Dukes' staging of carcinoma of the colon A Tumour involves mucosa and submucosa only, good prognosis. B Tumour has penetrated muscle, moderate prognosis. C Tumour has spread to lymph nodes, poor prognosis. D Distant metastases are present. The corresponding 5-year survival rates are also shown.

double-contrast barium enema is carried out. If abnormalities on a barium enema are equivocal, a colonoscopy is essential. Investigations are necessary to determine the extent of spread of the tumour by performing abdominal and pelvic imaging with ultrasound, CT 1 or MRI. For rectal carcinomas local spread is best assessed by per-rectal ultrasound. Raised serum hepatic alkaline phosphatase may indicate hepatic metastases. Serum Carcinoembryonic antigen (CEA) is elevated in some patients at diagnosis (36% in Dukes' A and B tumours) and can be of use in monitoring response to treatment and diagnosis of recurrence.

Management A laparotomy is indicated unless a carcinoma is localized within a resectable polyp. For right-sided tumours a right hemicolectomy is performed, and a left heniicolectomy for left-sided tumours. Very low rectal tumours are usually treated with abdominoperineal resection of the tumour and rectum. There is some interest in treating very small rectal tumours by local diathermy or laser techniques. These techniques may also be used palliatively in advanced inoperable lesions causing obstruction or bleeding. For operable lesions the role of pre- or postoperative radiotherapy is still being assessed. There is now evidence that adjuvant chemotherapy postoperatively improves the prognosis in patients with Dukes' B or more advanced tumours. Embolization of hepatic metastases can be helpful for pain, as can external beam irradiation.

The prognosis depends on the extent of spread at the time of surgery. The overall 5-year survival rate in the UK is approximately 25% (80% for Dukes' stage A, 60% for stage B and 20% for stage C). Most patients with Dukes' stage D die within a year. Early diagnosis is therefore critical. In patients with Dukes' stage A tumour there is a greater risk from new primary tumours than from metastatic spread, and 3-yearly colonoscopic surveillance should be performed.

15

OTHER MALIGNANT COLORECTAL TUMOURS Malignant colorectal tumours other than adenocarcinomas constitute less than 5% of primary malignant colorectal neoplasms. They behave in a similar way to adenocarcinomas, and include carcinoids, lymphomas, leiomyosarcomas, fibrosarcomas and melanomas. Squamous carcinoma of the anus may also invade the rectum and be clinically indistinguishable from a rectal adenocarcinoma.

DIVERTICULAR DISEASE (DIVERTICULOSIS) Incidence and aetiology Colonic diverticula are herniations of mucosa and submucosa through the wall of the large bowel, usually at the point where arteries pass through the submucosa. The most common site is the sigmoid colon, but colonic diverticula may occur throughout the large bowel and can be very numerous. They are rare before the age of 30, but occur in about 50% of people by the age of 70. They are uncommon in the Middle and Far East and in Africa. Diverticula occur because of a weakness in the bowel wall and increased intracolonic pressure. The former is associated with an age-related reduction in strength of colonic connective tissue. The latter is thought to be due to reduced colonic contents resulting from a low fibre intake.

Clinical features Uncomplicated diverticulosis may be asymptomatic or be associated with colonic symptoms similar to those of irritable bowel syndrome (p. 817). Symptoms may arise as a resut of the complications of diverticular disease, the most important of which is an inflammatory mass (diverticulitis). This presents with abdominal pain, guarding tenderness and fever, and sometimes as an emergency with ileus, septicaemic shock, pericolic abscess or perforation. Diverticulitis can result in stricture formation. In diverticular disease there may be profuse colonic bleeding or a chronic iron deficiency anaemia. The differential diagnosis will then include other causes of chronic blood loss, such as a carcinoma.

FIG. 15.23 Double-contrast barium enema showing widespread diverticula, worst in the descending and sigmoid colon

Investigation Anaemia suggests blood loss, and a raised white cell count and ESR suggest inflammation. If patients are acutely unwell, a plain abdominal X-ray may show subdiaphragmatic gas due to perforation. When the problem is not acute a double-contrast barium enema is useful to show the extent of the diverticula (Fig. 15.23). If the patient is persistently unwell, with a pyrexia, ultrasound should be performed to detect an abscess following a localized perforation. A difficult diagnostic problem is the presence of a carcinoma in a segment of colon affected by diverticular disease. The barium enema is often difficult to interpret, and so colonoscopy is indicated.

Management There is no specific treatment for uncomplicated diverticular disease. A high-fibre diet may prevent further diverticula developing. The patient should be reassured that complications are unlikely to develop. Antispasmodic drugs such as mebeverine may help some patients with colicky pains. Broad-spectrum antibiotics are indicated for diverticulitis. Surgery is indicated for severe haemorrhage, local abscess, perforation and stricture causing obstruction.

803

PNEUMATOSIS COLI Pneumatosis coli is a rare syndrome where cysts of gas (comprising air and variable amounts of hydrogen and methane) up to 2 cm in diameter form in the submucosal and subserosal surfaces of the colon (and occasionally the small intestine). It can occur as a result of a peptic ulcer. In some patients there may be no evidence of a mucosal break, and most have chronic obstructive airways disease. Most patients are asymptomatic, but symptoms can include bloody diarrhoea, lower abdominal pain and increased rectal flatus. The cysts may be palpable rectally and sigmoidoscopy usually shows them as multiple blue swellings. Biopsy causes deflation, and the biopsy specimen often floats on the surface of the formol saline fixative. Plain abdominal X-ray shows multiple translucent cysts along the course of the colon and may be used to follow progress.

is reassured and encouraged to take a high-fibre diet to avoid constipation. The anus is kept clean by washing twice daily, without the use of strong soaps, ointments etc. If the problem is persistent and there is no obvious treatable cause, local 1% hydrocortisone and oral antihistamines may help.

FAECAL INCONTINENCE Faecal incontinence is usually due to weakness of the smooth muscle of the internal anal sphincter or to impacted stool in the rectum. It affects 7% of healthy adults over 65. The most common cause is childbirth, but it also occurs after surgery and in patients with neurological disease, particularly multiple sclerosis. It can be defined by anal manometry and ultrasonography. Antidiarrhoeal drugs such as loperamide and codeine are often effective in patients without faecal impaction. In more severe cases specialist investigation and surgery may be required.

Management Treatment is symptomatic, but breathing high concentrations of oxygen in a hyperbaric chamber accelerates resorption of the cysts. This is usually effective within 5 days, but there is a tendency to relapse.

PRURITUS ANI Itching around the anus is a common symptom and can have a number of causes (see Table 10.31, p. 433). A careful history to identify an aetiological factor is followed by anal inspection, rectal examination and, if necessary, proctoscopy or sigmoidoscopy. In the UK many patients use proprietary ointments and creams for their 'piles', and these treatments commonly produce contact dermatitis and pruritus. Unless there is a thrombosed pile or hygiene is poor, pruritus rarely complicates haemorrhoids. Investigation of infections requires the laboratory examination of anal smears and, sometimes, faeces for threadworms and infections such as Candida. Tight-fitting clothing and obesity cause local warmth and moisture, and predispose to skin infection and inflammation. In some patients no cause can be identified; some of these have an obsessive and neurotic attitude, which may be the primary problem.

Management Treatment is directed at the underlying cause. The patient

1

804

MCQ 15.14

2

MCQ 15.15

INFLAMMATORY BOWEL DISEASE 1 2 Inflammatory bowel disease can be defined as chronic inflammatory disease of the gut of uncertain aetiology. It is divided into two types: • Crohn's disease, which affects the gut from mouth to anus and is characterized pathologically by noncaseating giant cell granulomata; • Ulcerative colitis, which affects only the large bowel. In Crohn's disease small aphthoid ulcers, larger ulcers and strictures are seen macroscopically. Histology shows the whole bowel wall to be infiltrated by lymphocytes and plasma cells. In contrast, ulcerative colitis (UC) is primarily a superficial mucosal inflammation. In both diseases the large bowel may initially be the only site of involvement, and differentiation between them is then less important as the presentation, complications and management are very similar. It is possible that the two diseases are different responses to the same aetiological agent.

CROHN'S DISEASE Crohn's disease is a granulomatous disease that most commonly affects the terminal ileum and right side of the colon. It is named after Dr Burrill Crohn of New York, who described 52 cases in 1932. Crohn's disease is common in the western world but rare in underdeveloped countries. The annual incidence of new cases appears to be rising in all countries studied (England, Wales, Scotland, Scandinavia and North America) and is currently 3-6 per 100000 population. It occurs equally in the sexes. The initial presentation can be at any age but is most common in early adult life.

CASE STUDY 15.4 BLOODY DIARRHOEA IN A 23-YEAR-OLD AFRO-CARIBBEAN WOMAN History A 23-year-old female secretary of Afro-Caribbean descent was seen in clinic with a 3-week history of diarrhoea with bleeding, abdominal pain, anorexia and weight loss. She was opening her bowels 6-8 times a day and twice at night. The stool was liquid and dark brown, containing fresh and altered blood and mucus. She had lower abdominal cramping pain, which was associated with the urge to defecate and was relieved by defecation. She had lost 6 kg in weight over the 3 weeks. There was nausea but no vomiting. Her symptoms started while on holiday in Spain, but none of her friends who went with her were ill. However, further questioning revealed that her stools had been looser than normal and she had mild abdominal discomfort over the 3 months before her holiday. There was no relevant previous medical history and her only medication was the oral contraceptive pill. She had been treated with loperamide and ciprofloxacin for 1 week, with no effect on her symptoms. There was no relevant family history and she was a non-smoker. Examination The patient was pale and tired, with a temperature of 38°C and pulse of 95 beats/minute. Her abdomen was slightly distended, with generalized tenderness to deep palpation. There was no guarding or rebound tenderness and bowel sounds were normal. Perianal inspection was normal and rectal examination revealed bloody liquid stool on the glove.

Questions l. What is the differential diagnosis? 2. Does the drug and social history help with the differential diagnosis?

Investigations Haemoglobin 7.9g/dL, MCV 70 fl, MCH 64 pg,WCC 12.5 (neutrophil leukocytosis), platelets 550, serum albumin 30 g/L; the rest of the routine biochemistry was normal; CRP 15 mg/L; stool microscopy and culture, negative; plain abdominal X-ray, no dilatation or evidence of perforation, but the air-contrast appearance suggested a colitis extending from rectum to mid-transverse colon; rigid sigmoidoscopy, inflamed rectal mucosa with contact bleeding and muco-pus in the lumen. Question 3. What histological changes could be seen in the rectal biopsy? Rectal biopsy - typical appearances of acute ulcerative colitis. Management A diagnosis of acute colitis was made on the basis of initial sigmoidoscopy. Question 4. How is the severity of colitis assessed? , I She was admitted because the colitis was severe, as indicated by significant anaemia and hypoalbuminaemia, fever, tachycardia and nausea. She was treated with intravenous fluids and hydrocortisone 100 mg qds and kept nil by mouth. Her fluid balance and urine output were monitored and her stools were charted. She had daily plain AXR, full blood count, blood urea and electrolytes until she was significantly better. After 24 hours she was improved with no nausea, less abdominal pain and diarrhoea and normal temperature. At 48 hours she had no further blood PR. After 3 days she was passing three semiformed blood-free stools a day and had no abdominal pain. She started a low-lactose light diet without any relapse in her symptoms.

15

She was changed from hydrocortisone i.v. to prednisolone 40 mg daily by mouth. After 6 days she was discharged on prednisolone 30 mg daily and iron supplements, feeling well and passing two soft stools a day. At review in clinic 1 week later she was still improving. She was started on Asacol (mesalazine) 800 mg t.d.s. and told to progressively reduce the prednisolone dosage over the next 2 months so that the drug could be stopped. After 3 months she was well on Asacol. Iron and prednisolone had been stopped. Her hemoglobin was 12.5 g/dL and her albumin and CRP were normal. Colonoscopy showed a quiescent colitis from the proximal transverse colon to the anal margin. Multiple biopsies confirmed the diagnosis of ulcerative colitis. Discussion The sudden onset of bloody diarrhoea on holiday in Spain suggested an infective colitis. This could have been caused by bacteria, such as Salmonella, Shigella or Campylobacter species. Viral gastroenteritis does not cause bleeding and amoebic dysentery is not a disorder likely to be acquired in Spain. However, stool microscopy and culture were negative. She had no diarrhoeal contacts as far as she knew, and no response to ciprofloxacin. Other causes of bloody diarrhoea needed to be considered, and particularly in a young adult, inflammatory bowel disease (Table 15.17). Close questioning revealed that her bowel had not been right before her holiday. The confirmation of colitis was made at sigmoidoscopy. This examination must be performed gently and carefully, with minimal air insufflation to minimize the risk of precipitating toxic megacolon. The plain AXR is performed in all severe colitics to exclude toxic megacolon. In addition, the colitis and its extent can often be seen, as the air in the colon shows a featureless colon with

805

CASE STUDY 15.4 CONTINUED oedematous mucosa. Rectal biopsy confirmed ulcerative colitis, but it is possible that the exacerbation of her symptoms was precipitated by a concurrent enteral infection. She was a non-smoker, which is more common in ulcerative colitics than in patients with Crohn's disease. Questions 5. Why was she started on a low-lactose diet? 6, What are the therapeutic options if she had not responded to corticosteroids? Her colitis was severe (see Table 15.42) and required in-patient treatment. Mild or moderate attacks can be treated on an outpatient basis. All severe colitics are given i.v. fluids and are kept nil by mouth. Ill patients must be started on an intravenous corticosteroid and, if infection is suspected and cannot be excluded, antibiotics must also be

given. Similarly, very ill patients with ulcerative colitis or Crohn's colitis who are toxic and/or have guarding and rebound tenderness should be given antibiotics. Patients can be fed orally after 3-4 days if they are recovering, rapidly as in this case. However, patients with very severe fulminating colitis and/or toxic megacolon need 5 days' intensive treatment without oral feeding. If at 5 days they are no better, or if they relapse on refeeding, emergency colectomy should be considered. These severely ill patients may need parenteral nutrition. Ciclosporin intravenously may help acutely ill non-responding ulcerative colitics who do not have a toxic megacolon. If food is tolerated, intravenous corticosteroids are replaced by oral medication. Lactose intolerance due to primary lactase deficiency is a problem in many populations and can cause troublesome diarrhoea and flatulence, particularly in patients

recovering from colitis. Seventy-five per cent of adult Afro-Caribbeans have primary lactase deficiency, so it was sensible to start feeding her with a low-lactose diet. After discharge prednisolone is gradually reduced and, as the drug has no role in maintaining remission, is stopped. Reduction of the dose and cessation of therapy can lead to relapse, but it is unusual not to be able to stop prednisolone completely. This is in contrast to Crohn's colitis. 5-Aminosalicylic acid preparations have no role in the treatment of moderate or severe acute colitis but have a major role as maintenance treatment because they reduce the risk of relapse. Colonoscopy allows assessment of the extent of disease and confirmation of its nature. Sometimes it is not possible to distinguish ulcerative colitis from Crohn's colitis even on histology (indeterminate colitis). Colonoscopy can be left until the acute attack has settled.

CASE STUDY 15.5 ABDOMINAL PAIN, WEIGHT LOSS AND DIARRHOEA IN A 23-YEAR-OLD ASIAN WOMAN History A 23-year-old South Asian (Indian) housewife presented with a 6-month history of right iliac fossa pain, anorexia, weight loss of 6 kg and diarrhoea. The pain was episodic and occasionally severe but was not associated with eating, defecation or micturition. She had no upper GI symptoms. She opened her bowel 3-5 times a day with a soft or watery stool of normal colour containing no blood or mucus. She did not awake at night to defecate. She had no musculoskeletal or urinary symptoms. She had not had a period for 3 months, although she did not think she was pregnant. There was no relevant previous medical history and she had never been pregnant. She

806

was taking no regular medication. There was no family history. She was a non-smoker who had been born in England. She had been to India on one occasion as a 12-year-old. Examination She was thin and pale, with tenderness but no guarding to deep palpation in the right iliac fossa. No masses were palpable. Perianal inspection and rectal examination were normal. Questions 1. What are the possible diagnoses? 2. What blood investigations are necessary?

Investigations Haemoglobin 9.5g/dL; MCV 69 fl; MCH 22 pg; WCC normal; platelet count 456 x 109/L. Biochemical profile normal except for serum albumin - 29g/L. C-reactive protein, 40mg/L (n = < 5). Serum iron was low and TIBC was raised; serum B12 and folate were normal. Endomysial antibodies were absent. TSH was normal. Question 3. What is the most likely diagnosis and how would you confirm it? Pregnancy test was negative. Stool microscopy was negative. Flexible sigmoidoscopy was normal. A small

15

CASE STUDY 15.5 CONTINUED bowel follow-through showed the typical radiological appearances of Crohn's disease affecting the ileocaecal area. A colonoscopy was later performed which showed an inflamed contracted caecum. Biopsies from the caecum and terminal ileum showed the typical features of Crohn's disease with non-caseating granulomas.

Question 4 Was a colonoscopy absolutely necessary in this patient?

Management A diagnosis of ileocaecal Crohn's disease was made. In view of the severity of her illness she was started on prednisolone 40 mg daily, as well as iron supplements. After 1 month she was much better, with loss of diarrhoea, lessening of the abdominal pain, return of appetite and weight gain of 8 kg. She looked slightly Cushingoid and was no longer tender in the right iliac fossa. Her haemoglobin had improved to 11.0, her serum albumin to 33g/L and her CRP was normal. She was able to progressively cut down her prednisolone dose to 10 mg daily over the next 8 weeks without relapse. She did not attend follow-up appointments and was re-referred as an urgent appointment 18 months later, when she was admitted with a 4-week history of increasingly severe attacks of colicky central abdominal pain associated with nausea and vomiting. There was no diarrhoea and between attacks she felt reasonably well. She had lost 2 kg in weight. On examination she was not ill but was in pain. Her abdomen was slightly distended, with diffuse mild tenderness but no guarding. Bowel sounds were loud. Her blood tests were normal.

Questions 5. What is the diagnosis? 6. What single test would you do next? A plain abdominal X-ray showed distended small bowel with fluid levels on the erect film. A diagnosis of subacute intestinal obstruction was made and was treated with i.v. fluids and analgesics. She was not vomiting and so a nasogastric tube was not inserted. Her pain settled over the next 24 hours and she was able to tolerate liquids and then a light diet, she had a small bowel meal which showed an extensive tight stricture involving the terminal ileum. She was started on prednisolone and discharged. Her attacks of pain persisted and 6 weeks later she was admitted electively for a right hemicolectomy and ileal resection. She made a good postoperative recovery. Prednisolone was stopped and she was started on mesalazine (as Pentasa) 2 g daily. Six months later she was very well on mesalazine. Question 7. What is the rationale for mesalazine use? Discussion Although the commonest cause of abdominal pain and diarrhoea continuing for several months in a young adult is irritable bowel syndrome, this diagnosis is not the most likely one. Weight loss is unusual in IBS, so special care must be taken to exclude other causes. Crohn's disease may present in this way. The absence of blood per rectum and of typical colonic pain (abdominal pain worse before and relieved by defecation) suggested that colitis or another colonic disorder was less likely. Apart from weight loss there were no typical features of malabsorption, although their absence does not exclude such a

disorder. There was nothing to suggest an infective aetiology. Malignancy was unlikely, although small bowel lymphoma (which is rare) can present like this in younger adults and children. Thyrotoxicosis must be considered in patients presenting with diarrhoea and weight loss, although abdominal pain is not usually a feature. The blood tests showed iron deficiency anaemia, hypoalbuminaemia and raised CRP. Therefore, the diagnosis of IBS was unsustainable. These tests with the clinical features suggested inflammatory bowel disease, particularly Crohn's disease of the small bowel. However, infection and malignancy remained possibilities. The barium study of the small bowel showed appearances typical of ileocaecal Crohn's disease. However, these are not diagnostic and tuberculosis and lymphoma may give similar appearances. In practical terms Crohn's disease is by far the most likely cause of these radiological appearances, and therefore the patient can be treated as a presumptive case of Crohn's disease. Tuberculosis was an unlikely diagnosis, although the patient was of Indian origin; she was born in Britain and had only been to India once 10 years previously on holiday. There was no family history of TB. A colonoscopy is not essential but does allow a definitive diagnosis. It would be indicated if there was a higher suspicion of another cause for the clinical and radiological features. The patient's disease was active, as indicated by her gastrointestinal and systemic symptoms and by the presence of anaemia, hypoalbuminaemia and raised inflammatory markers (CRP). The main treatment for active Crohn's disease remains corticosteroid therapy. Her initial response was encouraging and, unlike many patients with Crohn's disease, she was able to stop therapy without relapse.

807

CASE STUDY 15.5 CONTINUED On her return to hospital 18 months later she had developed typical symptoms of subacute intestinal obstruction. This was confirmed by plain abdominal X-ray during an attack of pain. Subacute obstruction in Crohn's disease does not require emergency surgery and the symptoms will settle with 'drip and suck' and

analgesia. A terminal ileal stricture was later shown on a small bowel meal. Stricture can be due to mucosal oedema secondary to active Crohn's disease or to scarring as a result of previous disease. The normal blood tests and the lack of response to corticosteroids suggested the latter. The treatment of choice was ileal

Aetiology

SUMMARY 7 Clinical presentation of Crohn's disease

The aetiology is unknown but evidence suggests that the cause is largely environmental. For example, when a population migrates from an area of low to one of high incidence, the incidence in that population subsequently rises. A 1995 report from North Tees suggested 7.3 cases/ 100000/year for Crohn's and a rise to 22 for ulcerative colitis. Prevalence rates of up to 100 in 100000 for ulcerative colitis and up to 147 in 100000 for Crohn's are recorded. The following theories have been put forward: • Infection. No infectious agent has yet been shown to transmit the disease or produce a similar granulomatous reaction in animals. At present there is interest in the possible role of Mycobacteria, Campylobacter and measles. • Immunological. The characteristic granuloma of Crohn's disease suggests a possible immunological origin. Both food and bacterial antigens have been implicated, but there is no clear evidence for either. As with other inflammatory bowel diseases, minor, nonspecific, immunological changes have been reported. • Dietary. A high intake of refined sugars and a low intake of fibre have been suggested as aetiological factors, but the evidence that they are primary factors is weak. • Genetic. There is a higher incidence of both Crohn's disease and UC in relatives of patients with Crohn's disease. Concordance rates in twin pairs and siblings are high, with a relative risk of about 25 in patients with inflammatory bowel disease. • Smoking. In direct contrast to UC, patients with Crohn's disease are more likely to be cigarette smokers than are appropriate controls. • Others. Oral contraceptives, detergents, mercury and psychological trauma have all been proposed as aetiological agents, but with little firm foundation.

1 808

resection with right hemicolectomy: 80% of patients show endoscopic if not clinical or biochemical evidence of relapse 6-12 months after resection. There is now good evidence that mesalazine in a dose of at least 2g/day as maintenance therapy after resection reduces the risk of relapse.

Fig. 15.23

2

Fig. 15.24

General Fever Malaise Weight loss Amenorrhoea In children: failure to thrive, growth retardation, delayed puberty Terminal ileal disease Abdominal pain Diarrhoea (with or without steatorrhoea) Right iliac fossa tenderness or mass Signs of small intestinal obstruction Colonic disease Abdominal pain Diarrhoea Blood and mucus per rectum Tenesmus Abdominal tenderness Anal disease Skin tags, fissures and abscesses Incontinence Fistulae between anal canal and skin

Mouth

Cobblestone mucosa Aphthous ulcers Glossitis Stomatitis Fistula formation Entero-enteric or enterocutaneous Urinary frequency, dysuria, haematuria and pneumaturia Faeces per vaginam and per urethram Complications due to nutritional deficiencies Anaemia Oedema due to hypoproteinaemia Gallstones Tetany Osteomalacia Vitamins A and K deficiency Water-soluble vitamin deficiency Ion deficiencies (potassium, magnesium and zinc)

Pathology At laparotomy, Crohn's disease is seen as thickened bowel with associated enlarged lymph nodes. Microscopically the inflammation is transmural, in contrast to ulcerative colitis, where only the mucosa is usually involved. The other hallmark of Crohn's is the microscopic finding of noncaseating granulomata. As a result of transmural inflammation there may be local abscesses or fistulae between two pieces of bowel (entero-entero or enterocolonic) or between the bowel and the skin (enterocutaneous), usually in the anterior abdominal wall or around the anus. 1

Fistulae may also form between the bowel and bladder or vagina.

Clinical features The commonest initial presentation is with terminal ileal disease, often with coexistent oral and anal involvement. Less common initial presentations are acute right iliac fossa pain simulating acute appendicitis, and acute colitis identical to that of UC. General symptoms When the disease is active the majority of patients have a mild fever and general malaise. Loss of appetite and malabsorption contribute to weight loss. In prepubertal and pubertal children there may be no history of diarrhoea or abdominal pain. Children usually present with non-specific symptoms: failure to thrive, decreased growth velocity, delayed puberty and symptoms of anaemia. Gastrointestinal symptoms The gastrointestinal symptoms depend primarily on the site of disease. As a result of stricture formation, many patients with ileal disease complain of colicky abdominal pain, usually in the right iliac fossa. Diarrhoea, usually modest and containing excess fat, is usual. Disease in the right side of the colon produces similar symptoms, although colonic disease alone does not cause steatorrhoea. It is often combined with ileal disease. Left-sided colonic disease results in diarrhoea without steatorrhoea, but with blood per rectum. Urinary frequency and dysuria may result from inflamed bowel adjacent to the urinary tract (especially the right ureter and bladder). Pneumaturia (air in the urine), and faeces via the urethra or vagina indicate a fistula from the bowel to the genitourinary tract. Complications due to nutritional deficiency The complications of nutritional deficiency may cause symptoms. Anaemia may be due to blood loss or reduced

SUMMARY 8 Investigation of Crohn's disease Sigmoidoscopy and rectal biopsy Stool culture and microscopy Abdominal and chest X-ray Small bowel enema or meal Barium enema Colonoscopy Upper gastrointestinal endoscopy (when indicated) Full blood count and acute-phase reactants, e.g. ESR, CRP or orosomucoids Serum iron, iron binding capacity, folate and B12 Liver function tests, including serum albumin Radiolabelled white blood cells (when indicated)

absorption of iron, folate and vitamin B12. Oedema secondary to hypoproteinaemia is a consequence of both protein loss into the lumen of the gut (protein-losing enteropathy) and reduced dietary intake. Gallstones are due to terminal ileal bile acid malabsorption, which results in cholesterol-supersaturated bile and causes cholesterol gallstones. Hyperoxaluria and renal oxalate stones can occur in patients with steatorrhoea. Vitamin D and calcium malabsorption can result in hypocalcaemia (manifest as tetany) or, when long-standing, pseudofractures (Looser's zones) or fractures due to osteomalacia (p. 962). Night blindness and spontaneous bleeding may rarely occur as a result of vitamin A and K malabsorption. Angular cheilitis, glossitis and even scurvy can result from vitamin B and C deficiency. Hypokalaemia results in malaise and muscle weakness, and hypomagnesaemia can cause tetany. Zinc deficiency can cause a rash similar to that seen in zinc-deficient children with acrodermatitis enteropathica.

15

Non-intestinal manifestations Both Crohn's disease and UC may be associated with nonintestinal manifestations (Table 15.38). Acute large joint arthritis, seen commonly, affects the knees and ankles. Erythema nodosum and pyoderma gangrenosum usually affect the limbs, particularly the legs (p. 407). If urticaria and erythema multiforme are seen they should be assumed initially to be due to sulphasalazine. If there are ophthalmic symptoms, an assessment with a slit lamp is needed to exclude uveitis. Sacroiliitis occurs in up to 10% of patients. The association with ankylosing spondylitis is rare. Abnormal liver function tests occur in about 10% of patients with colitis and are usually asymptomatic. The commonest lesions are pericholangitis and fatty infiltration. Primary sclerosing cholangitis (p. 858) is the most common clinically significant problem. Cholangiocarcinoma and autoimmune chronic active hepatitis are rare. Cholesterol gallstones, secondary to bile acid malabsorption, are common in patients with terminal ileitis or resection. Acute colitis is associated with an increased risk of venous thromboembolism. Signs Physical examination of a patient with Crohn's disease should start with a search for nail changes (including clubbing, leukonychia and koilonychia), anaemia, pyrexia, tachycardia and mouth ulcers. 0 In the abdomen there may be operation scars, evidence of fistulae and signs of intestinal obstruction, with visible peristalsis. Palpation may reveal a mass, most commonly in the right iliac fossa, due to disease in the terminal ileum and the right side of the colon. Auscultation may suggest intestinal obstruction. Perianal inspection, sigmoidoscopy and rectal biopsy are essential. Sigmoidoscopy may show ulceration, a granular inflamed mucosa, spontaneous bleeding and increased mucus formation. Even in the absence of obvious inflammatory changes, the rectal biopsy may demonstrate diagnostic granulomata.

809

TABLE 15.38 Non-Intestinal manifestations of inflammatory bowel disease Affected part

Manifestation

Reflecting acute illness Joints Skin

Eyes

Acute large joint arthritis Pyoderma gangrenosum Erythema nodosum Vasculitis Conjunctivitis Episcleritis Uveitis

TABLE 15.39 Differential diagnosis of Crohn's disease Symptoms

Differential diagnosis

Terminal ileal disease

Tuberculosis Yersinia Tumours, especially lymphoma and primary carcinoid Irritable bowel disease

Colitis

Ulcerative colitis Infective colitis Ischaemic colitis Irritable bowel syndrome

Weight loss and diarrhoea

Carcinoma (especially of the large bowel in older people) Other causes of malabsorption, e.g. coeliac disease Hyperthyroidism Anorexia nervosa (with or without surreptitious purgative abuse) Addison's disease Hypopituitarism

Venous thromboemboiism Not reflecting acute illness Joints Liver

Sacroiliitis Ankylosing spondylitis Fatty change Pericholangitis Primary sclerosing cholangitis Carcinoma of bile duct (cholangiocarcinoma) Autoimmune chronic active hepatitis

Investigation The diagnosis is confirmed by demonstrating the characteristic granulomatous histology, often by rectal biopsy. In the presence of active disease the haemoglobin is usually low and the neutrophil count and acute-phase reactants are raised. The serum albumin is low if there is widespread disease. Stool culture and microscopy should be performed; infections, particularly Yersinia (p. 816) and Campylobacter (p. 324), may mimic Crohn's disease. A plain abdominal X-ray will determine whether there is colonic dilatation. A chest X-ray may reveal tuberculosis, which may present with ileal disease similar to Crohn's disease. A small bowel enema or meal shows involvement of the small bowel, which is characterized by stricture formation, rose-thorn ulcers, proximal dilatation and fistula formation (Fig. 15.24A). The colon and caecum (and often the terminal ileum) are visualized by a double-contrast barium enema (Fig. 15.24B). Multiple aphthous ulcers, seen at colonoscopy or barium enema, are typical of early Crohn's colitis. Colonic changes include thickening of the wall, deep ulcers (rose-thorn and collar-stud types) and skip lesions, with areas of normal bowel interspersed between diseased segments. Colonoscopy allows direct inspection 1 and biopsy of the colon and terminal ileum. A useful technique is to label the patient's white cells in vitro with an isotope such as indium-Ill. After reinjecting them into the patient, the localization and activity of the

1

810

Fig. 15.25

disease can be assessed by determining the site of uptake using a gamma camera.

Differential diagnosis The differential diagnosis of Crohn's disease is shown in Table 15.39. In many parts of Africa, India and the Far East, ileocaecal tuberculosis is quite common and Crohn's disease is rare. The barium radiological appearance is similar to that of Crohn's disease; however, in the case of tuberculosis, caseating granulomata are found on histology, acid-fast bacilli may be seen on microscopy, and Mycobacteria are isolated on culture of the biopsy. Yersinia can present with an acute ileitis in young people, but symptoms and signs disappear within 3 months. Some patients with Crohn's disease may be wrongly diagnosed and treated for anorexia nervosa. Addison's disease and hypopituitarism can also present as weight loss and diarrhoea.

Complications • Acute colonic (toxic) dilatation, indistinguishable from that seen with UC and sometimes complicated by perforation and haemorrhage. • Subacute small bowel obstruction, due to stricture formation and obstruction by a food bolus. • Fistula formation. • Renal stones. • Carcinoma of the colon may complicate long-standing (usually more than 20 years) Crohn's colitis. Small intestinal carcinomas rarely complicate Crohn's disease.

15

FIG. 15.24 Barium enemas demonstrating Crohn's ileitis and colitis A Small bowel enema showing stricture formation and ulceration of the terminal ileum, typical of Crohn's ileitis. B Barium enema showing deep ulceration of the distal colon, suggesting Crohn's colitis.

• Amyloid may occur in the bowel and other organs, and may regress with resection of diseased bowel. • Perforation may occur in acute toxic megacolon or silently in acute disease during steroid therapy. • Severe colonic haemorrhage.

Management Crohn's disease can cause considerable morbidity, but with joint medical and surgical specialist care the overall quality of life should be good. Life expectancy for most patients is virtually the same as for normal individuals, but patients with extensive disease and with onset before 20 years of age still have higher mortality than normal.

Medical Management of patients with Crohn's disease depends on the site, severity and extent of disease (Table 15.40). Smoking may adversely affect prognosis and should be stopped. In general, patients should be encouraged to take a balanced diet with an adequate amount of fibre. Dietary fibre should only be restricted when there is stricture formation, as it may then contribute to intestinal obstruction. Replacements are given if there is evidence of haematinic, fat-soluble vitamin or mineral deficiency.

Active ileitis and colitis Active ileal disease causing diarrhoea, abdominal pain and a right iliac fossa mass is usually treated with a 3-month course of oral prednisolone (starting at 30-40 mg/day). If the disease is resistant, azathioprine (2mg/kg) or 6mercaptopurine may be used. Clinical trials have shown that 5-aminosalicylate preparations are of minor value when the disease is restricted to the ileum. An elemental diet (a solution of essential amino acids, carbohydrate, fat, vitamins and trace elements) may also be of value in the treatment of active ileal disease. Active colonic disease is also treated with prednisolone. Sulphasalazine, which is 5-aminosalicylic acid bound to a sulphonamide (sulfapyridine), may be of some use in Crohn's colitis, although its main role is in the prevention of relapse in UC. However, because of the numerous and potentially serious side-effects (Table 15.41), mostly due to the sulfapyridine, 5-aminosalicylic acid preparations have been developed. One of the enteric-coated preparations uses an acrylic-based resin outer coat (mesalazine) which, like the semipermeable membrane of the slow-release preparation of sulfasalazine, is pH dependent and allows the release of 5-aminosalicylic acid into the ileum and colon. Azodisalicylate consists of two molecules of 5aminosalicylic acid bound by an azo bond which is split by lower intestinal bacteria. Another option is the use of

811

TABLE 15.40 Medical treatment of Crohn's disease Symptoms

Treatment

Nutritional deficiencies

Good balanced diet Folicacid Vitamin B12 Iron Vitamins A, D, E and K Potassium, zinc and calcium

Ileal disease

Prednisolone Elemental diet Azathioprine/6-mercaptopurine 5-aminosalicylate drugs

Colonic disease

Prednisolone Sulfasalazine Azathioprine/6-mercaptopurine 5-aminosalicylate drugs

Severe active ileitis and/or colitis

Intravenous hydrocortisone Intravenous fluids or total parenteral nutrition Antibiotics

Anal disease and fistulas

Metronidazole Azathioprine Elemental diet/total parenteral nutrition and nil by mouth

Non-responsive active disease with or without fistulae

Infliximab

Non-intestinal manifestations

Local measures Prednisolone (if acute illness)

CROHN'S DISEASE

TABLE 15.41 Role and side-effects of salicylate drugs used In Crohn's disease and ulcerative colitis Drug

Side-effects

Sulphasalazine

Prevention of relapse in UC Skin rashes Marrow suppression Some use in ileal and colonic Crohn's disease Haemolysis Male subfertility Gastrointestinal symptoms Orange urine

5-aminosalicylic acid (mesalazine)

Headache Diarrhoea Nephrotoxicity

Alternative to sulfasalazine

Azodisalicylate (olsalazine)

Headache

Alternative to sulfasalazine

5-aminosalicylic acid enemas

Headache

Alternative to Sulfasalazine enemas

1

812

RECENT ADVANCES IN TUMOUR NECROSIS FACTOR- (TNF- ) ANTIBODY IN

Role

TNF- is a cytokine which is thought to have a major role in the initiation and propagation of the inflammatory response and apoptosis in patients with Crohn's disease and rheumatoid arthritis. Mouse-derived monoclonal antibodies (e.g. infliximab and CDP571) have subsequently been raised and assessed in these diseases. The respective antibodies, IgG 1 and IgG4 do not cause an immune response in humans but do neutralize human TNF- . If the antibodies are infused into patients with active Crohn's disease there is a beneficial effect at 2 and 4 weeks on symptoms and laboratory markers of inflammation, but by 6 months, the effects have disappeared. Endoscopic and biopsy studies have shown a significant healing of endoscopic inflammation and ulceration and a reduction in the inflammatory cell infiltrate. A randomized double-blind placebo-controlled trial of three doses of infliximab in patients with fistulae showed an improvement in fistula healing. The most common sideeffects of infliximab were headache, nausea, upper respiratory tract infection and fatigue. Tubercluosis may occur. TNF-a antibodies are a promising new treatment for patients with severe Crohn's disease who do not respond to other agents. 5-aminosalicylic acid enemas. The new 5-aminosalicylate preparations are as effective as sulphasalazine and are replacing it. Resistant patients may respond to azathioprine or 6-mercaptopurine. Patients with severe active ileitis and/or colitis have constitutional symptoms and require hospital admission. Intravenous hydrocortisone sodium succinate (usually l00mg 8-hourly) is usually given for about 5 days and then oral prednisolone is substituted. Intravenous fluids are needed, and if the patient is malnourished total parenteral nutrition should be started. If there is a possibility of infection or abscess the patient is treated with intravenous broad-spectrum antibiotics (e.g. ampicillin, gentamicin and metronidazole). Anal disease Anal disease with sinuses, abscesses or fistulae can be treated with metronidazole for several weeks, but the patient should be warned to report any symptoms suggestive of drug-induced peripheral neuropathy. Resistant anal disease may respond to dietary measures or require surgical resection with the formation of a stoma. In severe cases an elemental diet will greatly reduce stool frequency and help healing. Enterocutaneous and enteroenteric fistulae usually require surgery.

MCQ 15.16

Non-intestinal manifestations Disorders reflecting acute illness settle with removal of the

affected bowel (usually colon), but those due to the chronic illness do not. Severe joint, skin or eye disease requires prednisolone. The treatment of sacroiliitis and ankylosing spondylitis is described in Chapter 22, page 1147.

Surgical Surgery is required for a severe acute attack unresponsive to treatment, persisting active disease resistant to corticosteroids, symptomatic strictures, non-healing fistulae and complicating carcinoma. Most Crohn's patients will at some time require intestinal surgery. Limited resections and end-to-end anastomoses are performed as necessary, usually to relieve obstruction or fistula formation. Strictureplasties may relieve obstruction without the need for small bowel resection. Occasionally, panproctocolectomy with the formation of an ileostomy is indicated for severe colitis. Restorative proctocolectomy with pouch formation is generally thought to be contraindicated in Crohn's colitis. The complications of ileostomy are discussed on page 815. In patients with small intestinal resections, oral 5aminosalicylic acid preparations reduce the number of endoscopic recurrences.

ULCERATIVE COLITIS

1

Ulcerative colitis (UC) is a chronic inflammatory condition of unknown aetiology affecting the colon and rectum. Like Crohn's disease, it is most common in the western world (Fig. 15.24) and its incidence appears to be increasing (see p. 804). It is slightly more common in women than in men. Most patients present in early adult life, but all ages can be affected.

Aetiology and pathology Aetiological factors are similar to those for Crohn's disease (p. 808). There is evidence that relapses can be associated with stopping smoking cigarettes; the explanation for this is not known. UC primarily involves the mucosa; only in very severe disease is the muscularis mucosae penetrated. Microscopy shows plasma cells and neutrophils, particularly in the crypts of Lieberkiihn, which can break down to form 'crypt abscesses' communicating with the luminal surface. If the disease is confined to the rectum (proctitis) it is clinically less severe and there is no increased long-term risk of carcinomatous changes. The extent of disease varies from patient to patient, but in almost all cases the rectum is involved.

Clinical features Acute attack The usual presenting symptom is the passage of frequent loose, small-volume brown motions, with fresh blood and

TABLE 15,42 Ulcerative colitis: clinical features of mild, moderate and severe attacks Clinical feature

Mild

Diarrhoea Blood per rectum Mucus Abdominal pain Tenesmus Anorexia Weight loss Acute extracolonic manifestations Fever Tachycardia Anaemia Distended abdomen Abdominal tenderness Sigmoidoscopy

+ + + +

Abnormal plain abdominal X-ray High ESR High neutrophil count Hypoproteinaemia

+ + ± Fine granularity

+ + -

Moderate

15

Severe

++

+++

++

+++

++

+++

+

++

+

++

+

++

+

++

+

++

+

+

+

+

+

++

±

+

±

+

Coarse granularity Spontaneous bleeding -

Ulceration Pseudopolyps Mucopus +

+ + +

++ ++ ++

mucus either on the surface of the stools or intermixed with them. Blood per rectum alone or diarrhoea alone are common first symptoms. Attacks can be usefully divided into mild, moderate and severe (Table 15.42). Patients with mild attacks open their bowels up to five times a day, with moderate attacks up to 10 times a day, and with severe attacks over 10 times a day. The faeces become increasingly soft with more severe relapses. Cramping abdominal pains are common in attacks of moderate severity, whereas rectal pain and a painful feeling of incomplete evacuation (tenesmus) are common in the more severe cases. Fever and tachycardia indicate severe disease - usually widespread and active - rather than infection. Anaemia, if present, is usually hypochromic and microcytic and is due to blood loss. Sigmoidoscopy and rectal biopsy are essential for the assessment of proctocolitis. In proctitis the mucosa becomes oedematous, diffusely reddened and granular, when it loses its normal 'wet' appearance. The blood vessels and mucosal folds are lost, and contact bleeding and spontaneous haemorrhages may be present. 'Pseudopolyps' are islands of residual mucosa around which mucosa has been destroyed. Sigmoidoscopy in active proctitis should be brief; no attempt should be made to insert the instrument far into the rectum and only a little air insufflated. Colonic blood loss causes anaemia, whereas protein loss

813

TABLE 15.43 Investigation of an acute attack of presumed ulcerative colitis • • • • • • •

Sigmoidoscopy and rectal biopsy Full blood count and acute-phase reactants Liver function tests and plasma albumin Stool microscopy and culture Plain abdominal X-ray Barium enema or colonoscopy (delay until after acute attack) Radiolabelled white cell scan (selected cases)

leads to hypoproteinaemia and peripheral oedema. With acute colonic dilatation the patient becomes ill and febrile, with abdominal tenderness and reduced bowel sounds. Signs of peritonitis (abdominal rigidity, extreme tenderness, guarding and absent bowel sounds) suggest colonic perforation, a complication with a mortality rate approaching 50%. Emergency colectomy may therefore be necessary in acute colonic dilatation to avoid this complication. Chronic ulcerative colitis Following recovery from an acute attack, patients may return to a normal bowel habit. Acute exacerbations of varying severity usually occur intermittently. In some patients these periods of remission may be very short, so that treatment becomes almost continuous. Other patients have persistent symptoms, which typically become worse when their prednisolone dose is reduced. Development of malignant change Patients with total colitis for more than 10 years are at an increased risk of developing carcinoma of the colon, hence the need for regular colonoscopic screening. Non-intestinal manifestations The non-intestinal manifestations of UC are similar to those of Crohn's disease (p. 809 and Table 15.38).

Investigation The relevant investigations in UC are shown in Table 15.43. Stool microscopy and culture are necessary to exclude infection. A plain abdominal X-ray may demonstrate colonic dilatation, usually best seen in the transverse colon (Fig. 15.25). Barium enemas or colonoscopies can precipitate toxic dilatation and should therefore be avoided in acute disease. When used, they show loss of haustra, dilatation of the lumen and superficial ulceration. 1

1 814

Figs 15.26-15.28

FIG. 15.25 Plain abdominal X-ray showing a dilated transverse colon outlined by air (arrowed) - the toxic megacolon of colitis

Management Acute attack In mild attacks of distal disease, corticosteroid or 5aminosalicylic acid enemas are usually effective. In mild but more extensive disease, oral 5-aminosalicylic acid preparations or prednisolone are effective. In moderately severe attacks, oral prednisolone (30-40 mg daily) is the usual therapy, with the dose then being tapered off over 3 months. In severe attacks intravenous hydrocortisone (l00mgt.d.s.) is given. Antibiotics and total parenteral nutrition are indicated in sick toxic patients. Intravenous, followed by oral, ciclosporin can lead to remission in 70% of very severe cases not responding to intravenous corticosteroids, but relapse may occur on stopping treatment. 5-aminosalicylic acid compounds (e.g. mesalazine, 800mgt.d.s.) reduce symptoms in patients with mild attacks. Some patients benefit from the exclusion of dietary milk products. Maintenance treatment Numerous studies have shown that 5-aminosalicylate preparations reduce the relapse rates of patients in remission (see Table 15.41). Azathioprine can also be helpful. There is no indication for steroids in patients in remission. Management in pregnancy Most patients with inflammatory bowel disease improve during pregnancy, although a minority (about 20%) have relapses. There is no contraindication to the use of 5aminosalicylate preparations, azathioprine and prednisolone in any trimester. Patients tend to relapse

postpartum because of a fall in plasma steroid levels. They should be warned of this possibility and the disease treated vigorously if it occurs.

15

Management in childhood Children should be treated actively in the same way as adults to maximize growth. Prednisolone should be used sparingly because, although it reduces disease activity, it tends to inhibit growth. Particular attention should be paid to diet. In medically resistant disease a colectomy should be performed. Surgery The standard operation for UC has been panproctocolectomy with ileostomy. However, subtotal colectomy is now usually performed and, at a second operation, a pelvic pouch is constructed from the terminal ileum to act as a reservoir and anastomosed to the anus (restorative proctocolectomy). Inflammation of the pouch ('pouchitis') is a complication. There are three major indications for surgery: 1. Patients with severe acute disease, especially where the colon is dilated more than 6cm on plain abdominal Xray and the patient is toxic with a pyrexia and tachycardia. The operative mortality becomes much higher (50%) if patients perforate their colon and develop a faecal peritonitis (compared with 3% for elective and 10% for urgent operations). Acute severe colonic bleeding is also an indication for surgery. 2. Patients with chronic severe extensive disease which persistently requires high-dose corticosteroid treatment. Patients usually feel dramatically better after colectomy, which also avoids the development of further complications (apart from chronic liver disease). 3. In patients with total colitis for more than 10 years there is a significant risk of carcinomatous change (Fig. 15.26). Carcinomas may be multifocal and are often preceded by the development of epithelial cell dysplasia. After 10 years' total colitis, regular colonoscopies are therefore indicated. Severe dysplasia or carcinoma are indications for total colectomy. Complications of an ileostomy Complications of ileostomy are usually minimal, apart from minor problems with the abdominal wall around the stoma. There may be high fluid outputs (more than 1L/day), especially if there has been small bowel surgery or Crohn's disease. This can be helped by avoiding hyperosmolar food and by the use of drugs that slow intestinal motility (e.g. loperamide, codeine phosphate). Impotence due to damage to the sacral nerves is unusual. Dehydration, oliguria and hyponatraemia can occur in hot weather, with poor oral intake of fluids. Reduced urine output can lead to urate stones. However, most patients with an ileostomy have excellent health. Psychological problems can be severe and patients require sensitive counselling by medical and nursing staff. Many hospitals have full-time stoma nurses, who care for

FIG. 15.26 Barium enema findings with pancolitis Long-standing UC with disease throughout the colon, mucosal ulceration (white arrows at right-hand side) and loss of haustrations. There is an 'apple-core' lesion of the ascending colon (single white arrow) typical of carcinoma.

patients' equipment and basic medical and psychological needs, both in and out of hospital. Many patients also find it helpful to meet other stoma patients. Colitis and stoma societies and self-help groups are of great value in helping patients to lead a normal life. Prognosis Less than 1 % of patients suffering from UC die within a year of presentation. In contrast, after their first attack settles, 10% of patients will have at least 15 subsequent asymptomatic years. Up to 25 % of patients require colectomy within 5-10 years of developing the disease. Death can be avoided by performing colectomy before advanced carcinoma develops. Overall, the majority of patients have a normal life expectancy.

INFECTIVE, MICROSCOPIC, ISCHAEMIC AND RADIATION COLITIDES

PSEUDOMEMBRANOUS COLITIS Pseudomembranous colitis is usually due to a toxin produced by Clostridium difficile and nearly always follows the use of broad-spectrum antibiotics, e.g. amoxycillin, and cephalosporins. It is much more commonly hospital than

815

community acquired. The symptoms develop 4 days to 6 weeks after starting the drugs.

Clinical features and investigation The patient develops cramping abdominal pains and diarrhoea with mucus and, sometimes, blood. Sigmoidoscopy reveals diarrhoea, proctitis and, sometimes, multiple yellow elevated plaques (pseudomembranes) up to 20mm in diameter. Investigations may demonstrate Cl. difficile toxin in the faeces. A rectal biopsy is usually diagnostic and shows necrotic foci with inflammatory cells migrating through to the lesion ('the summit lesion') and pseudomembranes.

Management The treatment of choice is vancomycin for 7-14 days, but metronidazole is also effective. Colestyramine (cholestyramine), an anion-exchange resin, can be used to bind the toxin and reduce diarrhoea. Supportive measures, which often include intravenous fluids, are important. The prognosis is usually good, except in the elderly or if there is a delay in making the diagnosis. Relapse can occur.

tuberculosis is ingested to cause disease. The affected bowel becomes ulcerated and may alternate with normal bowel (skip lesions). The radiological and macroscopic appearances can be very similar to those of Crohn's disease. Histology shows caseating granulomata, and acidfast bacilli may be seen. Clinical features and investigations There is a gradual onset of abdominal pain, anorexia, weight loss, diarrhoea and fever. Intestinal obstruction, perforation and haemorrhage can occur, and if there is extensive small intestinal involvement malabsorption can be a feature. In tuberculous colitis there may be blood and mucus in the stool. Perianal tuberculosis may cause ulceration, fistula and abscess. Tuberculous intestinal lesions are shown on small bowel meal and/or barium enema. These are not specific, and so colonoscopy with biopsy or even laparotomy with biopsy is indicated to obtain a histological diagnosis. The differential diagnosis for small intestinal tuberculosis includes Crohn's disease, yersiniosis and lymphoma, and for large bowel tuberculosis, Crohn's disease, ulcerative colitis, infective colitis and carcinoma.

Tuberculous peritonitis YERSINIA Yersinia enterocolitica and pseudotuberculosis are organisms which may cause an acute ulcerative inflammation of the ileum, appendix and colon. They may be associated with erythema nodosum and an acute arthritis. Barium studies show a swollen terminal ileum with tortuous folds and filling defects, giving a cobblestone appearance or a colitis. Some patients may have a laparotomy because a mistaken diagnosis of acute appendicitis is made. The condition is self-limiting and resolves completely within 3 months. Tetracycline or cotrimoxazole are indicated in patients with severe systemic symptoms.

ABDOMINAL TUBERCULOSIS This is rare in the west but relatively common in developing countries, particularly the Indian subcontinent. Tuberculosis can affect any part of the gastrointestinal tract, but is commonest in the ileocaecal area, proximal colon and peritoneum. Fifty per cent of patients have evidence of previous or active pulmonary tuberculosis.

Intestinal tuberculosis

Tuberculous peritonitis is usually due to haematogenous spread from pulmonary tuberculosis and is rarely associated with intestinal tuberculosis. It presents with progressive ascites, accompanied later by anorexia, weight loss, fever and minor abdominal pain. Oedema, hepatosplenomegaly and generalized lymphadenopathy may appear late. A dry form of the disease can occur which presents with palpable abdominal masses due to multiple intra-abdominal adhesions. Other causes of ascites, particularly malignancy and cirrhosis, must be considered in the differential diagnosis. M. tuberculosis organisms are rarely seen in ascitic fluid (5%), which is an exudate and may contain numerous lymphocytes, but culture is positive in about 80% of cases. Sometimes diagnosis is made by biopsy and culture of peritoneal tissue obtained at laparoscopy or laparotomy. Management Standard antituberculous chemotherapy (see p. 645) leads to complete resolution of abdominal tuberculosis and should be continued for 12 months. Occasionally, longterm problems are caused by intestinal strictures or intraabdominal adhesions.

COLLAGENOUS (MICROSCOPIC) COLITIS

Either the human or the bovine form of Mycobacterium

1

816

Fig. 15.29

Collagenous colitis is a rare cause of chronic watery diarrhoea. Macroscopically the colon looks normal, but microscopically there is a thickening of the subepithelial layer by collagenous bands and an inflammatory cell infiltrate of the lamina propria. The aetiology is unknown, but some patients also have coeliac disease. Treatment can be

difficult, but regimens used in UC, such as maintenance mesalazine and courses of prednisolone for acute relapses, can help. Symptoms may last for months or years, but spontaneous resolution can occur.

GAY BOWEL SYNDROME Homosexuals have an increased incidence of diarrhoea from infections such as Shigella, Salmonella, Campylobacter, Giardia, amoebae and Cryptosporidium. They also commonly develop perianal warts (condylomata acuminata). These patients may have a proctitis due to gonococcal or herpes simplex type 2 infection and, occasionally, anal syphilis (see Chapter 9). The acquired immune deficiency syndrome (AIDS) can be associated with fulminant bowel infections as well as the development of plaques of Kaposi's sarcoma throughout the bowel (p. 449). Eating and swallowing may be painful owing to oral and oesophageal Candida infection.

ISCHAEMIC COLITIS Aetiology and pathology Ischaemic colitis usually affects the region of the splenic flexure or the rectosigmoid area. Rarely, ischaemia of the superior mesenteric artery may affect the ascending and proximal transverse colon. Atheroma is a more common cause than embolus, but aetiological factors are otherwise similar to those seen in small bowel ischaemia (p. 795).

Clinical features and investigation The patient is usually elderly and presents with bloody diarrhoea and abdominal pain. Examination shows mild peritonism and the rectal mucosa may be normal or show a non-specific proctitis. Sometimes a plain abdominal Xray shows a dilated colon with 'thumb printing', due to oedema of the bowel wall, which is confirmed by a barium enema. 1

Management In most patients the symptoms resolve with analgesia and intravenous fluids. Perforation is rare. A late complication is stricture formation.

RADIATION COLITIS Damage to the transverse colon may occur during upper abdominal radiotherapy and damage to the sigmoid colon and rectum during pelvic irradiation. In the acute phase

(during or immediately after treatment) the symptoms are bloody diarrhoea, tenesmus and abdominal pain. These are treated with a low-residue diet and rectal steroids if there is a proctitis. Later complications include the development of a chronic proctitis or colitis, with telangiectatic lesions which may bleed, benign ulcers and stricture formation. The proctocolitis and ulcers may sometimes improve with rectal or systemic steroids.

15

FUNCTIONAL BOWEL DISORDERS Incidence In the UK it is estimated that the prevalence is 10-22% of adults; 14-50% of these patients visit their GP with these symptoms. Patients with functional bowel disease make up 25-50% of new patients in gastroenterology outpatient clinics. These patients have abdominal symptoms with no evidence of underlying organic pathology. Many individuals do not seek medical advice, presumably because they understand the nature of their symptoms and may be able to attribute them to a particular cause; they thereby learn to avoid or live with them. Others will consult their GP, sometimes because of the discomfort experienced but often out of fear that the symptoms might indicate more serious disease. Some will then be referred to a specialist. This may be because of the severity or intractability of the symptoms, or because the presence of atypical features causes concern. More often, however, the referral will be made in order to reinforce the reassurance already given and, hopefully, to make treatment more effective. Specialists therefore tend to see a selected sample of patients who will often form the basis of clinical studies. Because the cause or causes are not known, and because the range of symptoms is wide and relates to the whole of the gastrointestinal tract, it is sensible to consider all the manifestations of functional bowel disorder as a single entity associated with multiple factors. At different times the same patient may experience either predominantly lower gastrointestinal symptoms (irritable bowel syndrome, IBS) or upper gastrointestinal symptoms (functional or non-ulcer dyspepsia). Functional bowel disorders commonly present in the third and fourth decades, with a slight predominance in women. In retrospect many patients can recall intermittent symptoms in childhood, often at times of stress, to which various labels may have been applied, e.g. bilious vomiting, grumbling appendix, abdominal migraine. Symptoms first presenting in the mid-40s onwards should be regarded more seriously, and as possibly indicating organic disease. Many of these older patients will none the less be found to have sigmoid diverticular disease, which has pathophysiological features in common with IBS. However, there is no convincing evidence that IBS in young adults progresses to diverticular disease in the middle-aged. Surveys of the general population show that symptoms of IBS are common: 30% of such a sample of 301 appar-

817

TABLE 15.44 Organic vs functional disorders in a sample of 2000 gastroenterology outpatients* Diagnosis

Number of patients

Definite non-organic Spastic colon Painless diarrhoea Painless constipation Depression and pain Anxiety and Gl symptoms Functional dyspepsia Others

888 449 107 39 50 24 77 59

Definite organic Peptic ulcer (Duodenal ulcer) (Gastric ulcer) Oesophagitis Carcinoma of stomach Postgastrectomy Inflammatory bowel disease (Crohn's disease) (Ulcerative colitis) Coeliac disease Infective diarrhoea Carcinoma of colon Gallstones Hepatitis Cirrhosis Alcoholism Other

980 197 (135) (62) 188 13 14 168 (57) (111) 26 21 28 48 11 21 16 229

* After Harvey, Salih and Read 1983. Lancet 1: 632-634. The study was based on 2000 new attenders at a gastroenterology outpatient clinic over a period of 5 years. Of these, 57 had a non-gastroenterological diagnosis and 75 remained undiagnosed.

ently healthy people had either abdominal pain, constipation or diarrhoea. When 2000 new attenders at a gastroenterology outpatient clinic in Bristol were prospectively studied over a 5-year period, about half turned out to have functional disorders. It is interesting to contrast these patients with the remainder found to have organic disease, as many of the organic conditions have to be considered in the differential diagnosis of functional bowel disorders (Table 15.44).

Aetiology Many aetiological factors have been postulated for functional bowel disorders. Whether the bowel is inherently irritable, or is normal until irritated, cannot be determined. No clear pattern of inheritance of functional bowel disorders has been demonstrated; although it may sometimes appear to be familial, this can probably be explained by common environmental factors.

1

818

Fig. 15.30

Motility disturbance 1 Hyperreactivity of the gut, both in the sense of provoking reflex contraction and of a lowering of the threshold to stimuli causing pain, has been demonstrated in many ways, including balloon distension studies. Pressure transducer studies have shown that some patients have increased frequency and strength of small intestine motility.

Previous gut infection The postinfective variant of IBS may comprise up to a third of all patients. Damage to the epithelium by microorganisms, their toxins or bile acids may allow stimulation of mast cells or paracrine cells initiating the syndrome. Dietary factors Many patients with IBS have diets high in refined carbohydrates and low in fibre. Food intolerance is often suspected by patients: in different studies its estimated frequency varies between 6% and 60%. Although exclusion diets, with recurrence of symptoms when the suspected food is reintroduced, would seem to support this possibility, relatively few double-blind studies have been undertaken. True allergy, in an immunological sense (using such criteria as a rapid reaction to small doses of antigen accompanied by raised IgE levels and positive skinprick tests) is exceedingly rare. The usually transient cows' milk intolerance of infants is the best-known example. Alternative non-immunological mechanisms have been postulated to account for some apparent reactions, which require large amounts of the suspect food and come on so slowly that the offending food may not be recognized. It has been suggested that changes in gut flora to give high counts of facultative anaerobes (analogous to changes produced by antibiotic therapy) may enhance the production of toxic metabolites from undigested food residues. Psychological The psychosomatic mechanisms postulated seem to be borne out by the frequently found evidence of stress and

SUMMARY 9 Contributory factors in functional bowel disorders • • • • •

Motility disturbance Altered sensory threshold Previous gut infections Altered gut flora Dietary Fibre deficiency Food intolerance Immunological ('allergy') Non-immunological Food additives • Psychological Anxiety Depression

CASE STUDY 15.6 ABDOMINAL PAIN AND CHANGED BOWEL HABIT IN A 47-YEAR-OLD MAN History A 47-year-old white Caucasian housewife presented to clinic with a 4-month history of altered bowel habit and worsening abdominal pain. Before her illness she opened her bowels once a day with a formed stool. Now her bowel habit was erratic, with a few days of constipation during which she passed hard pellet-like stools with difficulty once a day, alternating with several days of opening her bowel up to five times a day, with a semiformed stool containing mucus. She saw spots of blood on the toilet paper after straining when constipated. She never awoke at night to defecate, but on 'diarrhoea' days she opened her bowels several times during the first 2 hours after getting up, and had tenesmus. She had episodes of severe colicky left iliac fossa pain, which tended to occur before and was partially relieved by defecation. She had abdominal bloating, low back pain and urinary frequency. She had no upper gastrointestinal symptoms. Her appetite was poor but she had not lost weight. She felt tired. She had had a hysterectomy for uterine fibroids 5 months previously. There was no other relevant history and she was on no medication. Her mother had died of bowel cancer at the age of 62. Examination This revealed a tense, anxious but not ill middle-aged woman. Her abdomen was slightly distended with gas and her sigmoid colon was tender and palpable in the left iliac fossa. Questions 1. What is the most important disease that must be excluded? 2. What other disorders should enter the differential diagnosis? 3. What investigations are necessary?

Investigation The full blood count, biochemical profile, C-reactive protein, urinalysis, MSU and pelvic ultrasound were normal. Colonscopy showed a 5 mm stalked polyp in the sigmoid colon 30 cm from the anal margin. The rest of the colon, from anal margin to caecum, was normal. The polyp was excised by diathermy snare. Questions 4. What are the advantages of colonoscopy in the investigation of suspected colonic disease? 5. What types of polyp can be present in the colon? Histology showed that the polyp was metaplastic and not adenomatous. Biopsies from the rest of the colon were normal. Management A diagnosis of irritable bowel syndrome (IBS) was made. Question 6. Outline the management of

IBS.

The benign nature of the disorder was explained and she was advised to take a high-fibre diet. An antispasmodic, mebeverine, and additional fibre (sterculia) were prescribed. On review in 2 months she was a little better, although she still found the abdominal pain troublesome. She remained very tense and was concerned about her continuing symptoms. Amitriptiline 25 mg nocte was started and on review 2 months later she was much more relaxed and largely symptom free. She was more positive about the future. Six months later she was well off all treatment. Discussion In a middle-aged woman the presenting symptoms raised the

15

possibility of colonic cancer, particularly with the family history of the disease. She had blood on the toilet paper but the amounts were small and the nature of the bleeding suggested that it was from the anal canal rather than from higher up. The location of the pain and the type of bowel irregularity suggested a colonic disorder. The nature of the change in bowel habit, which did not include persistent diarrhoea, suggested that she did not have inflammatory bowel disease. There were no worrying symptoms such as weight loss or vomiting. Many of the features suggested irritable bowel syndrome (IBS), namely erratic bowel habit, increased frequency of defecation without liquid stools, mucus PR, tenesmus and early morning stool frequency. The absence of nocturnal defecation supported the diagnosis of non-organic disease. The easing of abdominal pain after defecation is seen in both non-organic and organic colonic disease. From middle life onwards colonic diverticulosis may occur, but the symptoms are similar to those of IBS and may have a similar aetiology. Patients with IBS may have non-gastrointestinal symptoms, such as malaise, tiredness, backache and urinary frequency. Headache, dysuria and dyspareunia can also occur. In IBS investigations are normal because the bowel is not diseased, although its motility is disturbed. In this case a colonoscopy was essential to exclude serious colonic disease, especially carcinoma. Sigmoidoscopy and barium enema are probably as good in the diagnosis of cancer, but in a patient with a family history of colon cancer identification of adenomatous polyps is essential, and so colonoscopy is preferred to barium radiology. Moreover, colonoscopy allows biopsy of lesions and the excision of polyps. The aetiology of IBS is unclear (see p. 818), but life events and the patient's psychological response to

819

CASE STUDY 15.6 CONTINUED

them may be important. This patient had had a hysterectomy just before the onset of IBS symptoms. Explanation and reassurance are vital in the management of IBS. The patient must realize that improvement is gradual and episodic, and is almost never rapid. The role of stress in precipitating and exacerbating symptoms must be explained. She was started on a highfibre diet, which may be beneficial in most cases of IBS, especially those with pure constipation or constipation alternating with increased stool frequency. In the early stages of treatment additional fibre in the form of sterculia is helpful, being less associated with gas

formation than methylcellulose or isphagula. However, in some patients diarrhoea, bloating and pain are exacerbated by extra fibre. She was started on an antispasmodic, although evidence for their benefit is largely anecdotal. The placebo response is a major feature in the treatment of IBS. In the present case the psychological state of the patient inhibited the response to treatment, and she improved greatly after the introduction of the tricyclic amitriptyline. This drug has been found to be of particular benefit in functional gastrointestinal disorders associated with anxiety and/or depression. Benzodiazepines were used in the past with benefit, but

anxiety. None the less, attempts to modify autonomic nervous activity pharmacologically have met with limited success, perhaps partly because the less accessible enteric nervous reflexes may be at fault. Much remains unexplained. It is not known what determines which of the various symptom patterns will develop in an individual patient, nor what makes some individuals accept the situation without recourse to orthodox, or even complementary, medical advice. A study screening consecutive non-complaining blood donors showed that the psychoneurotic symptomatology in the non-complaining IBS subjects more closely matched that of IBS patients than that of the remaining blood donors. It is not clear whether the differences between patients and the normal population are quantitative or qualitative.

Clinical features 1 The descriptive names given to functional bowel disorders indicate many of the symptoms - spastic colon, mucous colitis (in fact a misnomer, as the colon is not inflamed), and functional dyspepsia. Globus hystericus is difficulty with swallowing which has a non-organic basis. Most patients will experience alteration of bowel habit, many will have pain and bloating, and distension and excess flatus are common in the predominant lower gastrointestinal IBS (Table 15.45).

1 MCQ 15.17 820

their potential for addiction has severely limited their use. The prognosis for IBS is good for most patients: follow-up studies show that 80% of patients are well on no treatment 5 or more years later. The remainder have recurrent attacks or persisting symptoms Of varying severity. Good prognosis is associated with a short history, onset after bowel infection, constipation and male gender. The patient was found to have a metaplastic polyp at colonoscopy which, unlike adenomatous ones, has no malignant potential and is of no significance.

TABLE 15.45 Symptoms more commonly associated with functional than organic disease. The probability of functional bowel disease inireases the more these symptoms are present The criteria Ire better for women than men and are less predictive with increasing age Upper abdominal symptoms (non-ulcer dyspepsia) Oral flatulence and aerophagy Morning nausea and vomiting Symptoms worse with stress Intestinal symptoms (irritable bowel syndrome) Borborygmi Abdominal distension or bloating Pain relief with bowels open More frequent stools with the onset of pain Looser stools with the onset of pain Passage of mucus per rectum Increased rectal flatulence Sensation of incomplete rectal evacuation Urgency of defecation Straining to finish defecation Proctalgia fugax Oesophageal symptoms (irritable oesophagus) Globus

Irritable oesophagus This is a recurring retrosternal pain which may be severe and associated with globus, dysphagia, reflux-like symptoms or dyspepsia. Anxiety levels are high and angina is suspected, but cardiac and gastro-oesophageal tests are normal. Patients often have associated irritable bowel symptoms.

Non-ulcer dyspepsia (p. 782) Pain after food, with nausea, distension, belching and regurgitation, in patients who have had negative gastroscopy and gallbladder studies, are the hallmarks of nonulcer (or functional) dyspepsia, though the symptoms may overlap those of IBS. Unlike IBS, which does not progress to any other condition, up to 20% of patients with nonulcer dyspepsia may subsequently develop demonstrable peptic ulceration or oesophagitis. Fifty per cent of patients have Helicobacter pylori in their stomachs.

Pain It has long been recognized that pain arising from the gut is not always localized to the midline. In addition to 'classic' subumbilical griping, pain in many other sites and radiations, sometimes extra-abdominal, can occur. This can lead to a wide differential diagnosis (Table 15.46). Although not feasible as a routine diagnostic procedure, it can be useful to perform studies in which a balloon attached to a colonoscope is inflated at different sites as it is withdrawn from the caecum. This not only confirms that colonic distension can produce pain at both conventional and unconventional sites, but may also exactly reproduce a patient's pain. The severity and quality of pain may vary considerably, even mimicking biliary colic and other acute

TABLE 15.43 Differential diagnosis of typical and atypical IBS and functional dyspepsia Functional bowel disorder

Differential diagnosis

Typical IBS

Inflammatory bowel disease (Crohn's>ulcerative colitis) Colorectal cancer Diverticular disease Infective diarrhoea Malabsorption (coeliac disease) Endocrinopathies Thyrotoxicosis Hypothyroidism Carcinoid syndrome

Atypical IBS Right iliac fossa pain Right hypochondrial pain Lett hypochondrial pain Back pain Loin pain Menstrual exacerbation Dyspareunia

Appendicitis, Crohn's disease Gallstones Ischaemic heart disease, chronic pancreatitis Pancreatic disease Spinal osteoarthritis Renal disease Endometriosis Pelvic inflammatory disease

Functional dyspepsia

Peptic ulcer Oesophagitis Gastric cancer Ischaemic heart disease

Irritable oesophagus

Gastro-oesophageal reflux Oesophageal dysmotility Angina

abdominal emergencies. Palpable and tender bowel in the left iliac fossa is a relatively non-specific finding, but evidence of spasm (by feeling transmitted pressure waves in the reservoir bulb on sigmoidoscopy) or the reproduction of pain by the above procedure are particularly helpful in making a positive diagnosis. Proctalgia fugax is a transient shooting pain in the rectum, seen in IBS patients.

15

Constipation Constipation may be the predominant presenting symptom in functional bowel disorders, but often alternates with diarrhoea and may be associated with pain. The stools may be pellet- or ribbon-like and be commonly associated with the passage of mucus. Rectal bleeding must always be fully investigated, even if only to confirm that it is due to haemorrhoids. Although some patients may have large amounts of dry stool in the rectum with little awareness of its presence, others, despite not having a bowel action for several days, may have an empty rectum with greatly delayed colonic transit. Constipation dating back to childhood commonly emerges in the history, when 'potty training' may have been at fault, the urge to defecate repeatedly ignored, and purgatives regularly administered.

Diarrhoea Diarrhoea is often a distressing symptom because of the uncertainly and anxiety it may generate in an already anxious patient. The fear of faecal incontinence is a real one and, as the diarrhoea frequently occurs during the morning, the journey to work may be dreaded. Eating may precipitate the urge to defecate. The stool may vary between liquid and semiformed. There is often much flatus, and another uncertainty is thus whether the urge to pass wind may result in passage of diarrhoea and consequent soiling. The patient may be greatly relieved to be asked about this, as he or she may be too embarrassed to volunteer the symptom. A sense of wanting to evacuate when there is nothing there (tenesmus), or of incomplete defecation, is also particularly common. Although the patient may think large amounts are being passed, stool weights are in fact within the normal range. Diarrhoea occurs less commonly at other times of the day, but unlike inflammatory bowel disease or other organic diarrhoeas does not usually disturb sleep. Rumbling and gurgling (borborygmi) accompany diarrhoea and flatus.

Associated symptoms Other associated, non-specific symptoms may confuse the diagnosis. Sweating, tremor, hyperreflexia and tachycardia in association with diarrhoea are usually due to anxiety rather than thyrotoxicosis; a marked sinus arrhythmia, indicating autonomic overactivity, is frequently present. Gynaecological symptoms, e.g. increased severity of pain premenstrually, low back pain and dyspareunia, are common and often lead to unnecessary gynaecological

821

investigation. Headache, low back pain and fibromyalgia syndrome occur. Some patients have urinary symptoms, especially frequency ('irritable bladder'). Psychological symptoms are common, and underlying stress and disorders of mood and personality are frequently invoked as major factors in pathogenesis. None the less, the results of formal studies have often been conflicting. The perennial problem of whether anxiety provokes symptoms or vice versa cannot be clearly resolved, although the vicious circle of symptoms (anxiety -> more symptoms —> more anxiety) is real enough. Although depression may present with predominantly functional bowel symptoms (which regress with antidepressants), this is not as common as was once thought. Even if there were initially other reasons for a patient's anxiety, the symptoms frequently reinforce it. An important part in history taking is to ask the patient what they consider the cause to be. Frequently a fear of cancer, which they were reluctant to voice, will be revealed. Many will also be worried about inflammatory bowel disease, a worry perhaps reinforced by a previous vague diagnosis of mucous colitis or peptic ulcer.

Examination The patient often shows signs of anxiety, including sweaty hands and a tachycardia. There are no signs of masses or organomegaly. The abdomen may be diffusely and nonspecifically tender and the patient's response to the tenderness may be exaggerated. Sigmoidoscopy shows normal rectal mucosa, but air insufflation may reproduce the pain.

Prognosis Morbidity from functional bowel disorders can be considerable. There have been few long-term studies, but in one such, of 77 of 91 patients recalled after 6 years 29 had no bowel problems and many of the 44 who still had IBS symptoms found the condition easier to cope with. In view of the difficulties in making a positive diagnosis, it is notable that only four patients had a new diagnosis made, invariably prompted by a change in the clinical picture.

Investigation and management

822

The first problem that arises in the management of functional bowel disorder is the confidence with which one can make the diagnosis and the extent to which further investigation should be undertaken. The need for a full history and examination (including sigmoidoscopy) is self-evident; the history may be further supplemented at the follow-up appointment, which should be given to every patient. Many important psychological and social factors may then be disclosed to a doctor who has earned the patient's confidence. All investigations should be ordered at the first visit and the patient told that they are expected to be normal. To go on asking for 'just another test' counters the essential reassurance given. In hospital practice the stool

should be examined for pathogens and parasites if the history is of recent diarrhoea, especially after foreign travel. A barium enema or colonoscopy to exclude other colonic disease is not necessary in all patients, particularly if they are young with typical symptoms. Sometimes, however, the patient will require this investigation for complete reassurance. The most important diagnoses to exclude are inflammatory bowel disease in young adults and colorectal tumours in older patients. Patients with dyspepsia or chest pain may need upper GI endoscopy to exclude upper GI disorders and provide reassurance. The less typical the clinical presentation the more the exclusion of other diseases may be necessary, but a positive diagnosis should be made rather than arriving at one purely by exclusion. Such a minimalist approach is therapeutically effective and safe, with the proviso that fuller investigation or even reinvestigation is undertaken in the following circumstances: • Patients over 40 years of age presenting with symptoms for the first time; • Any change in the pattern of symptoms, or the development of new symptoms which sound organic; • The presence of additional symptoms which are not part of IBS, such as weight loss, rectal bleeding or fever. At the end of the first consultation the physician should be able to make a positive diagnosis and suggest an explanation for the cause of the symptoms. Even if the symptoms are bizarre, their existence should not be questioned. Patients can be specifically reassured that the physician does not think that they have cancer, peptic ulceration or intra-abdominal inflammatory conditions. The patient should be asked what they feared might be the cause, and then be reassured by pointing out how common the problem is and the generally good prognosis. After identifying any provocative factors, the doctor should suggest modifications in the patient's lifestyle that might be beneficial. The constipated may need to increase dietary fibre intake as well as adequate daily fluid. Smoking and excess alcohol consumption should be stopped and the effects of other medications reviewed. At a subsequent interview the results of normal tests can be used to reinforce the fact that there is no serious underlying cause for the symptoms. Patients should be encouraged to return if symptoms alter or complications such as weight loss develop.

Medication Drugs play only a minor role in management. When properly conducted, all drug trials in IBS and functional dyspepsia have shown very high rates of response to placebo (often over 50%). Antispasmodics Anticholinergic drugs may relieve pain but are limited in dosage by their side-effects. The directly acting smooth muscle relaxant mebeverine is widely used, and in some studies has been shown to be superior to placebo. An old

remedy, peppermint oil, in a new colonic-release capsule, is probably as effective. Opiate analgesics should always be avoided. Antidiarrhoeals Disabling morning diarrhoea can be treated with codeine phosphate on rising, repeated if necessary. If this is not effective, loperamide can be substituted. Constipation should be avoided. Laxatives and purgatives Only bulking agents should be encouraged. Many constipated patients may have been using purgatives for many years and these should be stopped, despite the fears and reluctance of patients to abandon them. Simple suppositories, or even periodic enemas, may have to be given temporarily while attempts are made to re-educate the bowel. Psychotropics Long-term use of benzodiazepine or other tranquillizers must be avoided, but they may be necessary as a short-term measure to combat periods of stress. Tricyclic antidepressants, especially amitriptyline or selective serotonin reuptake inhibitors (e.g. fluoxetine) may be helpful, particularly if diarrhoea is present, as the anticholinergic side-effect tends to have a constipating effect. Prokinetics The postprandial distension and nausea of non-ulcer dyspepsia is often greatly helped by metoclopramide syrup or tablets before meals. If this is ineffective, or if the drug is not tolerated because of somnolence or dyskinetic movements, domperidone should be substituted. Antiadd drugs. Simple antacids, H2-receptor antagonists or proton pump inhibitors may be helpful in patients with dyspepsia. Helicobacter pylori eradication This may benefit some patients with non-ulcer dyspepsia and Helicobacter in their stomachs. Calcium channel blockers and nitrates These may help some patients with oesophageal symptoms. 5 HT-3 antagonists These may benefit some patients with abdominal pain and diarrhoea.

SELF-INDUCED BOWEL DISEASE Self-induced bowel disease may be a variant of Munchausen's syndrome, where patients deliberately invent symptoms (p. 239), or be a result of drug abuse. Both problems are rare, and tend to occur in young females. The diagnosis is often well disguised. Invention of symptoms The patient who is inventing symptoms most commonly presents with severe abdominal pain or with gastrointesti-

nal bleeding (sometimes fictitious and sometimes simulated by the ingestion of animal blood). The history may be convincing but background information is vague. Examination often reveals an abdomen with multiple scars from when surgeons have been misled in the past. If this diagnosis is suspected, investigations should be minimal and the treatment is observation and reassurance. Psychiatric support rarely helps.

15

Laxative abuse The patient abusing laxatives usually presents with profuse watery diarrhoea secondary to purgative abuse. There may be several litres of watery diarrhoea a day, with consequent weight loss, dehydration and hypokalaemia. The diagnosis may be suggested by the appearance of pigment deposition in the rectal mucosa at sigmoidoscopy (melanosis coli). It may be confirmed by chemical analysis that purgatives such as phenolphthalein, bisacodyl, sennosides, danthrone or magnesium are present in the faeces or urine, or by finding the patient surreptitiously taking them. Confrontation should be avoided, and psychiatric treatment is often disappointing. Some patients may also abuse diuretics.

DISEASES OF THE PERITONEUM

PERITONITIS Acute bacterial peritonitis This follows perforation of an abdominal viscus or abdominal diverticulum as a result of ulceration, inflammation, ischaemia, trauma or (rarely) malignancy, or follows infection in the female genital tract. The patient develops an acute abdomen, with severe abdominal pain, nausea, vomiting, fever and tachycardia, followed by shock. Septicaemia and multiorgan failure can occur. On examination in the early stages there may be guarding, rebound tenderness and scanty bowel sounds. Later the abdomen becomes rigid and extremely tender, with absent bowel sounds. There is often a neutrophil leukocytosis and there may be metabolic acidosis. A plain erect abdominal or chest X-ray shows free gas under the diaphragm. Treatment requires a laparotomy if there is a ruptured viscus. Preoperatively the patient will need nasogastric suction, intravenous fluid and electrolytes, oxygen, opiate analgesia and broad-spectrum antibiotic therapy, including cover for anaerobic organisms (e.g. amoxicillin, gentamicin and metronidazole).

Primary (spontaneous) peritonitis This is acute or subacute bacterial inflammation of the peritoneum. It occurs in patients with cirrhotic or malignant ascites, immune deficiency states and after splenectomy, and in children with nephrotic syndrome or urinary tract infections. The commonest organism is Streptococcus pneu-

823

moniae. Pneumococcal vaccination or lifelong penicillin therapy will reduce the risk to splenectomized patients.

Tuberculous peritonitis (see p. 816) Chemical peritonitis develops when a sterile but irritant fluid, e.g. bile, blood or pancreatic secretions, leaks into the peritoneal cavity. Granulomatous peritonitis can be caused by infections (tuberculosis, fungi, parasites), secondary adenocarcinoma, sarcoidosis, Crohn's disease or foreign bodies (e.g. starch). Vasculitic peritonitis occurs when there is gut involvement in SLE, polyarteritis nodosa, systemic sclerosis, dermatomyositis or Henoch-Schonlein purpura. Familial Mediterranean fever is characterized by recurrent episodes of acute self-limiting serositis, especially involving the peritoneum (see p. 1263).

PERITONEAL MALIGNANCY Secondary carcinoma is the commonest cause of peritoneal malignancy. Most cases result from adenocarcinoma of the stomach, colon, pancreas, ovary, lung or breast. Primary mesothelioma is rare, slightly more common in men, and usually associated with asbestos exposure. About 50% of patients have signs of asbestosis on chest X-ray.

Clinical features and investigations The main symptoms are abdominal pain, nausea, vomiting, weight loss and abdominal swelling due to ascites. The ascitic fluid has a high protein content, may be bloodstained, and may contain malignant adenocarcinoma or mesothelial cells. Laparoscopy or laparotomy is sometimes necessary to make a diagnosis. Differential diagnosis includes tuberculous peritonitis, cirrhosis and Budd-Chiari syndrome.

Management and prognosis The prognosis is poor in secondary carcinoma and mesothelioma. Radiotherapy or chemotherapy may sometimes help in palliation of secondary carcinoma. Patients' symptoms may be helped by regular paracentesis.

FURTHER READING ON COLORECTAL DISEASE

824

Ferzoco L B, Raptopoulos V, Silen W 1998 Acute diverticulitis. N Engl J Med 338:1521-1526. Halligan S, Fenlon H M 1999 Virtual colonoscopy. Br Med J 319:1249-1252. Jones E M, MacGowan A P 1998 Back to basics in management of Clostridium difficile infections. Lancet 352:505-506. Kamm M A 1998 Faecal incontinence. Br Med J 316:528-532. Midgley R, Kerr D 1999 Colorectal cancer. Lancet 353:391-399. Present D H et al 1999 Infliximab for the treatment of fistulas in patients with Crohn's disease. N Engl J Med 340:1398-1405

Rodriguez L A G , Ruigomez A 1999 Increased risk of irritable bowel syndrome after bacterial gastroenteritis: cohort study. Br Med J 318:565-566 Satsangi J, Parkes M, Jewell D P, Bell J I 1998 Genetics of inflammatory bowel disease. Clin Sci 94:473-478.

THE PANCREAS The pancreas is a retroperitoneal organ which, along with the common bile duct, drains through the medial wall of the second part of the duodenum via the ampulla of Vater. The pancreas has exocrine and endocrine functions. Duodenal release of the hormone secretin results in acinar bicarbonate secretion. The pancreatic enzymes trypsin and chymotrypsin break down proteins, lipase digests fats, and amylase acts on starch. These enzymes are released in response to cholecystokinin and the activity of the parasympathetic nervous system. The endocrine cells of the pancreas are localized to the islets of Langerhans and make up about 1% by volume of the pancreas. The most numerous endocrine cells are the (3 cells secreting insulin but also present are a cells secreting pancreatic glucagon, D cells secreting somatostatin and PP cells secreting pancreatic polypeptide. With disorders of the pancreas, exocrine function is usually lost before endocrine function.

Clinical features of pancreatic disease The symptoms of pancreatic disease include pain (severe, epigastric or central abdominal, and often radiating through to the back), steatorrhoea (indicating lipase deficiency), weight loss, nausea, vomiting, and those of hyperglycaemia. Patients with acute pancreatitis may present with an acute abdomen. Other features include anaemia, jaundice due to common bile duct obstruction or associated liver disease, an abdominal mass, ascites and pleural effusion (usually left-sided).

Investigation of pancreatic disease General A full blood count is done to look for anaemia and for the white cell count (high in acute pancreatitis and abscesses etc., low in Schwachman's syndrome; see p. 827). A serum amylase is only indicated if a diagnosis of acute pancreatitis or continuing pancreatic obstruction due to an abscess or pseudocyst is being considered. Pancreatic proteases play a role in making vitamin B12 available for absorption in the terminal ileum, and a deficiency may result in megaloblastic anaemia. Three-day faecal fats are useful to assess the degree of steatorrhoea, and very high levels (more than 100 g fat per day) are typical of pancreatic insufficiency. Tube pancreatic function tests Tube pancreatic function tests involve the passage of a thin catheter into the second part of the duodenum. The duodenal contents are then aspirated for measure-

15

CASE STUDY 15.7 EPIGASTRIC BACK PAIN IN A 68-YEAR-OLD MAN A 68-year-old white Caucasian male coalminer presented with a 3-month history of epigastric pain. The pain had become increasingly severe and was worse with eating. It radiated through to his back. His appetite had become reduced and he had lost 12kg and now weighed 68kg. He was a lifelong smoker, smoking 15 cigarettes a day. He drank about 25 units of alcohol a week. Examination He looked thin and unwell. There was early jaundice and rectal examination showed him to have steatorrhoea. Examination of the heart and lungs was normal. Abdominal examination showed erythema ab igne (CD Figure 15.29) over the upper abdomen and epigastric tenderness. Question 1. What is the most likely cause of the pain and why? Investigations Haemoglobin 13.8 with a normal white cell and platelet count. Urea and electrolytes normal. Liver function tests showed a bilirubin of 85 umol/L with a cholestatic pattern of liver enzymes showing an alkaline phosphatase of 910 IU/L, an ALT of 105 U/L and -GT of 272 U/L.

ducts and gallbladder were all dilated. A CT scan confirmed the findings. Endoscopic retrograde cholangiopancreatography (ERCP) (CD Fig. 15.30) was performed and showed strictures in the head of the pancreas, with proximal dilatation of both the bile duct and the pancreatic ducts. An imaging wire was passed through the bile duct stricture and the stricture was brushed for cytology. A plastic stent was then placed through the bile duct stricture and produced good drainage of bile. Cytology of the brushings showed adenocarcinoma cells. Questions

3. What further investigations are needed? 4. What is the treatment of choice? A CT scan was performed (Fig. 15.31, p. 832) and showed that the tumour was invading the superior mesenteric artery and the portal vein. The patient was reviewed at a joint radiological and surgical meeting and it was decided that a Whipple's procedure to remove the tumour was not practical. Six weeks later the plastic stent was removed and replaced by an expanding metal stent to allow long-term relief from jaundice and itching.

Question

Question

2. What further investigations are indicated?

5. What further treatment is indicated?

An abdominal ultrasound showed a 3 cm solid mass in the head of the pancreas. The bile duct, intrahepatic

The aim of treatment in this patient is palliation. In particular, pain needs to be controlled by analgesics,

ment of bicarbonate and pancreatic enzymes, after stimulation by an intravenous pharmacological dose of cholecystokinin and secretin, or by a Lundh test meal, which consists of a 500 mL standard mixture of fat, protein and carbohydrate.

including opiates if necessary. Non-steroidal anti-inflammatory drugs may help. The patient needs to be advised about diet and given supplements if necessary. Patients with a blocked pancreatic duct have problems with fat malabsorption because of lack of pancreatic lipase. They may benefit from taking pancreatic enzyme supplements and substituting dietary fat with carbohydrate. Discussion Severe epigastric pain radiating to the back and associated with weight loss is commonly due to carcinoma of the pancreas in older patients. Epigastric pain after eating commonly occurs in patients with gastric ulcers, gastric cancers and pancreatic disease, whereas duodenal ulcer pain is often improved by eating. Alternative diagnoses are carcinoma of the stomach, benign gastric ulceration or chronic pancreatitis. All patients under the age of 80 should be considered for surgery, but advanced disease is usually present at the time of presentation. If patients are treated palliatively the prognosis is very variable and depends on tumour growth and the development of metastases. Some patients only survive a few weeks, whereas others can survive 5 years. When a pancreatectomy is performed for carcinoma of the head, body or tail of the pancreas, the 5-year survival rate is under 10% because the tumours spread early to adjacent lymphatics and the liver. The 5-year survival rate for carcinoma of the ampulla of Vater is higher, probably because they present earlier.

Tubeless pancreatic function tests Tube tests are uncomfortable for patients, and oral tests (Fig. 15.27) are fortunately now available. In the PABA test, pancreatic chymotrypsin cleaves N-benzoyl-

825

L-tyrosyl-p-aminobenzoic acid; the resulting paraaminobenzoic acid (PABA) is absorbed and the urinary PABA is measured in a 6-hour collection. To exclude possible intestinal malabsorption, a trace dose of 14C-PABA or para-aminosalicylic acid (PAS) is given at the same time and urinary radioactivity or PAS measured. The test is invalidated by paracetamol and cotrimoxazole and is dependent on normal gastric emptying and renal function. A similar test involves the cleavage of fluorescein dilaurate by pancreatic esterase, and subsequent measurement of urinary fluorescein.

FIG. 15.27 Tubeless pancreatic function tests A The PABA test involves chymotrypsin cleavage of N-benzoyl-tyrosyl and the subsequent measurement of PABA in urine or plasma. B The fluorescein test involves esterase cleavage and the subsequent measurement of fluorescein in urine. If either test is abnormal, the basic chemical (i.e. PABA and fluorescein) should be given on a separate day to exclude a problem with intestinal malabsorption.

Imaging techniques The pancreas is relatively difficult to image because of its retroperitoneal location. Plain abdominal X-ray is often unhelpful, but pancreatic calcification may be seen with chronic pancreatitis and a 'sentinel loop' of dilated duodenum may be seen over the pancreas in acute pancreatitis. Ultrasound, CT or MRI are useful in delineating abnormal anatomy and can be used to direct needle biopsies. Endoscopic retrograde pancreatography (ERP) is usually combined with cholangiography (ERCP). ERP is particularly useful in demonstrating abnormalities due to chronic pancreatitis or carcinoma. Angiography may help to diagnose pancreatic tumours and to assess operability.

SUMMARY 10 Investigations of pancreatic disease CONGENITAL PANCREATIC ANATOMICAL ABNORMALITIES

General Full blood count Liver function tests and serum proteins Serum amylase Blood glucose Serum B12 Three-day faecal fat excretion Tube pancreatic function tests Cholecystokinin-secretin test Lundh meal Tubeless pancreatic function tests PABA test Fluorescein dilaurate test Imaging techniques Plain abdominal X-ray Ultrasound CT scan ERCP Angiography

1

826

MCQ 15.18

2

MCQ 15.19

Embryologically the pancreas is derived from an anterior and a posterior bud, which normally fuse and drain into the medial wall of the duodenum. Developmental abnormalities include: 1. Accessory pancreatic tissue, which is separate from the body of the pancreas. This is found in about 10% of autopsies, with ectopic tissue most commonly located in the wall of the stomach and intestine. Rarely it may ulcerate and cause gastrointestinal bleeding, or be the site of carcinomatous change. 2. Annular pancreas, where the second part of the duodenum is encircled by pancreatic tissue. This is a rare abnormality which usually presents in the first months of life with duodenal obstruction; it requires a bypass gastrojejunostomy. Mild cases are asymptomatic, but the condition may present for the first time in adults with abdominal pain and vomiting, sometimes associated with peptic ulcer or obstructive jaundice. 3. Pancreas divisum, caused by failure of the two pancreatic ducts to fuse. Most of the gland is drained through the accessory ampulla. It is associated with an increased incidence of acute pancreatitis. Sphincterotomy of the accessory ampulla may reduce the number of attacks.

TABLE 15.47 Pancreatic and intestinal complications of cystic fibrosis Effect

Complication

Pancreatic insufficiency

Enzyme deficiency Diabetes mellitus Intestinal obstruction Volvulus Peritonitis following perforation Polyhydramnios in utero Intussusception Intestinal obstruction

Meconium ileus

Meconium ileus equivalent

CONGENITAL PANCREATIC FUNCTIONAL ABNORMALITIES Cystic fibrosis 1 Cystic fibrosis is the most important congenital disease of the pancreas. The pancreatic and intestinal effects of the disease are summarized in Table 15.47. Other complications include cirrhosis of the liver (Ch. 16), pulmonary disease (Ch. 13) and male infertility due to atrophy or absence of the vas deferens, seminal vesicles and epididymis. Cystic fibrosis is characterized by hyperviscous secretions containing excessive sodium, mucus and protein from exocrine glands, which result in the plugging of ducts leading to destruction of glandular tissue (Table 15.47). Diagnosis is made by demonstrating excessive sodium in sweat. Pancreatic insufficiency The destruction of pancreatic exocrine acini is often rapid. The patient develops steatorrhoea and oedema secondary to hypoproteinaemia. A good response is seen with the introduction of pancreatic enzyme supplements taken with each meal, and proton pump inhibitors and H2-receptor antagonists can help by raising duodenal pH to increase pancreatic enzyme activity. High-strength protease preparations can result in fibrous strictures of the colon. Supplements of vitamins A, D, E and K also have a role in the presence of steatorrhoea. With severe malnutrition, dietary supplements of easily absorbed medium-chain triglycerides and protein hydrolysates in the form of milkshakes or cooking additives may be helpful. Diabetes mellitus due to the destruction of pancreatic endocrine tissue is a late event. Meconium ileus In cystic fibrosis the meconium protein content at birth is about eight times the normal level. As a result, neonates with cystic fibrosis often present with intestinal obstruction, which may be complicated by a volvulus or peritonitis due to a perforation. Instillation of high-osmolarity

diatrizoate (e.g. Gastrografin) causes the plug to shrink and allows its passage down the gut. If a perforation or volvulus has occurred, surgery is usually needed. Fetal meconium ileus in utero causes maternal polyhydramnios.

15

Meconium ileus equivalent 'Meconium ileus equivalent' is, as the name suggests, a problem similar to meconium ileus, but it occurs later in life and is complicated by the presence of faeces. A highfibre diet to soften the stool, and laxatives, are effective in mild cases. N-acetylcysteine can be used as a mucusclearing agent. In severe cases diatrizoate is usually effective in shrinking the faecal mass; surgery should be avoided if possible, because of the high mortality from intestinal and pulmonary complications. Prognosis The prognosis is variable, but many affected children are now surviving into their 30s and 40s.

Macroamylasaemia Macroamylasaemia is a condition of unknown aetiology characterized by large molecular weight amylase in the serum. It accounts for 2.6% of elevated serum amylase levels. The amylase may be bound to immunoglobulins or glycoproteins, and is too large to be excreted by the glomerulus like normal amylase. Consequently, very high serum amylase levels occur. The diagnosis can be confirmed by chromatography and by showing low urinary amylase and normal plasma trypsin and lipase levels. It is important to make the diagnosis to avoid mistaking the condition for acute pancreatitis. There is an association with coeliac disease. No therapy other than reassurance is needed for macroamylasaemia.

Schwachman's syndrome Schwachman's syndrome is the association of pancreatic enzyme deficiency, neutropenia, short stature and metaphyseal dysostosis. Sometimes there is also anaemia, thrombocytopenia and low intelligence. It is inherited as an autosomal recessive. The usual presentation is with steatorrhoea and failure to thrive in young children; a normal sweat sodium and chloride excludes cystic fibrosis. The pancreatic insufficiency is treated with pancreatic enzyme supplements and H2-receptor antagonists or proton pump inhibitors. Infections should be treated aggressively in the presence of neutropenia, which is associated with a deficiency of white cell precursors. The prognosis is usually good.

ACUTE PANCREATITIS 2 Acute pancreatitis is an acute inflammatory condition of the pancreas which results in necrosis of exocrine tissue. It

827

causes abdominal pain and is associated with high plasma and urinary pancreatic enzyme levels. The incidence is 10 per 100000 per annum in the UK and North America, and in Copenhagen in 1970-80 was 28.1 per 100000 in people over the age of 20. The median age is 53 years, with an equal sex incidence.

Aetiology Pancreatitis occurs as a result of pancreatic duct obstruction, bile and duodenal content reflux and direct action of toxins. In the UK about 80% of cases are due to gallstones or alcohol abuse. In some parts of the world (e.g. South America) alcohol abuse is the most important factor. Gallstones are thought to block pancreatic enzyme outflow, but may be passed via the ampulla of Vater before investigations are undertaken. In this situation gallstones are usually found in the gallbladder or in the common bile duct. The mechanism by which alcohol causes acute pancreatitis is controversial. It may be due to a rise in pancreatic duct pressure or to a direct action on acinar cell membranes. ERCP may cause acute pancreatitis, possibly as a result of increased pressure in the pancreatic duct. Exploratory laparotomy, other intra-abdominal operations and translumbar aortography may also be complicated by acute pancreatitis. Patients with a Polya partial gastrectomy with a 'blind' duodenal loop are at increased long-term risk. Acute pancreatitis due to viral infection (e.g. mumps) is rare, and is caused by viral destruction of acinar cells. Pancreatic tumours may cause pancreatitis by ductular obstruction. Many drugs have been reported as possible causes. It is often stated that types I, IV and V hyperlipoproteinaemia and hypercalcaemia are causes of acute pancreatitis, but studies have suggested that these associations are doubtful.

SUMMARY 11 Aetiology of acute pancreatitis • • • • • • • • •

Gallstones in common bile duct Alcoholism latrogenic - usually post-ERCP Trauma Viral - mumps and coxsackie B Pancreatic tumours Drugs, e.g. corticosteroids, oral contraceptives, thiazide diuretics Possibly hyperlipidaemia and primary hyperparathyroidism Idiopathic

1

828

MCQ 15.20

TABLE 15.48 Complications of acute pancreatitis Early Clinical Cardiac failure Renal failure Respiratory failure (ARDS) Hyperglycaemia Anaemia Jaundice Disseminated intravascular coagulation

Biochemical Uraemia Hypocalcaemia Hypoalbuminaemia Abnormal liver function Hyperglycaemia Metabolic acidosis Hypoxaemia

Late Pseudocyst Abscess Pancreatic duct stricture Ascites/pleural effusion Diabetes mellitus

Clinical features The main symptom of acute pancreatitis is severe abdominal pain. Its usual site is epigastric or hypochondrial, but it may occur around the umbilicus. In about half the patients it radiates through to the back. The onset is usually sudden, such that the main differential diagnosis is intestinal perforation. The pain settles within 48 hours in most cases. The other common initial symptom is vomiting. Examination reveals a patient in great pain with severe abdominal tenderness, and often guarding and rigidity. Bowel sounds are usually reduced. Extra-abdominal signs may include fever, tachycardia and hypotension. In severe cases, complications (Table 15.48) may supervene.

Investigation The diagnosis is made by finding a serum amylase at least four times the upper limit of normal. A smaller rise in serum amylase may be seen in patients with perforated peptic ulcer, intestinal obstruction, other intra-abdominal emergencies, diabetic ketoacidosis, after opiate analgesia, macroamylasaemia and renal and hepatic failure. If the serum amylase is less than four times the upper limit of normal, or if it is thought that the high amylase may be due to macroamylasaemia, a high amylase:creatinine urinary excretion, serum lipase or serum trypsin may be used as confirmatory evidence of acute pancreatitis. A plain abdominal X-ray may show a 'sentinel loop' of air-filled duodenum lying over the pancreas. It may also show evidence of ascites or intestinal ileus, which may complicate acute pancreatitis. Ultrasound or CT may demonstrate an oedematous enlarged gland, gallstones, and complications such as pseudocysts. The differential diagnosis of acute pancreatitis includes gastric or intestinal perforation, peptic ulcer disease, acute cholecystitis, intestinal obstruction, acute mesenteric ischaemia, ruptured aortic aneurysm and myocardial infarction.

indicator of a severe attack and is independent of the Glasgow Score. Initial treatment is to restore fluid and electrolytes by intravenous infusion. A central venous pressure line is helpful to monitor fluid replacement. Analgesia should be maintained with pethidine or buprenorphine; morphine is contraindicated because it causes spasm of the sphincter of Oddi. A sedative such as diazepam may be helpful. Nasogastric suction is indicated initially, and should be continued until there is evidence that gastric emptying is normal and any intestinal ileus has resolved. Neither drugs such as cimetidine, anticholinergics, glucagon, aprotinin (a trypsin and kinin inhibitor), antibiotics, somatostatin and calcitonin, nor peritoneal lavage or parenteral nutrition, have been shown to have overall benefit in clinical trials. Early endoscopic sphincterotomy is indicated in severely ill jaundiced patients with stones in the distal common bile duct.

15

Complications of acute pancreatitis

FIG. 15.28 Management steps in acute pancreatitis

TABLE 15.49 Indicators of severe acute pancreatitis • • • • • • • • •

White blood cell count over 15 x 109/L Glucose over 10mmol/L Urea over 16 mmol/L Pa02 below 8.0 kPa Calcium below 2.0 mmol/L Albumin below 32 g/L LDH over 600 units/L AST over 100 units/L Over 55 years of age

Management (Fig. 15.28) Acute pancreatitis may have a self-limiting short course, be associated with the development of late complications, or be very severe and result in early death from circulatory collapse. In order to predict severity and prognosis, a number of indicators are helpful (Table 15.49); if more than three of these are present there is a risk of early death or complications (Glasgow coma scoring system). A serum CRP >210mg/L in the first 4 days is an

The treatment of these severe problems (Table 15.48) is supportive (usually in an intensive care unit), with fluids, cardiac inotropes, oxygen, ventilation, dialysis, calcium and insulin as necessary. Pseudocysts are collections of pancreatic juice in the lesser sac, and may need surgical or endoscopic drainage into the stomach. Pancreatic abscesses are usually due to Escherichia coli infection and require surgical drainage and antibiotics. Persistent high serum amylase levels suggest pseudocyst or abscess formation. Ascites and pleural effusions due to a fistula from the pancreatic duct system into the peritoneal or pleural spaces are characterized by very high amylase concentrations in the fluid aspirate. The effusions can often be treated with drainage via a cannula, but sometimes surgical closure is required. The overall mortality rate is about 10% per attack; this rises with age and other indicators of severity. The mortality of initial attacks is more than 10 times higher than that of recurrent attacks, and is twice as high with gallstones as with alcoholism. The recurrence rate is greatly reduced by successful treatment of gallstones or alcoholism.

CHRONIC PANCREATITIS 1 Chronic pancreatitis is associated with the destruction of pancreatic exocrine tissue which persists even when the underlying cause has been removed. Inflammation is followed by fibrosis and atrophy. Chronic relapsing pancreatitis is a variant in which symptoms, such as pain, are intermittent. The disease may be further classified according to its aetiology and whether steatorrhoea or diabetes mellitus is present. Steatorrhoea implies loss of lipase due to advanced acinar damage; diabetes mellitus indicates severe loss of islet tissue and advanced disease.

829

In 1978-79 the prevalence of chronic pancreatitis in Copenhagen was 27.4 per 100000 adult population per year, with a 3:1 male preponderance. The median age at onset was 39. The annual incidence in Minnesota in 1960-69 was 3.5 per 100000 population. The incidence rises in areas with high alcohol intakes.

SUMMARY 12 Aetiology of chronic pancreatitis • • • •

Alcohol Gallstones Hereditary Pancreas divisum

• Hyperlipidaemia • Hypercalcaemia • Idiopathic

Aetiology Alcohol is by far the most important factor in cases where a cause can be identified. Protein precipitates have been found in the pancreatic juice of alcoholics with chronic pancreatitis, and it is suggested that these form plugs which block the small pancreatic ducts and cause acinar damage. Gallstones are a common cause of acute pancreatitis but a rare cause of the chronic form. In the hereditary form of chronic pancreatitis the children in affected families usually present in adolescence. It is inherited as an autosomal dominant, but penetrance is variable. Pancreas divisum, which occurs when the ventral and dorsal embryological buds do not join, is a rare cause of chronic pancreatitis, as are hyperlipidaemia and hypercalcaemia. The disease may also occur in primary biliary cirrhosis, sclerosing cholangitis, and with choledochal cysts. In the tropics, chronic pancreatitis associated with malnutrition may be seen in young adults. In a significant number of patients no cause can be found.

Clinical features The main symptom of chronic pancreatitis is upper abdominal pain. 1 This is usually severe and chronic, although in about 10% of patients pain is not a feature. Radiation to the back is common, and sometimes the pain can be relieved by leaning forward. It may be exacerbated by eating. Anorexia, nausea and vomiting may be prominent. Steatorrhoea occurs with very high faecal fat output, and is associated with weight loss despite a good appetite. After 8 years of symptoms, about 70% of patients have Steatorrhoea. If there is gross Steatorrhoea deficiencies of vitamins A, B12 (see p. 824), D, E and K may occur. Coexistent alcoholic cirrhosis, or thrombosis of the portal or splenic vein as a complication of pancreatitis, may cause oesophageal or gastric varices. Gastric outflow obstruction due to swelling of the pancreatic head or a pseudocyst, and pancreatic ascites due to a fistula from the pancreatic duct to the peritoneum, occasionally occur. Biliary obstruction secondary to swelling of the pancreatic head is seen particularly in alcoholics, and can cause obstructive jaundice. About a third of patients develop diabetes mellitus. Calcific pancreatitis is associated with

1

830

Fig. 15.31

SUMMARY 13 Investigations for chronic pancreatitis Biochemical

Radiological

Faecal fat estimation and other tests for malabsorption Secretin-cholecystokinin test Lundh test PABA test Liver function tests Blood glucose Serum amylase (only useful in acute attacks)

Plain abdominal X-ray Ultrasound, possibly CT scanning Endoscopic retrograde pancreatography (ERP)

alcohol abuse and has a high risk of diabetes mellitus and Steatorrhoea.

Investigation Biochemical Serum amylase levels are not usually raised in chronic pancreatitis. Although faecal fat excretion may be very high (more than 100g/24h) this is not diagnostic of pancreatic disease, as amounts approaching this level may be seen in patients with small bowel disease. Low pancreatic enzyme and bicarbonate outputs can be measured by pancreatic function tests (p. 824). Mild abnormalities of liver function tests are common. Rarely, the head of the pancreas may swell and cause an obstruction to the lower end of the common bile duct; this results in jaundice and cholestatic liver function tests. Radiological A plain abdominal X-ray may show pancreatic calcification which is diagnostic (Fig. 15.29A). Ultrasound, CT and ERCP may show abnormal anatomy, including duct dilatation, calcification and tissue damage (Fig. 15.29B).

Management It is important, whenever possible, to treat the underlying cause, although this may not prevent progression and pain. Mild analgesia is rarely effective. Codeine, pethidine or buprenorphine may control pain, and destruction of the coeliac ganglion with injection of 100% alcohol may give relief. Surgery in the form of total or subtotal pancreatectomy should be considered in patients with severe pain, preferably before addiction to opiates has developed. However, the results are variable and often poor if patients

SUMMARY 14 Treatment of chronic pancreatitis • Treatment of aetiological problem. No alcohol • Pain Simple analgesics Avoid opiate analgesics (if possible) Coeliac ganglion block Surgery • Steatorrhoea Low-fat/high-carbohydrate diet Enzyme supplements H2-receptor anatagonists/Proton pump inhibitors MCTs/fat-soluble vitamins • Diabetes mellitus Diet Oral hypoglycaemics Insulin • Metabolic complications Protein supplements Vitamin B12

15

Prognosis The prognosis, if not the quality of life, of patients with chronic pancreatitis is generally good. Patients rarely die directly of complications and most survive 20 years from the onset. Pain may settle spontaneously; this may reflect 'burning out' of the pancreas and be complicated by Steatorrhoea and diabetes. Sometimes patients whose pain is poorly controlled become dependent on narcotics.

TUMOURS OF THE EXOCRINE PANCREAS FIG. 15.29 Chronic pancreatitis A Plain X-ray showing pancreatic calcification. B ERCP in chronic pancreatitis showing a dilated main pancreatic duct with associated dilated side branches.

continue to abuse alcohol. A minority of patients require regular methadone. Steatorrhoea and the resulting malnutrition may be helped by a low-fat, high-carbohydrate diet. Pancreatic enzyme supplements should be taken with food. They can be inactivated by acid, and therefore intraduodenal pH can be raised by H2-receptor antagonists or proton pump inhibitors. Food may be cooked in, or be supplemented by, midchain triglycerides (MCTs), as these can be absorbed directly without lipolysis. Supplements of vitamins A, D, E and K should be prescribed as necessary. Diabetes mellitus requires standard treatment. A particular problem is variable absorption of carbohydrate owing to exocrine deficiency or after surgical procedures, as this may result in poor control of blood glucose ('brittle diabetes'). Protein supplements may help raise plasma proteins, and parenteral vitamin B12 is occasionally required.

The overall incidence of carcinoma of the pancreas is about 10 per 100000 per year, but this rises to 100 per 100000 in those over 65 years of age (Fig. 15.30). It accounts for 5.5% of cancer deaths and is the fourth most common cause of such deaths in men. The incidence is rising throughout the world, with particularly high rates seen in Afro-Americans and New Zealand Maoris. In England and Wales in 1995-1997 there were 5000 deaths per year due to carcinoma of the pancreas.

Aetiology and pathology The aetiology is unknown. The disease is twice as common in cigarette smokers and in females with diabetes mellitus; chemists and non-oven coke workers also have an increased incidence. There is no definite association with alcoholism, coffee drinking or acute or chronic pancreatitis (except of the rare familial chronic type). However, there may be an increased incidence in chronic calcific pancreatitis. In cases where a histological diagnosis is established the tumour is almost always an adenocarcinoma, usually arising from the ductular epithelium. It may be mucin producing. About 5% of tumours arise from the acinar cells

831

FIG. 15.30 Age-specific incidence of pancreatic carcinoma per 100000 population in the UK It is commoner in males and the incidence increases with age.

and, rarely, sarcomas may occur. The tumour metastasizes to local lymph nodes adjacent to the vascular supply and, by invasion of blood vessels, to the liver and distant sites. Sixty-five per cent of tumours occur in the head of the pancreas, 30% in the body and 5% in the tail. The tumour infiltrates locally into the duodenum and the retroperitoneal space. It may involve the portal vein, causing thrombosis and splenomegaly, and may also seed into the peritoneal cavity causing ascites.

FIG. 15.31 CT scan of abdomen This demonstrates a dilated gallbladder (filled with contrast medium following an ERCP) and a large tumour mass in the head of the pancreas (arrowed).

SUMMARY 15 Symptoms and signs of carcinoma of the pancreas Symptoms

Signs

Weight loss Abdominal pain Jaundice Vomiting Steatorrhoea Pruritus Diabetes

Hepatomegaly Jaundice Palpable pancreatic mass Palpable gallbladder Thrombophlebitis Ascites

Clinical features Symptoms tend to occur late and are often non-specific. Consequently there may be a delay in diagnosis, and the majority of patients already have inoperable large tumours and/or metastases by the time the diagnosis is made. The most important symptoms are epigastric pain radiating through to the back, weight loss and jaundice. Jaundice and pruritus indicate obstruction to the common bile duct or multiple hepatic metastases. Left upper quadrant pain can occur in tumours of the tail of the pancreas. Vomiting may be due to gastric or duodenal invasion or peritoneal metastases. Haematemesis and melaena may also occur as a result of invasion of the stomach or intestine. Steatorrhoea may occur as a result of bile duct obstruction. The development of diabetes may predate all other symptoms. Hepatomegaly and/or jaundice are found in the majority of patients at presentation, and the coexistence of a palp-

1

832

Fig. 15.32

able distended gallbladder suggests the presence of a carcinoma obstructing the lower end of the common bile duct (Courvoisier's sign). This is, however, an unreliable sign, as it can also occur with cholelithiasis and is often absent with pancreatic cancer. Thrombophlebitis at remote sites, or deep venous thrombosis, occurs in about 10% of patients as a paraneoplastic syndrome (Ch. 6, p. 154). Carcinoma of the ampulla of Vater presents early with obstructive jaundice.

Investigation A full blood count and liver enzyme tests may show anaemia and an obstructive jaundice, but these are nonspecific and do not help to make the diagnosis. In experienced hands ultrasonography can be diagnostic; it also allows the passage of an ultrasound-directed needle to biopsy from the tumour for histology. CT scanning (Fig. 15.31) has an equal or greater accuracy and is of great help if ultrasound is negative or equivocal. Endoscopy may allow a direct view and biopsy of ampullary lesions, and

TABLE 15.50 Pancreatic endocrine tumours Peptide

Cellular source

Title

Clinical syndrome

Pancreatic glucagon

celI

Glucagonoma

Insulin Somatostatin

celI D cell

Insulinoma Somatostatinoma

Gastrin

G cell

Gastrinoma (Zollinger-Ellison syndrome)

Pancreatic polypeptide Vasoactive intestinal polypeptide

Pancreatic polypeptide cell Nerves

P Poma VIPoma (Verner-Morrison syndrome)

Adrenocorticotrophic hormone Parathyroid hormone Growth hormone releasing hormone

Unknown Unknown Unknown

Cushing's syndrome Ectopic hyperparathyroidism Pituitary acidophil hyperplasia

Necrolytic migratory erythema Diabetes mellitus Hypoglycaemia Diabetes mellitus Steatorrhoea Cholelithiasis Duodenal ulcers and malabsorption secondary to hyperacidity in the duodenum None Watery diarrhoea, hypokalaemia and achlorhydria (WDHA syndrome) Cushing's syndrome Hypercalcaemia Acromegaly

ERCP often demonstrates main duct lesions and gives information about the biliary tree. 1 In the case of ductular lesions, pancreatic juice may be aspirated for cytology. Laparoscopy and biopsy under direct vision are helpful in some difficult cases. Selective angiography may help to make the diagnosis and is important in determining resectability. Management The surgical treatment of carcinoma of the pancreas is very unsatisfactory owing to the organ's inaccessibility, the frequency of local spread and metastases. Radical surgery may be performed for localized tumours and a small number of patients are cured. The prognosis for ampullary tumours is considerably better than for those elsewhere in the pancreas. However, even in operable cases the 5-year survival is less than 15%. Hepatic metastases and portal vein thrombosis are contraindications for surgery. Nonsurgical palliation is now usual, the procedure of choice being to stent the biliary obstruction at ERCP to relieve jaundice. Alternatives are percutaneous stents or biliary drainage. For most patients with carcinoma of the pancreas, treatment is directed to symptomatic relief. Analgesics and coeliac plexus block may be needed. There are few cytotoxic drugs that are effective in pancreatic carcinoma, although responses are sometimes seen with alkylating agents, nitrosoureas and 5-fluorouracil.

PANCREATIC ENDOCRINE TUMOURS The islet cells of the pancreas are the primary site for a number of rare functioning endocrine tumours (Table

15

15.50), which may be benign or malignant and can produce characteristic syndromes. These tumours are important because surgical cure by resection may be possible. The symptoms caused can be predicted by the known physiological action of the hormone produced, the exception being pancreatic polypeptide, which is secreted by many of these tumours and can be used as a tumour marker. Multiple endocrine neoplasia type 1 (MEN type 1) is inherited on an autosomal dominant basis and is the association of hormone-synthesizing parathyroid adenomas (parathyroid hormone), pituitary adenomas (prolactin, growth hormone) and pancreatic endocrine tumours. MEN type 2a (see p. 938) is similarly inherited, and is the association of hormone-synthesizing parathyroid adenomas (parathyroid hormone), phaeochromocytomas (catecholamines) and medullary carcinoma of the thyroid (calcitonin). Occasional islet cell tumours are non-functioning.

Investigation Diagnosis of these tumours is by demonstrating a high plasma level of the appropriate peptide by radioimmunoassay. The tumours can often be demonstrated by ultrasound or CT scanning. On occasions it is necessary to sample blood along the portal vein and to demonstrate a peak of hormone secretion corresponding to the site of tumour venous drainage. Arteriography can also be used, and is helpful in demonstrating multiple primary tumours (common with gastrinomas) and hepatic metastases.

Management Treatment is surgical resection if possible, and this often results in cure. In the presence of metastatic disease the cytotoxic drugs streptozotocin and 5-fluorouracil can

833

produce worthwhile remissions. The symptoms of most of the tumours may be helped by somatostatin therapy. The rash of the glucagonoma syndrome can be improved by zinc, and the hypoglycaemia of insulinomas by diazoxide. The symptoms of gastrinomas are controlled by high-dose proton pump inhibitors, and the diarrhoea of VIPomas by a variety of agents, including corticosteroids, metoclopramide and trifluoperazine.

834

FURTHER READING ON THE PANCREAS Baron T H, Morgan D E 1999 Acute necrotizing pancreatitis. N Engl J Med 340:1412-1417. Glazer G, Imrie C W, Mann D V 1998 United Kingdom guidelines for the management of acute pancreatitis. Gut 42 (suppl 2):S1-S13. Mergener K, Baillie J 1997 Chronic pancreatitis. Lancet 350:1379-1385.

16

Liver and Biliary Tract Disease

disease and chronic hepatitis in countries with well-developed liver transplant programmes continue to die without having access to transplantation. The factors limiting the application of liver transplantation include cost, the limited availability of organs, and the lack of effective strategies to prevent the recurrence of alcoholic and viral disease.

STRUCTURE AND FUNCTION

John O'Grady

ANATOMY

Introduction 835

Chronic hepatitis 863

Structure and function 835

Primary biliary cirrhosis 866

Investigations in liver disease 839

Haemochromatosis 868

The jaundiced patient 845

Wilson's disease 869

Unconjugated hyperbilirubinaemia 846

Other causes of cirrhosis 870

Acute hepatitis 846 Drug-induced jaundice 852 Acute liver failure 853 Jaundice in pregnancy 855 Acute alcoholic hepatitis 855 Cholestatic jaundice 856 Cirrhosis 859

Complications of cirrhosis 871 Hepatocellular carcinoma 875 The liver in systemic disease 877 Non-viral infections involving the liver 879 Congenital biliary and hepatic disorders 881 Liver transplantation 881

Chronic alcoholic liver disease 861

INTRODUCTION Liver disease occurs worldwide and the majority of cases are due to either chronic viral hepatitis or alcohol. However, there is a considerable geographical variation. Hepatitis B is particularly common in the Far East, but there is some evidence to suggest that active vaccination programmes will reduce the prevalance of infection there. Alcohol is the dominant aetiology in the west, but it is now clear that viral hepatitis pays a significant role there too, in the form of hepatitis C. As yet, there are no major strategies aimed at reducing the incidence of liver disease in the west. The common liver diseases lead to premature death by means of liver failure, or the development of hepatocellular carcinoma. Liver transplantation is changing the profile of liver disease, but the impact is greatest in the west and it is having its greatest benefit in those diseases that are relatively uncommon. The majority of patients with alcoholic liver

The classic division of the liver into right and left lobes has now been complemented by the description of eight segments defined on the basis of blood supply. The caudate lobe is segment 1, segments 2-4 make up the left lobe, and the right lobe contains segments 5-8 (Fig. 16.1). The liver has a dual blood supply: from the hepatic artery, which is usually a branch of the coeliac axis, and from the portal vein, formed from the mesenteric and splenic veins (Fig. 16.2). The total liver blood flow is about 1300mL/min, i.e. one-quarter of the resting cardiac output. The hepatic artery supplies about 25% of the total liver blood flow but 50% of the oxygen; the remaining 75% of the volume and 50% of the oxygen comes from the relatively deoxygenated low-pressure portal vein supply. Both vessels run in the free edge of the lesser omentum to the liver hilum before dividing into major right and left branches. The hepatic veins drain directly into the inferior vena cava, which runs in a groove along the back of the liver between the functional right and left lobes. The three large hepatic veins (right, middle and left) enter just below the diaphragm, but there are a number of smaller veins draining the posterior aspect of the right lobe which enter the vena cava directly. These assume importance in two clinical situations: they may be the only remaining draining vessels if there is thrombosis of the main hepatic veins (the Budd-Chiari syndrome), and they are particularly difficult to dissect during a right hemihepatectomy. The biliary system collects bile, stores and concentrates it in the gallbladder, and delivers it to the duodenum when required. The main right and left hepatic ducts join in the hilum of the liver to form the common hepatic duct. The gallbladder lies on the undersurface of the right lobe of the liver, with its fundus usually projecting slightly below the lower edge. The cystic duct runs upwards and backwards to join the common hepatic duct, and the resulting common bile duct then runs with the hepatic artery and portal vein in the free edge of the lesser omentum to pass behind the first part of the duodenum and the head of the pancreas before turning to the right to enter the duodenum. This part of the duct is often completely buried in the pancreas, and it is here that it is susceptible to obstruction by pancreatic tumours. The pancreatic and bile ducts usually share a common channel for the last 2-7 mm, which opens through a small papilla on the inner wall of the

835

FIG. 16.1 The liver is divided into eight segments according to the vascular tree This subdivision has replaced conventional anatomy (see Fig. 16.2) for the purposes of planning hepatic surgery.

FIG. 16.2 A simplified view of the gross anatomy of the liver and its major relationships Hepatic veins (A); hilar lymph nodes (B); portal vein (C); splenic vein (D); pancreatic duct (E); bile duct (F); inferior mesenteric vein (G); superior mesenteric vein (H).

duodenum (the ampulla and papilla of Vater). In this part of the duct the choledochal muscle is thickened to form the sphincter of Oddi.

efferent veins. Although the concept of the acinus as a functional unit has proved useful in understanding patterns of liver injury, it is still useful for descriptive purposes to refer to the lobule as the classic structural unit of the liver. Each lobule is centred on an efferent (or central) vein, surrounded by radially orientated sinusoids and with five or six portal tracts at the periphery.

Microanatomy At a microanatomical level the liver consists essentially of a series of channels, or sinusoids, that run between plates of hepatocytes. These are lined by endothelial cells which contain unusually large fenestrations, allowing the sinusoidal part of each hepatocyte surface to have free access to almost all the constituents in plasma (Fig. 16.3). The space of Disse contains specialized cells, including Kupffer cells and stellate cells. The Kupffer cells are fixed macrophages and are an important barrier to gut-derived infections. The normal function of stellate cells is unclear, but they appear to be important in the pathogenesis of fibrosis. Specialized phagocytic cells (Kupffer cells) are also present in the sinusoids. Biliary canaliculi form between the contiguous surfaces of hepatocytes, and both here and on the sinusoidal surface there are extensive microvilli in the cell membrane to increase the capacity for transmembrane transfer. Branches of the hepatic artery, portal vein and bile duct within the liver are carried in portal tracts. The smallest of these, carrying terminal branches, supply groups of sinusoids which together form a functional unit, the acinus. Blood from each acinus then passes into a number of

1

836

MCQ 16.1

PHYSIOLOGY Essentially the liver has four quite distinct functions: • It supplies bile salts and bicarbonate to assist in digestion. • It acts as a buffer between the gut and the systemic circulation. • It maintains stable levels of amino acids and glucose in blood. • It synthesizes a large number of specialized proteins, carbohydrates and lipids. • It is a major excretory pathway for larger and more hydrophobic physiological metabolites, foreign substances and drugs.

Bile acid metabolism and secretion The primary bile acids, cholic and chenodeoxycholic acids, are synthesized from cholesterol in the liver, conjugated with taurine or glycine to form bile salts, and then excreted by an active transport mechanism into the bile canaliculi and eventually into the duodenum. In the distal ileum there is another active transport system which ensures the conservation of bile salts when they have fulfilled their

16

FIG. 16.4 Enterohepatic circulation of bile acids Bile salts entering the duodenum via the bile duct (A); absorption of bile salts in the terminal ileum (B); secondary bile acids formed by bacterial degradation (C); uptake and re-excretion of bile salts by liver (D).

FIG. 16.3 The microanatomy of the liver A The classic liver lobule is centred on the central vein with the portal tracts at its periphery. The concept of the functional lobule emphasizes the portal tracts as the points of entry of the arterial and portal venous blood, with the sinusoids then extending from there to the 'central veins'. B A liver sinusoid with liver cell plates separated from the lining endothelial cells by the space of Disse. The Kupffer cells straddle the blood-filled centre of the sinusoid. C The endothelial cells, riddled with fenestrations, allow macromolecules to pass between the blood and the hepatocytes. The biliary canaliculi run between the hepatocytes in the liver cell plates.

digestive function (Fig. 16.4). This system of enterohepatic circulation is extremely efficient. The entire body pool of bile salts is circulated through the gut about twice during each main meal, and only about 5% is newly synthesized in the liver. The small fraction reaching the colon is modified by bacteria, which perform 7a-dehydroxylation, producing the secondary bile acids deoxycholic and lithocholic acids. All bile salts are powerful detergents, solubilizing lipids by enclosing them in bile salt aggregates called micelles. These are organized so that hydrophobic groups are orientated into the lipid-rich interior whereas the hydrophilic hydroxyl and carboxyl groups are on the outside. Cholesterol and lecithin in bile itself are carried in such micelles, and if the concentration of these molecules

rises too high relative to the concentration of bile salts, the bile becomes supersaturated. Precipitation of cholesterol may then occur, particularly if a nidus for crystal formation is present. This is the pathophysiological basis for gallstone formation, which is discussed further on page 856. The active transport of bile salts is the main driving force for bile secretion and its failure may be an important cause of intrahepatic cholestasis. Bile salts accumulate in cholestasis, whether owing to mechanical obstruction or to a defect in excretion, and deposition in the skin may account for the pruritus which is often a troublesome symptom. There is also a bile salt-independent contribution to bile secretion, which is probably driven by active sodium transport in the distal canaliculus. Bicarbonate is added in the small bile ductules and this process is stimulated by the hormone secretin. Cholecystokinin principally stimulates gallbladder contraction. 1

Protein metabolism The liver receives amino acids from the gut and the muscles and regulates their levels in plasma by controlling the rate of gluconeogenesis and transamination. It also converts ammonia into urea via the urea cycle. Ammonia is produced intrahepatically by deamination, in other tissues by nucleotide metabolism, and in the colon by bacterial action on proteins or urea. The liver is the major site of synthesis of almost all the plasma proteins, and for many is also the principal site of degradation. In an average 70kg adult total body protein is about 12kg and turns over at a rate of about 250g/day. There is a close relationship between the muscles, liver and gut with

837

SUMMARY 1 Liver physiology Normal function Bile acid metabolism Glucose and amino acid homeostasis Synthesis of plasma proteins, clotting factors, lipids Metabolism/excretion of bilirubin, ammonia/urea, drugs

Consequence in disease Fat malabsorption Encephalopathy Ascites, bleeding, xanthelasmata Jaundice, encephalopathy, drug toxicity

FIG. 16.5 Amino acid and protein metabolism in health Branched-chain amino acids (BCAA), marked in blue, constitute only 20% of the dietary intake, whereas aromatic amino acids (AAA), marked in black, make up 80%. The liver preferentially metabolizes AAA rather than BCAA, which are passed to the tissues (particularly muscles). The thickness of the blue and black lines represents the relative amounts of the amino acids in the different body compartments.

838

respect to amino acid fluxes (Fig. 16.5). Muscle contributes the biggest source of protein turnover - about 130g/day whereas dietary intake is about 90g/day. Only 23% of amino acid nitrogen absorbed into the portal bloodstream after a meal is passed on to the peripheral tissues, and in exercising this gate function the liver is responsible for extensively modifying the blood amino acid composition. The aromatic amino acids - phenylalanine, tyrosine and methionine - are preferentially processed to urea, whereas the branched-chain amino acids - valine, leucine and isoleucine - are selectively passed to the periphery, where they are predominantly metabolized by muscle. This selectivity is lost in severe liver disease, and it is possible that the resulting change in the ratio of the plasma concentrations of these two groups of amino acids may alter cerebral neurotransmitter metabolism and contribute to the pathophysiology of portosystemic encephalopathy (p. 873). Of the many proteins synthesized in the liver, albumin and prothrombin are of particular interest in relation to

liver disease and are discussed as liver function tests on page 839. Because albumin is also the most important mediator of the plasma colloid osmotic pressure, the fall in albumin concentration that accompanies chronic liver disease is relevant to the pathophysiology of ascites, one of the frequent complications of cirrhosis (p. 871).

Carbohydrate metabolism The liver acts as a gate for the large amounts of glucose delivered from the gut during digestion, by storing it as glycogen, and is also the main source of glucose during starvation. Initially, glycogenolysis is the principal metabolic pathway used, but as the glycogen stores become depleted, gluconeogenesis assumes a greater importance. The main substrates for this pathway are lactate, pyruvate, glucogenic amino acids (mainly alanine from muscle) and glycerol (from lipolysis in fat stores). In the fed state about 160g/day of glucose are needed for those tissues with an obligatory requirement for this substrate, principally the brain and haemopoietic tissues. During prolonged starvation the brain adapts to ketone bodies as an alternative energy source, and the total body requirement for glucose drops to about 40g/day. Although blood glucose levels may directly influence carbohydrate metabolism in the liver, the principal instruments for metabolic control are the hormones insulin, glucagon, catecholamines and glucocorticoids (Fig. 16.6). The fall in insulin levels during starvation releases glycogenolysis and gluconeogenesis from the inhibitory effects of this hormone, with glycogenolysis being particularly sensitive. The increasing glucagon and catecholamine levels act directly to stimulate glycogenolysis. In the periphery, the relative excess of glucocorticoids compared to insulin leads to increased release of amino acids into the circulation, and high glucagon levels stimulate alanine uptake by the liver. In liver disease the supply of glucose by gluconeogenesis is generally well preserved, and hypoglycaemia is only usually a problem in severe (acute) liver failure. Glucose intolerance is more common, but is rarely a clinical problem.

Lipid metabolism The major plasma lipids, cholesterol, cholesterol esters, phospholipids and triglycerides, are highly insoluble in water and are carried in macromolecular complexes of lipid and a protein carrier (apoprotein) called plasma lipoproteins. Only the intestine and the liver synthesize and secrete plasma lipoproteins. There are four main classes, which differ in size and density and have different proportions of the main lipids: • Chylomicrons are made in the mucosal cells of the small intestine during dietary fat absorption, and are the main carriers of triglyceride. • Very low-density lipoproteins (VLDL) are produced in the liver and intestine and have a core which contains

TABLE 16.1 Non-hepatic causes of impaired liver function tests Cause

Nature of abnormality

Haemolysis

Isolated hyperbilirubinaemia

Adolescence, Paget's disease, osteomalacia, bone tumours

Isolated elevation in (bone) alkaline phosphatase

Enzyme-inducing drugs (and alcohol)

Increased -glutamyl transpeptidase (GT) levels

Malnutrition, maiabsorption, protein loss and chronic illness

Decreased serum albumin

Fat maiabsorption, disseminated intravascular coagulation (DIC), warfarin therapy

Prolonged prothrombin time

Myocardial infarction, myositis

High serum aspartate aminotransferase (AST) levels

16

Drug metabolism

FIG. 16.6 The main routes of carbohydrate metabolism in the fed state and fasting

50-60% triglyceride. Their apoprotein content is low (about 10%). • Low-density lipoproteins (LDL) are formed in the plasma by catabolism of VLDL. They have a higher protein content (22%) and lower triglyceride proportion (10%) than VLDL. • High-density lipoproteins (HDL) are made to some extent within the plasma, but are also synthesized in the liver and intestine. They have the highest protein content (50%) and lowest triglyceride proportion (5%). Hypercholesterolaemia is a feature of any liver disease in which there is obstruction to bile flow, whether this is intra- or extrahepatic. The increase is due to a rise in free cholesterol levels. Although total plasma phospholipids are often normal, lecithin levels are increased as a result of regurgitation back into the blood from the obstructed biliary tree. Although the elevated free cholesterol levels may also be due to regurgitation, there is some evidence to suggest that the increased plasma lecithin stimulates hepatic cholesterol synthesis. A marked increase in LDL levels, and the appearance of an abnormal lipoprotein, lipoprotein X, is due to the lipids providing increased substrate drive. Among the clinical effects of these lipid changes are the yellowish plaques of cholesterol deposited in the skin (xanthomata) in long-standing biliary obstruction, and the appearance of 'target' red cells in the peripheral blood film, probably a reflection of altered membrane fluidity.

The liver is an important site for the clearance of many drugs. In general, the more lipophilic and the higher the molecular weight, the more likely the drug is to be cleared by the liver rather than the kidney. Such hepatic clearance usually involves metabolic biotransformation with or without conjugation. In liver disease, alteration of drug handling can have clinically important consequences (Ch. 1).

INVESTIGATIONS IN LIVER DISEASE LIVER FUNCTION TESTS Blood tests often described as 'standard' liver function tests are not true tests of liver function but reflect present or previous liver cell damage and biliary obstruction. Although they are useful in differential diagnosis and may be the first indication of liver disease, they are rarely of diagnostic value alone. Most of these tests are not specific indicators of liver damage, and the influence of disease in other organs needs to be carefully considered when evaluating their significance (Table 16.1).

Serum bilirubin The main source of bilirubin is the reticuloendothelial system, where haemoglobin from effete red blood cells is split to haem, then to bilirubin. This is lipophilic and is carried in the plasma, tightly bound to albumin. Uptake into the hepatocytes is mediated by a carrier system in the liver cell membrane, and by ligands in the cytosol (Fig. 16.7). Bilirubin is then converted to a water-soluble

839

TABLE 16.2 Clotting factors synthesized in the liver Fibrinogen Vitamin K-dependent factors (II, VII, IX and X) Labile factors (V and VIII) Contact factors (XI and XII) Fibrin-stabilizing factor (XIII)

SUMMARY 2 Liver function tests All non-specific, so consider disease in other organs Synthesis Albumin: long half-life, so slow response = chronic liver disease Prothrombin time: short half-life, so fast response = acute and chronic liver disease Excretion Bilirubin: conjugated bilirubin elevated in liver disease, whatever the cause FIG. 16.7 The formation and metabolism of bilirubin The main sites of congenital defects leading to jaundice are indicated. Failure of uptake into hepatocytes: probably the main abnormality in Gilbert's syndrome (A). Failure of conjugation, the main abnormality in more serious cases of unconjugated hyperbilirubinaemia in childhood (e.g. Crigler-Najjar syndrome) (B). Failure of excretion can occur in some rare congenital causes of conjugated hyperbilirubinaemia, such as Dubin-Johnson syndrome (C).

compound by conjugation with glucuronic acid, and excreted as the diglucuronide into bile. Conjugated bilirubin refluxing back into the blood can be excreted by the kidney, but the water-insoluble unconjugated bilirubin cannot. Hence jaundice without bilirubinuria, indicating high serum levels of unconjugated bilirubin, is a feature of haemolysis (increased formation), failure of bilirubin uptake, or conjugation in the liver. Excess conjugated bilirubin in serum occurs with cholestasis. In hepatitis there is intrahepatic cholestasis in addition to liver cell injury, but conjugated bilirubin is found in the plasma as conjugation is preserved even with severe liver damage. Tests reflecting synthetic function

Serum albumin Serum albumin levels are used to assess the liver's ability to perform a high-volume synthetic function. The half-life of albumin is about 21 days, and inability to maintain serum albumin levels between 35 and 50 g/L indicates impending liver failure in the context of chronic liver disease. Serum albumin levels are often normal in acute liver failure. 1

840

Case 16.1

Alkaline phosphatase: sensitive indicator of cholestasis -GT: not so sensitive but elevated with enzyme induction, including alcohol Cell damage AST/ALT: released from damaged liver cells = active liver inflammation

Other conditions causing low serum albumin levels need to be considered; they include urinary loss and, less commonly, severe malnutrition or hypercatabolism. Coagulation studies Analysis of coagulation factors is used to assess the liver's ability to synthesize important proteins immediately. The half-life of coagulation factors can be as short as 7 hours, and consequently these tests are used to identify massive liver injury, as seen in acute liver failure. The one-stage prothrombin time (Table 16.2) is a functional assessment of several coagulation factors and is widely used in acute and chronic liver disease. It is subject to significant methodological variation between laboratories and countries, and some aspects of this are overcome by reporting the result as an international normalized ratio (INR). Individual clotting factors can also be measured and, of these, Factor V assays are particularly useful. In chronic liver disease the prothrombin time or INR tends to parallel changes in serum albumin levels. In cholestasis the prothrombin time is often prolonged, even though synthetic liver function is intact. This is because several of the factors evaluated in the test are dependent for their activity on vitamin K, one of the fat-soluble vitamins which is poorly absorbed when bile salts are not reaching the intestine. Giving vitamin K parenterally

(10 mg daily for 3 days) is therefore a clinically useful test of liver synthetic function. The prothrombin time corrects completely in cholestasis but remains abnormal in severe hepatitis or chronic liver disease.

Tests reflecting cholestasis Serum alkaline phosphatase Different isoenzymes are found in bone, liver, intestine and placenta, and serum electrophoresis can be useful to identify the source of elevated levels. The liver isoenzyme is located mainly in the bile canalicular region of the hepatocyte surface membrane, and high serum levels are a sensitive indicator of segmental or global biliary obstruction, whether within or outside the liver. Much of the rise is due to increased production of the enzyme in areas of the liver proximal to an obstruction, together with reflux of a high molecular weight form of the liver type of isoenzyme, the 'biliary isoenzyme'. This is associated with vesicular structures which may be actual fragments of the canalicular membrane. Although high plasma bilirubin levels often accompany high alkaline phosphatase levels, bilirubin values can be normal in spite of markedly raised serum alkaline phosphatase levels. This usually indicates a patchy or incomplete obstructive lesion, with bilirubin being excreted by the non-obstructed areas. This pattern is a feature of space-occupying lesions in the liver, e.g. granulomata, cysts and abscesses, as well as secondary deposits. Serum y-glutamyl transpeptidase Serum y-glutamyl transpeptidase (y-GT) is a more liverspecific enzyme than alkaline phosphatase and also reflects cholestasis. However, it is less sensitive to biliary obstruction, and an important clinical point is its sensitivity to enzyme-inducing drugs, such as phenytoin, and alcohol, both of which can cause quite marked elevations in serum levels in the absence of significant liver damage. A useful practical guide to interpretation is to compare the levels of y-GT and alkaline phosphatase in serum. In cholestasis the alkaline phosphatase is usually higher than the -GT, whereas the reverse is usually true in alcoholic liver disease or in patients taking enzyme-inducing agents. 1

tion of the circulatory disturbance often results in a very rapid fall in AST levels.

16

Serum alanine aminotransferase Although more liver-specific, changes in levels of serum alanine aminotransferase (ALT) tend to parallel those of serum AST. However, ALT levels tend to be somewhat higher than AST levels, and this may be important when screening for necroinflammatory activity in patients with chronic hepatitis. ALT is predominantly a cytosolic enzyme, whereas 80% of AST is located in mitochondria.

SPECIAL INVESTIGATIONS Imaging techniques Ultrasound scanning Ultrasound scanning is often the first specialized investigation in patients with liver disease. It enables a rapid assessment of liver and spleen size, vessel size and patency, and the calibre of bile ducts (Fig. 16.8), cysts and tumours. It is particularly good at detecting gallstones, as they are very echogenic. Ultrasound is also widely used to screen for hepatocellular carcinoma in patients with cirrhosis. The sensitivity and diagnostic accuracy of ultrasound has been increased by new techniques enhancing patterns of vascularity. Liver scintiscanning This is now nearly obsolete in the investigation of liver disease. Liver scintiscanning is obtained by gamma camera imaging of the upper abdomen after the intravenous injection of technetium-99m-labelled sulphur colloid, which is taken up by Kupffer cells in liver sinusoids. Liver and spleen size can be accurately assessed. In cirrhosis,

Tests reflecting liver cell injury Serum aspartate aminotransferase Serum aspartate aminotransferase (AST) is found in particularly high levels in the liver and in skeletal and cardiac muscle; serum levels can rise after injury to any of these tissues. High levels in liver disease are classically associated with liver cell necrosis in acute or chronic hepatitis, but membrane damage without cell death may also allow enzyme to leak out into the circulation. Thus, high levels of serum AST can occur in conditions such as acute left ventricular failure and septicaemia, in which hypoperfusion and anoxia lead to transient liver cell injury. Correc-

FIG. 16.8 Ultrasound of the liver demonstrating dilated intrahepatic and common bile ducts

841

effective hepatic (sinusoidal) blood flow is reduced, and other parts of the reticuloendothelial system, notably the spleen and bone marrow, show markedly increased isotope uptake. Acute alcoholic hepatitis and Wilson's disease may yield a 'whiteout' because of complete failure of uptake of colloid by the Kupffer cells. Computed tomography (CT) CT is capable of providing high-resolution sectional views of the liver and surrounding structures in the upper abdomen (see Figs 16.22, 16.28B). Contrast enhancement is of value in further defining the nature of abnormal areas. The technique is best used in cases where ultrasound has failed to provide the expected morphological information, and for the examination of retroperitoneal structures. Magnetic resonance imaging (MRI) The role of MRI in the evaluation of liver disease is increasing (Fig. 16.9). It frequently complements the ultrasound and CT scans in the assessment of focal lesions, particularly since the availability of contrast-enhancing techniques, e.g. with manganese. MR cholangiography (MRCP) is being increasingly used to define anatomy and pathology, and MR angiography and venography (Fig. 16.10) are rapidly being developed.

FIG. 16.9 MR scan outlining a large cyst in the liver

Positron emission tomography (PET) The role of this emerging investigative technique has yet to be defined, but it may prove valuable in the management of patients with malignant disease. Contrast radiology Although ultrasound scanning is the best technique for detecting gallstones and dilated bile ducts, radiological contrast studies are of great value in further defining the morphology of the biliary tree. Oral cholecystography This investigation is also becoming obsolete. Oral cholecystography may be used as a screening procedure in patients with suspected gallbladder disease, but ultrasound is usually preferred. Like all procedures which rely on the excretion of contrast material by the liver, poor results are obtained in the presence of liver disease, particularly cholestasis. Percutaneous transhepatic cholangiography In percutaneous transhepatic cholangiography (PTC) a very fine flexible needle is introduced into the liver substance until a bile duct is punctured. Contrast medium is then injected to fill the biliary system. More widespread use of ERCP has meant that PTC is now used selectively.

1

842

Fig. 16.1

FIG. 16.10 MR venography outlining patient hepatic veins and portomesenteric venous system without the need for catheterization

Endoscopic retrograde cholangiopancreatography (ERCP) Endoscopic retrograde cholangiopancreatography involves cannulation of the ampulla of Vater via an endoscope, followed by the injection of contrast medium into the bile ducts. It allows excellent pictures of the biliary tree to be obtained (Fig. 16.11) and is the procedure of choice in most patients in whom biliary disease is suspected. Prophylactic broad-spectrum antibiotics are used immediately before the procedure in those with biliary obstruction to prevent ascending cholangitis.

The diagnostic information gained at ERCP may be increased by the extraction of bile for culture, brushing of the biliary epithelium for cytological analysis, and biopsy of suspicious lesions. Therapeutic interventions include the extraction of gallstones and dilatation and/or stenting of benign or malignant strictures. 1

16

Vascular studies Hepatic angiography following selective catheterization of the hepatic artery defines the vascular anatomy of the liver. It is used in the investigation of focal lesions and in conjunction with lipiodol injection, and CT is the most sensitive technique for the detection of small hepatocellular carcinomas. The patency of the portal vein is defined during the venous phase. Angiography is almost mandatory prior to major liver surgery. Therapeutic applications of angiography include embolization of bleeding points, e.g. following trauma or liver biopsy, and transcutaneous arterial chemoembolization (TACE) of hepatocellular carcinoma. Hepatic vein catheterization is used to indirectly measure portal vein pressures in the diagnosis and management of portal hypertension. Occasionally splenic puncture is required to determine the portal pressure accurately. Thrombosis of the hepatic veins in the BuddChiari syndrome is most reliably defined by contrast venography.

New techniques Endoscopic ultrasound is likely to have a significant impact on the investigation of biliary and pancreatic disease. Laparoscopic cholangiography is being assessed as an alternative to ERCP.

Liver biopsy Percutaneous needle biopsy of the liver has been increasingly accepted as part of the routine investigation of patients with liver disease. In experienced centres it is an extremely safe procedure, the overall mortality in one of the largest published series (79381 liver biopsies) being 0.015%. Nevertheless, it is an invasive procedure which needs to be approached with care. The guidelines for the use of liver biopsy issued by the British Society of Gastroenterology are shown in Table 16.3.

FIG. 16.11 Endoscopic retrograde cholangiopancreatography A ERCP demonstrating a normal duct system in a patient who has had a previous cholecystectomy (arrow). B ERCP in a patient immediately following cholecystectomy (note T-tube). There is a stone at the lower end of the common bile duct, which is dilated.

Indications and contraindications These are listed in Table 16.4. The patient must be able to understand the instructions, particularly to hold his or her breath at the appropriate moment. Determination of the patient's blood group is a sensible precaution, with some serum being saved for cross-matching should transfusion be required. The contraindications (Table 16.4) are not absolute, but in the presence of any of these risk factors the chance of complications is considerably increased.

843

TABLE 16.3 Guidelines for the use of liver biopsy • Establish a clear indication for the biopsy (diagnosis of primary liver disease, diagnosis of multisystem disease affecting liver, e.g. sarcoidosis, pyrexia of unknown origin, malignant disease, metabolic disease; monitoring of response to therapy) • Review ultrasound scan of the liver performed within 4 weeks of biopsy to detect abnormal anatomy • Review prothrombin time and platelet count performed within 7 days of biopsy prothrombin time - proceed if <4s prolonged; give fresh frozen plasma if 4-6s prolonged platelet count9 - proceed if >60 x 109/L; give platelet infusions if 40-60 x 10 /L do not proceed if either parameter exceeds these limits • Screen for other contraindications (most important are ascites and biliary obstruction; unusual causes include amyloid, hydatic cyst and large haemangioma) • Obtain informed consent • Prophylactic antibiotics should be given to those with valvular heart disease and those at risk of bacteraemia • No more than two passes should be made • Keep the patient under observation for at least 6 hours after procedure

Technique Sedation is rarely necessary, except in young children, when general anaesthesia is often used. The lateral intercostal route is usually preferred, although an anterior approach can be useful in specific lesions in the left lobe, particularly when combined with ultrasound guidance. Two different types of technique are used: one in which the biopsy is aspirated into the lumen of a large-bore needle (the Menghini needle), and one where a core of liver tissue is cut out by a needle sheath sliding over a notched introducer (the TruCut needle). The latter is often preferred because it preserves the integrity of the specimen in hard fibrotic or cirrhotic livers, although technically the Menghini is easier to use. After the biopsy the patient is advised to lie on the right side for 2 hours, and during this time quarter-hourly pulse and blood pressure measurements are recorded. Complications Bleeding at the site of liver puncture is inevitable and usually causes mild pain, either locally or at the right shoulder tip because of diaphragmatic irritation. Persistent bleeding into the peritoneum or biliary tree (haemobilid) is much rarer and more serious, and angiography and embolization or even laparotomy may be required to secure haemostasis. Sudden transient hypotension immedi-

1

844

MCQ 16.2

TABLE 16.4 Indications and contraindications for liver biopsy Indications For diagnosis of Primary liver disease Multisystem disease commonly affecting the liver, e.g. sarcoidosis, lymphoma Pyrexia of unknown origin and disseminated infection, e.g. tuberculosis Malignant disease Metabolic disease, e.g. glycogen storage disease For assessment of Progression of disease and response to treatment Contraindications Clotting defect Prothrombin time more than 6s prolonged Platelet count less than 40 x 109/L Ascites Hydatid cyst Risk of dissemination Haemangioma or amyloid Increased risk of bleeding Biliary obstruction Risk of biliary leak, cholangitis and peritonitis

SUMMARY 3 Complications of liver biopsy • • • • • • • •

Bleeding Sudden transient hypotension Perforation of colon/puncture of kidney or pancreas Pneumothorax Perforation of gallbladder Biliary peritonitis Cholangitis Septicaemia

ately after the biopsy, probably due to vagal stimulation, is not uncommon. Occasionally it may be dramatic and lead to loss of consciousness, but rapid recovery can be confidently anticipated. Perforation of the colon, or puncture of the kidney or pancreas, is rare but can usually be managed conservatively. Pneumothorax may occur in patients with obstructive airways disease and hyperinflated lungs, and in those who have been unable to cooperate at the time of the biopsy. Biliary peritonitis may follow puncture of the gallbladder or biliary tree or may be due to bile leaking from an appropriate puncture site in the liver. Biliary peritonitis can usually be avoided if biopsies are not performed in patients with biliary obstruction. It is a potentially very serious complication with a high mortality; early laparotomy is indicated. The introduction of a needle into an obstructed biliary tree also carries a high risk of cholangitis and septicaemia, another reason for avoiding biopsy in these patients.

FURTHER READING ON LIVER BIOPSY Grant A, Neuberger J 1999 Guidelines for the use of liver biopsy in clinical practice. Gut 4(Suppl 4): 1-11

THE JAUNDICED PATIENT

Differential diagnosis Four broad categories of disease cause jaundice: • • • •

Prehepatic (unconjugated hyperbilirubinaemia) Hepatitis Intra- or extrahepatic obstructive jaundice Chronic liver disease.

The initial aim should be to use the clinical history, physical examination, liver function tests and special investigations to establish a working diagnosis in terms of one of these broad groupings. This will then determine the next steps to be taken in establishing the final diagnosis and treatment. The organization of this chapter reflects this clinical approach.

Clinical features Although the jaundice itself is often the dominant feature of the illness for the patient, it is important to explore the first symptoms in detail as these often contain the most reliable clues to the diagnosis (Table 16.5). A prodromal 'flu-like' illness with anorexia and myalgia is characteristic of acute viral hepatitis, and usually starts from a few days to 2 weeks before the onset of jaundice. A longer history of mild malaise and anorexia (often 3-6 months in duration), associated with weight loss of more than 3kg, is common with carcinoma of the pancreas. Biliary colic with pain referred to the right scapula and shoulder a day or two before the jaundice suggests the passage of a stone down the common bile duct. Generalized itching is another symptom of considerable importance. It often precedes or parallels the jaundice in patients with biliary obstruction, but is usually absent or delayed in acute hepatitis, only becoming of major clinical

importance in those with an unusually prolonged cholestatic phase during recovery. Important exceptions to this general rule are acute alcoholic hepatitis, where cholestasis is sometimes a dominant early feature, and hepatitis associated with some drugs and chemicals, in which the dominant feature is a disturbance of biliary secretion rather than hepatocellular damage. Pale stools and dark urine are to be expected in jaundice due to obstruction, and in acute and chronic hepatitis. This finding is therefore of no help in differential diagnosis. In contrast, normal-coloured urine and stools in the presence of obvious jaundice strongly suggests a prehepatic cause. In addition to these general features specific risk factors should be identified, including homosexuality, drug addiction, foreign travel, contact with jaundiced patients and alcohol intake. It is essential to take a complete drug history. Drug-induced hepatitis can mimic a viral hepatitis, including a prodromal illness, or can present with all the features of a pure obstructive jaundice. 1

16

Physical examination In contrast to the history, which may point strongly to the diagnosis, the physical examination usually gives less specific information. Tender hepatomegaly is often present in both acute hepatitis and obstructive jaundice, and minor splenomegaly is present in a small proportion of patients with viral hepatitis. The finding of a palpable gallbladder is well recognized as a classic sign of carcinoma of the head of the pancreas, but is not often present. If the jaundice is due to hepatic metastases, hard, irregular and sometimes tender hepatomegaly is usually a prominent feature. Cutaneous stigmata of chronic liver disease, such as spider naevi, liver palms, leukonychia and Dupuytren's contracture, should be looked for carefully, as these are important clues that an apparently acute episode may instead be a manifestation of chronic liver disease. Conversely, in acute alcoholic hepatitis there may be signs suggestive of chronic liver disease, including splenomegaly, ascites and florid cutaneous stigmata, but no histopathological features of cirrhosis.

Liver function tests

TABLE 16.5 First symptoms before jaundice Symptom

Meaning

A few days of a 'flu-like' illness Months of malaise and anorexia Biliary colic Rigors Itching before jaundice Any medication

Acute viral hepatitis Carcinoma of pancreas Stone in common duct Biliary obstruction with sepsis Biliary obstruction or cholestasis Think of this as the cause

Although it is generally true that the ratio between the serum AST (or ALT) and alkaline phosphatase helps to distinguish between hepatitic (ratio high) and cholestatic (ratio low) types of jaundice, there is a large 'grey area' and the results of these tests are best regarded as an extension to evidence from the history and examination, rather than being diagnostic in their own right. A prolonged prothrombin time is an exception, the response to parenteral vitamin K clearly distinguishing between a hepatitic and a cholestatic cause. In isolated hyperbilirubinaemia the determination of conjugated and unconjugated serum

845

CASE STUDY 16.1 A 56-YEAR-OLD MAN WITH PAINLESS JAUNDICE A 56-year-old man presented with a 3-week history of jaundice which was preceded by the development of pale stools, dark urine and pruritus. On direct questioning he denied pain or fever. His weight had decreased by 4kg to 78kg. He was otherwise well. He had no significant past medical history apart from osteoarthritis of the hips, and he used diclofenac intermittently as analgesia. His alcohol consumption was 12-20 units per week and he had smoked 20-30 cigarettes daily for over 30 years. Physical examination confirmed the presence of jaundice but revealed no cutaneous stigmata of chronic liver disease. There was some skin

excoriation secondary to scratching. The liver could not be palpated but the gallbladder was palpable. There was no splenomegaly or ascites. The remainder of the examination was normal. The laboratory investigations were: haemoglobin 13.6g/dL, white cell count 7.8 x 109/L, platelets 182 x 109/L, prothrombin time 26 s (control 13s), serum bilirubin 474umol/L, aspartate aminotransferase 54IU/L, alkaline phosphatase 1293IU/L and serum albumin 39g/L.

Queition 1. What is the diagnosis?

bilirubin levels is of considerable value in confirming a diagnosis of prehepatic jaundice.

Special investigations The most useful special test in the jaundiced patient is an ultrasound scan. The aim is to determine whether or not the intrahepatic biliary system is dilated, which is one of the most reliable indicators of extrahepatic obstruction. If duct dilatation is detected it is usual to proceed to ERCP (p. 842). If the ducts are of normal calibre and the history is consistent with an acute hepatitis, then the specific diagnosis should be established with the appropriate virological tests. Liver biopsy is only indicated if there is real doubt as to the diagnosis, or if chronic liver disease is suspected.

UNCONJUGATED HYPERBILIRUBINAEMIA The two principal diagnostic categories are haemolysis and congenital defects in bilirubin uptake or conjugation.

HAEMOLYTIC JAUNDICE

846

The diagnostic features of haemolytic jaundice are acholuric jaundice (jaundice without bilirubinuria), unconjugated hyperbilirubinaemia and a raised reticulocyte count. The differential diagnosis, further investigation and treatment are discussed on page 1214.

This is one of the few situations where a clinical diagnosis of the cause of jaundice can be made with confidence. The symptoms and the biochemical profile are typical of obstructive jaundice. In that context, the absence of pain and the palpable gallbladder make carcinoma of the head of the pancreas the most likely diagnosis. Gallstones would be expected to cause severe pain, although this may not be true in elderly patients. An ultrasound of the abdomen confirmed dilation of the gallbladder with no evidence of gallstones. A CT scan located a tumour in the head of the pancreas.

DEFECTS IN BILIRUBIN UPTAKE OR CONJUGATION The commonest defect in bilirubin uptake or conjugation (see Fig. 16.7) is Gilbert's syndrome, probably owing to defective uptake of bilirubin into hepatocytes and secondary deficiency of hepatic glucuronyl transferase. The condition is of no functional significance but is frequently the cause of much diagnostic confusion, with 'recurrent hepatitis' being the most common misleading label. The tendency for the bilirubin to rise or jaundice to become clinically apparent in association with intercurrent infections or starvation can erroneously consolidate the association between symptoms and Gilbert's syndrome. The finding of unconjugated hyperbilirubinaemia and a normal reticulocyte count, with no other abnormalities on the standard tests of liver function, is sufficient to make the diagnosis with confidence. A full explanation to the patient is all that is required for successful management. More serious congenital defects in bilirubin conjugation, such as the Crigler-Najjar syndrome (Fig. 16.7), can produce higher unconjugated bilirubin levels in the blood and kernicterus in infants, but are rare.

ACUTE HEPATITIS The term 'viral hepatitis' is often used rather loosely to mean infection with one of the hepatotropic viruses, although jaundice can occur as part of a systemic infection with several other viruses (Table 16.6) or as a reaction to drugs or alcohol. The general features of hepatitis

16

TABLE 16.6 Viral causes of hepatitis Virus Hepatotropic viruses Hepatitis A vims (HAV) Hepatitis E virus (HEV) Hepatitis B virus (HBV) Hepatitis Delta virus (HDV) Hepatitis C virus (HCV) Other viruses which can cause hepatitis Epstein-Barr virus (EBV) (infectious mononucleosis)

Cytomegalovirus (CMV)

Herpes simplex and H. zoster

Lassa fever, Marburg virus and Ebola virus

Features

RNA virus; faecal-oral transmission; 1-month incubation RNA virus; faecal-oral transmission; 1-2 month incubation DNA virus; blood and sexual transmission; 3-month incubation RNA virus; blood (and sexual?) transmission; needs HBV RNA virus; blood (and sexual) transmission; 2-month incubation

DNA herpes-type virus, close contact and aerosol transmission; liver function tests abnormal in most cases but jaundice in only 15% DNA herpes-type virus; blood and close contact transmission, behaves like EBV infection Rarely involves the liver, but may cause severe liver damage in immunosuppressed patients Rare diseases; imported from Africa; severe organ damage, including liver

described earlier are similar whichever virus is responsible, but there are major differences in epidemiology and prognosis between the various hepatotropic viruses. Serological identification of the responsible agent is therefore of considerable help in management. Treatment is generally supportive, and in controlled clinical trials neither bed rest nor low-fat diets, which are often advised, have been shown to be of any value in speeding recovery. Similarly, although abstinence from alcohol is usually advised for about 6 months after recovery from the acute illness, there is no hard evidence that this makes any difference to the outcome. There is anecdotal evidence that vigorous exercise in the prodromal phase may increase the severity of the illness.

ENTERICALLY TRANSMITTED HEPATITIS VIRUSES Hepatitis A virus Virology Hepatitis A virus (HAV) is an RNA virus whose major polypeptides are similar to those of the enterovirus group (Fig. 16.12). Serologically the diagnosis of acute HAV infection is usually made by testing for the presence of specific IgM anti-HAV antibodies, which are present in

FIG. 16.12 The structure of the hepatitis A virus

serum for about 80 days after the acute illness. Although the mechanisms leading to cell necrosis have not been defined in detail, the role of immunological processes appears more important than direct viral cytotoxicity. Epidemiology HAV is transmitted by the faecal/oral route and rarely, if ever, parenterally. It may be endemic or sporadic, and epidemics sometimes occur. Childhood acquisition, often asymptomatic, is the norm in parts of the Mediterranean, the Middle East and India. In contrast, in northern Europe the vast majority of children reaching adulthood have no evidence of immunity to hepatitis A. They are then at risk of acquiring hepatitis A while travelling in endemic areas or with changes in eating habits, e.g. developing a taste for uncooked shellfish. The incubation period is about 1 month (2-7 weeks). Carriers have not been detected. Clinical features Acute hepatitis A is usually a mild illness preceded by a typical prodrome. This may be so similar to 'flu', with prominent myalgia, that the patient does not connect the prodrome with the jaundice, and should be asked about a preceding febrile illness. The patient is most infectious during the prodrome and less so during the first week of the clinical illness. Acute liver failure is a rare complication, and progression to chronic liver disease does not occur. Hepatitis A may follow a biphasic clinical course or, more frequently, may be followed by a protracted cholestatic phase. Aplastic anaemia, particularly in children, is a rare complication with a high mortality.

Hepatitis E virus This is an RNA virus which resembles the calciviruses in the size (7.5kb) and organization of its genome. It is spread by the faecal/oral route and first came to attention because of a major waterborne hepatitis epidemic in India, initially thought to be due to hepatitis A. It is common in the Indian subcontinent. The incubation period is about 6 weeks and the clinical illness it produces is in most respects like any viral hepatitis. However, there is an unusually high mortality among pregnant women who become infected. It does not produce chronic hepatitis. Diagnostic tests for the

847

virus and antibodies to its antigenic components are now available in specialist laboratories.

PARENTERALLY TRANSMITTED HEPATITIS VIRUSES Hepatitis B virus

1

Virology Hepatitis B vims (HBV) is a member of an unusual group called the HepaDNA viruses. It has an outer and an inner

protein coat, with double-stranded DNA and a DNA polymerase in the nucleocapsid (Fig. 16.13). The main antigenic protein of the outer coat, the hepatitis B surface antigen (HBsAg), is produced in excess during viral replication, and forms large numbers of smaller empty particles in the serum. The complete core protein coat reacts antigenically as hepatitis B core antigen (HBcAg). The open reading frame for the core gene of the virus has two start codons (positions 1814 and 1901 in Fig. 16.14). Transcription from the second one leads to production of the core subunits, which self-assemble to make the viral capsid, whereas transcription from the first start codon inserts a signal peptide

CASE STUDY 16.2 A YOUNG MAN WITH JAUNDICE RECENTLY RETURNED FROM INDIA A 23-year-old university student presented with jaundice, which he had noticed the previous day. He had been unwell for the previous 6 days, with nausea, vomiting and dark urine. However, these symptoms had improved once the jaundice appeared. He had recently travelled to India. He drank 20 units of alcohol per week and had used 'Ecstasy' most weekends for the previous 2 years. He was a sexually active heterosexual and had had numerous partners since the age of 17. There was no history of previous illnesses requiring hospitalization. Physical examination confirmed the presence of jaundice. His temperature was 38.2°. There were no cutaneous stigmata of chronic liver disease. The liver was palpable 3 cm below the costal margin and was tender. There was no clinical evidence of splenomegaly or ascites. His mental state was normal, with no evidence of drowsiness or confusion. The initial laboratory investigations were: haemoglobin 14.5 g/dL, white cell count 3.6 x 109/L, platelets 223 x 109/L, prothrombin time 16s (control 13s), serum bilirubin 176fimol/L, aspartate aminotransferase 1462IU/L, alkaline phosphatase 121 IU/L.

1

848

MCQ16.3

Question 1. What Is the most likely jdiagnfosis? The very high serum transaminase level indicates hepatitis, and in this age group this is most likely to be viral or drug related. He has a number of risk factors for hepatitis. Recent travel to India puts him at risk of hepatitis A and hepatitis E. Hepatitis B would also have to be considered on the basis of multiple sexual partners. Hepatitis C is unlikely as his recreational drug usage was confined to 'Ecstasy', and it is unusual for hepatitis C to cause clinical episodes of jaundice. 'Ecstasy' can cause jaundice but this is usually seen within the first few months of use. There is no reason to link alcohol to this episode of jaundice. The viral serology indicated that he was HAV IgM positive, HBsAg negative and HEV IgG negative. This established the diagnosis as acute hepatitis A. This was also the most likely diagnosis because of the fever that was present. He was due to attend the surgery 4 days after the initial consultation to review the results of investigations, but he did not arrive. Later that day he was

found unconscious by a flatmate and was taken directly to hospital. Physical examination confirmed he was unresponsive except to painful stimuli. The laboratory investigations were: haemoglobin 14.1 g/dL, white cell count 3.2 x 109/L, platelets 116 x 109/L, prothrombin time 101 s (control 13s), serum bilirubin 354 |imol/L, aspartate aminotransferase 434 IU/L, alkaline phosphatase 109 IU/L. Question 2. What complication has developed? The high prothrombin time and the impaired conscious level indicate that he has developed acute liver failure. This occurs in only about 0.1% of cases of acute hepatitis A infection. The prothrombin time was over 100 seconds and this indicated that he was unlikely to survive. Consequently, he was transferred to a specialist unit and underwent liver transplantation the following day. He made a full recovery. Question 3. What else should be done? His household contacts should be vaccinated against hepatitis A.

16

FIG. 16.13 The hepatitis B virus. A Structure; B genome

in front of the core protein, which leads to self-cleavage in the endoplasmic reticulum and the secretion of a soluble protein into the blood known as the hepatitis B 'e' antigen (HBeAg). This seems to be of crucial importance in establishing immune tolerance to the core protein, particularly in infants. Its presence in blood, with HBsAg, is also used as an indication of active viral replication in the liver and as a marker of infectivity. Each of these antigens can elicit a corresponding antibody response (anti-HBs, anti-HBe and anti-HBc). In an acute HBV infection the appearance of these antigens and antibodies in serum, and changes in serum AST levels, follow a well-defined pattern (Fig. 16.15). In most cases, HBV infection is detected by finding HBsAg in serum. The diagnosis of acute, as opposed to chronic, HBV infection is best made by testing serum for IgM-class anti-HBc antibodies. Epidemiology The main route for HBV infection is by parenteral inoculation, although direct transmission by blood transfusion is now very uncommon following the development of sensi-

FIG. 16.15 The main serological changes in (A) acute HBV and (B) chronic HBV infection The break in the x-axis of B has been used to indicate the very variable length of time that can elapse between the establishment of the chronic carrier state and the spontaneous cessation of viral replication.

tive serological tests to detect infectious blood donations. The virus has a wide global distribution (Fig. 16.16). Chronic carriers are especially prevalent in the Middle and Far East and in sub-Saharan Africa, where they may constitute 10-30% of the population. In the Far East vertical transmission from mothers to their children is the usual route of infection, whereas in Africa horizontal spread in early childhood, possibly through scarification rituals, is more common. In northern Europe and the USA, where chronic carriers are less common and where the screening of blood donations is accepted practice, the main reservoirs

849

FIG. 16.16 The distribution of HBV infection in the world as judged by the prevalence of hepatitis B surface antigen in different populations

for the virus are urban homosexual communities and intravenous drug abusers. The virus seems to be readily transmitted by rectal intercourse.

850

Immunopathology Recent evidence suggests that the complete hepatitis B virion has an area of its surface protein coat that resembles the IgA molecule and that this can bind to IgA receptors on hepatocytes. This probably explains the hepatotropism of HBV. During replication of the virus, viral antigens, principally HBsAg and HBcAg, appear on the surface membrane of infected hepatocytes and trigger a cytotoxic T-lymphocyte response (Fig. 16.17). This is largely directed at HBcAg and leads to cytolysis, seen clinically as acute liver damage (jaundice and AST elevation), and histologically as spotty necrosis of hepatocytes scattered throughout the liver lobules. Released virions are then neutralized by antibodies probably directed at the region responsible for hepatocyte binding. Complete recovery involves a coordinated T-cell and antibody response to different viral antigens. Failure of these complex mechanisms results in about 5% of apparently normal adults not clearing the virus, and becoming chronic carriers (Fig. 16.15). The carrier state is commoner in infancy and childhood and in those with an associated immunodeficiency, whether this is congenital, due to drugs (as in immunosuppressed transplant recipients), or secondary to other diseases. The human immunodeficiency virus (HIV) is the most dramatic of these associated virus infections, and AIDS and HBV infection often coexist in male homosexuals. In the case of HBV, an effective vaccine made by purifying HBsAg from the plasma of carriers and a recombinant HBsAg vaccine are available for those at increased risk. These include sexual and family contacts of carriers, as well as healthcare workers and children born to carrier mothers. In many of the developing countries in the Far East, extensive country-wide vaccination programmes have been in force for over a decade, and the incidence of HBV carriers has shown a marked fall. In Africa vaccination pro-

FIG. 16.17 The immunopathology of hepatitis B Liver cell damage in HBV infection is due to an immune attack by T cells against viral antigens expressed on the surface of infected hepatocytes. The T cells recognize the foreign viral antigens as small peptides bound to class I histocompatibility antigens (HLA class I).

grammes are less well developed, but HB vaccine is now being included in the childhood vaccination programme. Clinical features The prodromal illness may be particularly severe, with arthralgia, urticaria and occasionally a full 'serum sickness' syndrome with associated glomerulonephritis. Acute liver failure is a rare complication, and the main clinical concern is the relatively high rate of progression to a chronic carrier state. The resulting liver lesions are described on page 864. Careful follow-up with repeated serological tests is an essential part of the management. Although persistence of HBsAg for more than 6 months is the usual criterion for chronicity, persistence of HBeAg in serum for more than 8 weeks may give an earlier indication of failure to clear the virus normally.

Delta virus Virology This is a small RNA virus with a genome that does not code for its own protein coat; instead, it borrows HBsAg provided by a concurrent HBV infection (Fig. 16.18). Serologically, Delta infection can be detected by testing for anti-Delta antibodies in serum. These appear early in the course of the infection but may not persist. Epidemiology Delta infection is absolutely dependent on a helper function of HBV, and so is only found in two settings: either there is a dual acute infection of HBV and Delta virus, or Delta virus infects an existing HBsAg carrier. Like HBV it is spread by the parenteral route or via intimate contact. It is endemic in countries around the Mediterranean basin, and in northern Europe and the USA is found mainly in drug abusers.

16

SUMMARY 4 Clinical aspects of viral hepatitis Type Faecal/oral spread A E (Enteric non-A, non-B) Parenteral/sexual transmission C (Parenteral non-A, non-B) B

D (Delta)

Incubation period

Clinical features

Outcome

Vaccines

4 weeks 6 weeks

Mild often subclinical High mortality in pregnancy

No chronic disease No chronic disease

Widely available None

8 weeks

Mild otten subclinical

None

12 weeks

Variable severity

4-12 weeks

Severe in HBV carriers

20-40% progress to chronic disease Neonatal 90% carriers Childhood 10-40% carriers Adult 1-10% carriers Carrier on chronic HBV; Recovery in most coinfections

36 nm

FIG. 16.18 The Delta virus The Delta virus is a strange structure. It borrows its coat from the hepatitis B virus (HBsAg), so HBV must also be present in hepatocytes. The infective material of Delta is a small circular strand of RNA.

Immunopothology Clinical evidence and studies in primates suggest that the Delta virus is directly cytopathic to the cells it infects. The contribution of immune responses to cell damage and recovery is not known. Clinical features Coinfection produces an acute hepatitis that is clinically indistinguishable from HBV infection alone, although there is some suggestion that the illness may be more severe. Recovery from the HBV infection is always associated with the disappearance of Delta. Infection in an HBsAg carrier can produce a severe acute hepatitis, and in some cases acute liver failure. Chronic Delta infection is also more likely to occur and has a bad prognosis (p. 865).

Hepatitis C virus Hepatitis C, an RNA virus, was discovered in 1989. It belongs to the Flavivirus family and is quite common

Widely available and in national programmes None

throughout the world. There are significant geographical variations in the distribution of the four main genotypes of the virus which influence both the virulence of the infection and its responsiveness to current antiviral therapy. The results of epidemiological studies suggest that the carrier rate varies from about 0.05 to 0.2% in northern Europe, through 1 or 2% in the Mediterranean to 5% or more in the Far East. It is an important cause of chronic liver disease, cirrhosis and hepatocellular carcinoma. It is transmitted mainly by parenteral inoculation. Blood transfusion was one of the most important sources until the introduction of reliable blood tests for the detection of carriers around 1990-91. Drug abusers remain an important at-risk group. Sexual and neonatal transmissions can occur but are uncommon. Acute infection is not usually associated with a recognized clinical episode of hepatitis, even in adults. About 85% of cases become chronic carriers and 20-40% of these will develop chronic liver disease over a period of 20-30 years. Standard liver function tests screen for necroinflammatory activity but not for fibrosis. The assessment of the severity of the infection is best determined by measuring viral load in blood and by liver biopsy.

NON-HEPATOTROPIC VIRUSES Epstein-Barr virus Liver function tests are often abnormal in glandular fever, but jaundice occurs in less than 10% of cases. The Epstein-Barr virus (p. 287) does not seem to infect hepatocytes and liver cell necrosis is minimal. Recovery is almost always uneventful, although fatigue, as part of a postviral syndrome, can be troublesome for a few months.

Cytomegalovirus In adults cytomegalovirus (p. 288) can produce an illness like glandular fever, with accompanying jaundice in some cases, but recovery is usually rapid. Liver biopsy, although

851

not required to make the diagnosis (which is made serologically), may show the characteristic eosinophilic nuclear inclusions in infected hepatocytes.

Herpes simplex Herpes simplex (p. 286) is a rare cause of severe acute hepatitis, that usually affects immunosuppressed patients. There is usually a local oral lesion, but the striking feature is disseminated infection with encephalitis or myocarditis, in addition to hepatitis.

DRUG-INDUCED JAUNDICE The continued production of new drugs has been accompanied by an increasing problem with drug-induced liver damage (Table 16.7). It is generally stated that about 10% of cases of jaundice in hospital practice are due to medication. There are two main types of adverse reaction: predictable or dose-dependent (type A), and idiosyncratic or dose-independent (type B). In the former, liver damage is dose-dependent and relatively common, and can usually be reproduced in laboratory animals. Idiosyncratic cases are infrequent, not related to dose and not usually reproducible experimentally. Recognition of these cases is important for two main reasons: withdrawal of the drug usually leads to rapid recovery, and repeat prescription, which may be associated with a more severe reaction, can be avoided. The liver damage produced by drugs can mimic almost any acute or chronic hepatic or biliary disorder. In most cases liver damage is predominantly hepatitic or cholestatic.

TABLE 16.7 Examples of drugs causing liver damage Drug Antituberculous Para-aminosalicylic acid (PAS) Isoniazid (INAH) Rifampicin Pyrazinamide and ethionamide Other antibiotics Tetracycline Sulphonamides including co-trimoxazole Erythromycin Ketoconazole Nitrofurantoin Psychotropic drugs Monoamine oxidase inhibitors Phenothiazines Tricyclic antidepressants Analgesics Paracetamol Aspirin Non-steroidal analgesics Steroid drugs and hormones 17-alkyl/ethynyl substituted steroids (including contraceptive pill)

DRUG-INDUCED HEPATITIS Paracetamol- and halothane-induced liver damage are two examples of drug-induced hepatitis; both are potentially fatal. Paracetamol is a predictable hepatotoxin, but the liver damage induced by halothane occurs idiosyncratically.

Paracetamol

852

Paracetamol is safe in normal therapeutic doses but produces a severe hepatitis if taken in overdose (more than 10g or 20 tablets). Liver damage is due to the formation of highly reactive metabolites (Fig. 16.19). At normal dose levels these are inactivated by conjugation with glutathione. When a large amount of drug is taken, insufficient glutathione is available for detoxification and the metabolites damage the cell by binding to cellular macromolecules. Compounds acting as -SH donors can prevent this process. The most effective and non-toxic of these is 7V-acetylcysteine given intravenously (p. 28).

Anaesthetic agents Halothane Immunosuppressive drugs Azathioprine Methotrexate Corticosteroids Cardiovascular drugs Methyldopa Perhexilene Thiazide diuretics Oral hypoglycaemic drugs Chlorpropamide

Comments Idiosyncratic type; 1% get drug hepatitis in first 2 months Raised transaminases in 10%; 1% jaundice in first 2 months Up to 4% develop hepatitis; ?increases isoniazid toxicity Late-onset hepatitis in 1%

Microvesicular fatty liver Idiosyncratic; mimics viral hepatitis Cholestasis; especially estolate Hepatitis in 0.1% Cholestasis and chronic hepatitis Viral hepatitis lookalike in 1-2% after about 4 months Mixed cholestatic and hepatitic picture in 1-2% after 2-4 months; can persist Rare cholestatic hepatitis Predictable liver necrosis in overdose (more than 10g); /V-acetylcysteine antidote Acute anicteric hepatitis, especially in children; may cause Reye's syndrome Considered one of common causes, given scale of usage Predictable (dose/duration-dependent) Cholestasis, especially the 19-nortestosterones; increased risk of adenoma Commonly produces minor increase in transaminases, very rarely fatal hepatitis Rare Cholestasis Long continued treatment, as in psoriasis, may cause fibrosis and cirrhosis Fatty infiltration Acute and chronic hepatitis starting after 2-5 months Rare hepatitic illness, looks like alcoholic liver damage Cholestatic hepatitis; may be prolonged Cholestatic hepatitis in 5%, with granulomatous reaction

damage by direct hepatotoxicity. The agents include androgenic and anabolic steroids, progestogens and oestrogens. The cholestatic effect is dose related and the obstructive jaundice resolves following the cessation of treatment. Histologically there is evidence of centrilobular cholestasis without obvious hepatocellular damage. The exact mechanism is unknown.

16

Chlorpromazine Chlorpromazine is a rather dramatic example of an agent causing an idiosyncratic reaction. About 1-2% of those taking this drug develop a severe cholestasis, which may persist for weeks after the drug has been withdrawn. The reaction has been described following only one dose. The mechanism responsible is unknown.

FIG. 16.19 Paracetamol overdose Paracetamol is normally sale to take because the metabolites are inactivated by conjugation. After an overdose the conjugating material, particularly glutathione, is used up and reactive metabolites damage the liver cell proteins.

Halothane Halothane is an example of an unpredictable (idiosyncratic) hepatotoxin. Significant liver damage is very rare (1 in 30000 halothane anaesthetics), but when it occurs it is often fatal. It is more common after repeated halothane anaesthesia. The clinical picture is very similar to viral hepatitis, with a prodromal illness starting a few days after the surgical procedure. In about one-third of cases unexplained pyrexia has been noted following a previous halothane anaesthetic. Progression to acute liver failure is common, although the exact frequency of this complication is difficult to assess because of uncertainty about the diagnosis in mild cases. The mechanism of liver damage is thought to be immunological. A reactive metabolite, produced during oxidative metabolism of halothane, alters the normal antigens on the liver cell membrane. In patients with halothane hepatitis these new antigens have been shown to stimulate an immune response, which may produce the liver damage. The idiosyncrasy seems to be related to the ability of the immune system to recognize these subtle antigenic alterations.

DRUG-INDUCED CHOLESTASIS The cholestatic jaundice induced by C17-substituted testosterone derivatives and chlorpromazine is an example of predictable and idiosyncratic hepatotoxicity.

C17-substituted testosterone derivatives C17-substituted testosterone derivatives produce liver

ACUTE LIVER FAILURE

Pathophysiology Severe liver damage leads to secondary disturbance of function in other organ systems. This complex development is recognized clinically as acute liver failure. Encephalopathy is the commonest and most obvious complication. Although the exact cause is unknown, potential factors include the release of toxins from the damaged liver, failure to detoxify metabolites, and alterations in cerebral neurotransmitters owing to amino acid imbalances (p. 873). Renal failure and cardiovascular instability are also common and carry a particularly bad prognosis.

Epidemiology Acute liver failure is a rare complication of almost any of the causes of acute hepatitis. In most countries viral hepatitis (A, E, B and Delta) is the most common cause, but in the UK about half the cases are due to paracetamol poisoning, with most of the remainder being associated with hepatitis A or B, or seronegative hepatitis (also known as non A-E hepatitis). Hepatitis C virus seems to be a very rare cause of acute hepatic failure. A small proportion are due to medications such as isoniazid and rifampicin. In areas around the Alps and the Rocky mountains, ingestion of the poisonous mushroom Amanita phalloides is an important cause. The recreational use of Ecstasy may cause acute liver failure. Other less common causes include Wilson's disease, autoimmune hepatitis, Budd-Chiari syndrome, hyperthermia, and a range of conditions associated with pregnancy.

Clinical features Patients develop the signs of acute liver failure within 8-12 weeks of the onset of the illness and have no previous history of liver disease. This distinguishes acute liver failure

853

TABLE 16.8 Grades of coma in liver failure Grade

Clinical features

I

Slowness of mentation and affect, fluctuant mild confusion; reversed sleep rhythm; slurred speech Accentuation of grade I; inappropriate behaviour; drowsiness Marked confusion; sleeps most of the time but rousable; incoherent speech Unrousable; may or may not respond to noxious stimuli

II III IV

from rapidly progressive chronic liver disease. The first signs of the disease often relate to the encephalopathy, with the appearance of mild confusion, irrational behaviour or even euphoria. Progressive deterioration in mental condition then follows, leading to coma. The usual grading of the level of coma is according to a four-stage model (Table 16.8). In some rapidly progressive cases jaundice may never be apparent and the whole illness, from first symptoms to death, may last less than a week. The classic signs of encephalopathy, including fetor hepaticus and liver flap, are usually absent. In the early stages of grade IV coma hyperventilation is common, and the pupils are dilated and react sluggishly to light. Hypertonia and grasp reflex are readily elicited, and in deep coma decerebrate postures are seen. Terminally, the oculovestibular reflex is lost and hypotension, cardiac arrhythmias and respiratory arrest occur. The prolongation of prothrombin time is the most important of the liver function tests and is very influential in determining prognosis. However, extreme prolongations may be seen after paracetamol ingestion in patients who will survive, whereas modest prolongations are typical of syndromes that evolve over 4-12 weeks but ultimately lead to death in the vast majority of cases. The serum bilirubin increases with time and aids the determination of the prognosis. However, the range of serum transaminase levels seen is great and, as it bears no relationship to the clinical outcome, is of limited value. An early and often extreme acidosis is very suggestive of paracetamol toxicity.

Complications and management The complications and management of acute liver failure are listed in Table 16.9. Treatment consists of skilled supportive therapy based on a sound knowledge of the expected course of the disease and its complications. Admission to an intensive care area is essential and early ventilatory support is frequently required. Intravenous nutritional support is with 10% dextrose. Serum sodium and potassium levels are measured daily and any abnor-

1

854

MCQ 16.4

2

Fig. 16.2

TABLE 16.9 Complications and management of acute liver failure Problem

Treatment

Cerebral oedema Nutrition, fluid and electrolytes Renal failure

Mannitol infusions given promptly reverse pressure rises 10% dextrose into a central vein; extra potassium needed; low sodium is usually dilutional Common problem; half have 'functional renal failure', others acute tubular necrosis. Attention to fluid balance may help, but dialysis may be needed H2-receptor antagonists significantly reduce risk, and are given routinely Common late complication. Intubate early to protect airway

Gastrointestinal haemorrhage Respiratory failure

malities corrected. Hypoglycaemia and hypokalaemia are early life-threatening abnormalities which must be detected and corrected quickly. Erosive oesophagitis and gastritis leading to severe gastrointestinal haemorrhage used to be a major problem, but prophylaxis has proved effective. Renal failure is common, and haemodialysis or other forms of renal support may be needed. Coagulation disorders are also common, owing to failure of synthesis of clotting factors and, in some cases, the occurrence of disseminated intravascular coagulation (p. 1275). Cerebral oedema is an important cause of death, and mannitol infusions are useful in reducing increased intracranial pressure. The other major causes of death are sepsis and cardiovascular instability, which are unresponsive to antimicrobial agents and vasopressor drugs, respectively. Liver transplantation is a common treatment for the most severe cases, but it is essential to have a method of rapidly assessing prognosis to avoid transplanting those who will survive even though they appear to be desperately ill. Liver assist devices are being developed as a means of supporting patients until a liver becomes available for transplantation or until spontaneous recovery occurs. These use human hepatocytes from a tumour cell line or animal liver cells which restore some degree of liver function, but they await evaluation in controlled clinical trials. Combined medical and surgical management strategies are achieving survival rates of 40-90%, depending on the underlying aetiology. Survivors rarely develop cirrhosis or chronic liver disease, probably because of the remarkable regenerative capacity of the liver. 1

FURTHER READING ON HEPATITIS Da Villa G, Picciottoc L, Elia S, Peluso F, Montanaro F, Maisto T 1995 Hepatitis B vaccination: universal vaccination of newborn babies and children at 12 years of age versus high risk groups. A comparison in the field. Vaccine 13(13):1240-1243. Heathcote E J, Shiffman M L, Cooksley W G E, Dusheiko G M, Lee S S, Dalart L, Reindollar R, Reddy R K, Wright T L, Lin A, Hoffman J, De Pamphilis J 2000 Peg interferon alpha-2a in patients with chronic hepatitis C and cirrhosis. N Engl J Med 343:1673-1680.

RECENT ADVANCES LIVER ASSIST DEVICES A new generation of extracorporeal liver assist devices has started to emerge that have the potential to do for hepatology what haemodialysis did for renal failure. The devices currently under evaluation use either pig cells or cells from a human tumour cell line - Hep 3CA cells - to provide basic functions. The volume of cells used varies from 20 g of porcine cells to 200-400 g of Hep 3CA cells. The porcine cells are usually in series in the extracorporeal circuit with a charcoal filter designed to protect them from harmful toxins. Patients are treated intermittently for about 6 hours per day with the porcine cells, or continuously with the Hep 3CA cells for up to 10 days. The clinical evaluation of these systems is in progress. Preliminary results are sufficiently encouraging to proceed to formal controlled trials. The application of an effective device might initially be to 'bridge' very ill patients to liver transplantation by stabilizing their condition. However, if the devices are powerful supplements of liver function they may be effective in salvaging patients with acute or chronic liver failure without recourse to transplantation. These devices also have theoretical roles in extending the boundaries of feasibility for liver resection and the scope for using marginal donors in liver transplantation.

Hoofnagle J H, Di Bisceglie A M 1997 The treatment of chronic viral hepatitis. N Engl J Med 336:347-356. Makin A J, Wendon J, Williams R 1995 A 7-year experience of severe acetaminophen-induced hepatotoxicity (1987-1993). Gastroenterology 109(6):1907-1916. Mellor J, Holmes E C, Jarvis L M, Ya L, Simmonds P 1995 Investigation of the pattern of hepatitis C virus sequence diversity in different geographical regions: implications for virus classification. The International HCV Collaborative Study Group. J Gen Virol 76(pt 10):2493-2507. Tillman H, Trautwein C, Walker D, Michitaka K, Kubicka S, Boker K, Manns M 1995 Clinical relevance of mutations in the precore genome of the hepatitis B virus. Gut 37(4):568-573. Zeuzem S, Feinman V, Rasenack J, Heathcote E J, Lai M-U, Gane E, O'Grady J, Reichen J, Diago M, Lin A, Hoffman J, Drunda M J 2000 Peg interferon alpha-2a in patients with chronic hepatitis C. N Engl J Med 343:1666-1672.

JAUNDICE IN PREGNANCY CHOLESTASIS OF PREGNANCY Cholestasis of pregnancy is a rare condition in which pruritus, with or without painless jaundice, occurs in the last trimester of pregnancy. Some women taking the oral contraceptive pill seem to develop a similar condition, suggesting that it represents an unusual response to high oestrogen levels. The itching usually responds to cholestyramine. The symptoms disappear following the

birth, and liver function tests return to normal. The cholestasis tends to recur with each pregnancy.

16

ACUTE FATTY LIVER OF PREGNANCY Acute fatty liver of pregnancy is a serious but rare condition. It presents in the last trimester with a prodromal illness which is similar to acute viral hepatitis, and which starts about a week before the jaundice appears. The histological changes are quite distinct from those of viral hepatitis. There is a widespread, centrilobular microvesicular fatty change in hepatocytes. The aetiology is unknown, although identical changes are rarely seen following tetracycline treatment. Urgent delivery of the baby is essential and in most cases this triggers rapid resolution of the condition. Acute liver failure can occasionally develop but with optimal treatment the prognosis is good and liver transplantation is rarely required.

HELLP SYNDROME AND TOXAEMIA The HELLP syndrome is a combination of encephalopathy, elevated liver enzymes and thrombocytopenia. It also occurs in the last trimester and can be difficult to differentiate from acute fatty liver of pregnancy. Patients with toxaemia (pre-eclampsia or eclampsia) may develop infarction of the liver secondary to obliteration of small arteries and arterioles. These patients are characterized by especially high serum transaminases and can develop acute liver failure.

ACUTE ALCOHOLIC HEPATITIS Clinical features The term 'alcoholic hepatitis' has two quite different meanings. To the histopathologist it indicates certain specific features found in some patients with alcoholic liver disease, particularly perinuclear eosinophilic hyaline inclusion bodies (Mallory's hyaline) in hepatocytes in, or near to, areas of spotty parenchymal necrosis. An associated inflammatory infiltrate consists largely of polymorphonuclear leukocytes. There may or may not be associated fibrosis or cirrhosis. Although some patients with these changes on liver biopsy are entirely asymptomatic, the pathological features are important as they indicate a greatly increased chance of the development of cirrhosis. This progression occurs in 90% of patients with features of alcoholic hepatitis on liver biopsy if they continue to drink, whereas complete regression is the rule in 90% of those who subsequently remain abstinent. 0 To the clinician, the term 'alcoholic hepatitis' is also used to describe an acute illness associated with alcohol abuse,

855

in which deep jaundice, abdominal pain, fever, marked polymorphonuclear leukocytosis, and elevated prothrombin time are the principal features. Cutaneous stigmata of chronic liver disease are common, even in the absence of cirrhosis. Cholestasis is often prominent and may lead to a mistaken diagnosis of extrahepatic obstructive jaundice, and possibly a laparotomy, which can be fatal. The mortality of the condition is high (30-60%). Both the clinical illness and the pathological changes are commoner and more severe in women, and in people of northern European origin (including America). Although the histological changes can be found in southern European patients, the severe acute clinical illness is uncommon. There is no definite relation to the pattern of drinking or the type of alcoholic drink.

Management Trials of specific drug treatments have proved disappointing, although recent studies support the use of prednisolone 30-40 mg for periods of up to 4 weeks. Abstinence remains the most effective therapy (p. 251).

CHOLESTATIC JAUNDICE As with the diagnosis of acute hepatitis, a good history is the key to identification of cholestatic (including obstructive) jaundice. Although pale stools and dark urine are common to both, the early appearance of itching and, more rarely, high fevers with rigor (ascending cholangitis), is a general pointer to extrahepatic bile duct obstruction. The commonest causes are outlined in Table 16.10 and the specialized investigations that may be required to identify the site and cause of the obstruction are described on pages 842 and 843. An essential early investigation is an ultrasound scan of the upper abdomen. The presence of dilated bile ducts is an almost certain indication of major duct obstruction. Gallstones, which are echo dense, are also detected easily, but it can be more difficult to comment with certainty on the pancreas and other retroperitoneal structures. In some cases a CT scan may be helpful. The final determination of the site of obstruction is best made by ERCP or PTC.

GALLSTONES 1 Pathophysiology Mixed cholesterol stones Mixed cholesterol stones are the commonest variety in the western world, and consist largely of cholesterol crystals

1

856

MCQ 16.5

TABLE 16.10 Causes of cholestatic jaundice Intrahepatic Metabolic/unknown cause Pregnancy Contraceptive pill Methyltestosterone Parenteral nutrition Lymphoma Benign recurrent intrahepatic, and postoperative cholestasis Extrahepatic Benign Gallstones in bile duct Post-traumatic stricture Sclerosing cholangitis Biliary atresia Choledochal cyst Acute and chronic pancreatitis Retroperitoneal fibrosis Ascending cholangitis Haemobilia

Liver damage Primary biliary cirrhosis Late viral hepatitis Alcoholic hepatitis Biliary hypoplasia Secondary deposits

Malignant Carcinoma of the gall bladder Carcinoma of the bile duct Hilar lymphadenopathy Carcinoma of the ampulla of Vater Carcinoma of the pancreas

with a variable proportion of calcium salts. The main constituents of bile are cholesterol, phospholipids and bile salts. The ratio between cholesterol and the other constituents is crucial in maintaining cholesterol in solution. Bile supersaturated with cholesterol can develop because of an increased cholesterol concentration, or a decreased bile salt content. These changes in the constituents can be due to altered rates of secretion, either congenital or acquired, or changes in the relative proportions during concentration and storage of bile in the gallbladder. Pigment stones Pigment stones consist of bilirubin and salts such as phosphates and carbonates. They are common in the Far East. Increased delivery of bilirubin in chronic haemolytic anaemias is a frequent cause in the UK, whereas in the Far East parasitic and bacterial infection of the bile is the usual causative factor. These organisms promote the formation of insoluble deconjugated bilirubin by the release of enzymes such as glucuronidase.

Epidemiology The epidemiology of gallstone disease is changing. Cholesterol gallstones occur more often in women than in men (Fig. 16.20), and the prevalence increases with age, obesity and parity. Other suggested risk factors are diet (a vegetarian diet appears protective), a positive family history, low social class, and the use of oral contraceptives. There

FIG. 16.20 The increasing prevalence of gallstones with age in both men and women

are marked racial differences in prevalence, with Native Americans being particularly susceptible. There is a predisposition to cholesterol stone formation in patients with chronic liver disease, in whom there is reduced excretion of bile salts, and also in patients with a resection or disease of the terminal ileum. In these cases there is reduced bile salt reabsorption from the gut and a consequent fall in bile salt excretion into the bile.

Detection Only 10% of cholesterol gallstones contain enough calcium salts to be visualized on plain X-ray of the abdomen. Ultrasound is the investigation of choice to detect stones in the gallbladder. Several studies have shown that, using ultrasound, stones can be detected with an accuracy of more than 95%.

Clinical features Stones in the gallbladder are a common incidental finding and, if they remain there, may never give rise to symptoms. Stones in the biliary tree may cause biliary colic. Impaction in the cystic duct followed by infection produces acute cholecystitis, whereas passage down the common bile duct produces biliary colic, obstructive jaundice and cholangitis.

ACUTE CHOLECYSTITIS The presenting symptoms of acute cholecystitis are malaise, anorexia and pyrexia, with pain in the right hypochondrium sometimes radiating to the right shoulder. There is guarding and rigidity in the right upper quadrant. Tenderness on inspiration over the gallbladder area (Murphy's sign) is usually present. Jaundice suggests stones in the common bile duct. A polymorphonuclear leukocytosis is usually present but does not help in the differential diagnosis, which includes acute appendicitis and pancreatitis. There are two different approaches to treatment: either

early cholecystectomy or initial broad-spectrum antibiotic therapy (e.g. cephalosporin with metronidazole), fluid replacement and analgesia, followed by elective cholecystectomy several weeks later. Exploration of the common bile duct, with operative cholangiography, should be undertaken if there is jaundice, dilatation of the duct, or evidence on previous investigations of stones in the common bile duct. Complications include empyema and perforation of the gallbladder, and liver abscesses following portal pyaemia. Perforation is particularly frequent in the elderly. The signs are those of generalized peritonitis, unless previous episodes of cholecystitis have led to dense adhesions, when a localized abscess may be produced. Empyema results from infection of obstructed bile in the gallbladder, which fills with pus. There is a very tender mass in the right hypochondrium, with fever and sometimes Gram-negative septicaemia. Treatment is by emergency drainage of the biliary system under broad-spectrum antibiotic cover.

16

STONES IN THE COMMON BILE DUCT Although the classic symptoms of biliary colic followed by jaundice might be expected to accompany the passage of a stone down the common bile duct, it is important to realize that a substantial proportion of patients with biliary obstruction due to gallstones may present with painless jaundice. Cholangitis and Gram-negative septicaemia are frequent and dangerous complications, and the diagnosis must be established quickly. Ultrasound is again of great value in demonstrating a dilated biliary system, and may also show the stone in the common bile duct. It is usually necessary to proceed to cholangiography (PTC or ERCP, p. 843) to confirm the diagnosis, but antibiotics should be given first to minimize the risk of cholangitis. Treatment is usually surgical, involving exploration of the duct, the removal of stones and temporary T-tube drainage. It is best to start antibiotics, to give intramuscular vitamin K, and to correct fluid balance or electrolyte abnormalities before surgery. Many broad-spectrum antibiotics have been shown to be effective in this prophylactic setting and cephalosporins, with their low incidence of toxic reactions, are usually favoured. Mannitol infusion has been shown to reduce the incidence of postoperative renal failure. In the elderly or in bad-risk patients endoscopic sphincterotomy, followed by the removal of any remaining stones with a basket or balloon via the endoscope, is an alternative procedure and in many centres this is becoming the preferred treatment for all cases.

ACUTE CHOLANGITIS In any patient with obstructive jaundice the biliary system can become infected, often with dramatic clinical conse-

857

quences. Rigors usually accompany a high fever, and this in the presence of jaundice is sometimes known as Charcot's intermittent biliary fever, or more simply ascending cholangitis. Vascular collapse and renal failure are common complications, reflecting the high incidence of Gram-negative septicaemia. Urgent broad-spectrum antibiotic therapy, after taking blood cultures, is essential and must be coupled with early relief of the biliary obstruction. A combination of antibiotics active against aerobic and anaerobic organisms is usually chosen.

CHRONIC CHOLECYSTITIS Chronic inflammation of the gallbladder is usually found in association with gallstones, but whether or not this produces symptoms, and what these are, is a subject of continuing debate. The diagnosis of chronic cholecystitis is almost certainly made too frequently. The symptoms are often vague, with 'flatulent dyspepsia', fullness in the upper abdomen and excessive eructations (belching). In a sizeable minority of cases treatment by cholecystectomy does not alter these complaints. Other conditions, such as the irritable bowel syndrome (p. 817), can produce identical symptoms. Nausea is the commonest single symptom and is often triggered by fatty food. Gallstones are best detected by ultrasound, but are common in older women and should not be used as the sole criterion for making a diagnosis of chronic cholecystitis. Careful investigation of the upper gastrointestinal tract by endoscopy is often helpful in excluding another treatable cause, such as a gastric ulcer or gastritis. In the elderly, medical dissolution of gallstones with bile salts is an alternative to cholecystectomy, but the rate of dissolution is slow and the stones usually return after withdrawal of treatment. The biliary tree, including the gallbladder, can be infected by Cryptosporidium, particularly in an immunocompromised host, including those infected with HIV. Up to 10% of AIDS patients with cryptosporidiosis have biliary involvement, with right upper quadrant abdominal pain, nausea and vomiting. The infection is very difficult to control.

BENIGN BILE DUCT STRICTURE Benign bile duct stricture is almost always a consequence of prior surgical intervention. The bile duct may have been cut and repaired, or stenosis may develop after interference with its blood supply, or be secondary to local sepsis.

1

858

Figs 16.3, 16.4

2

Fig. 16.5

Clinical features The usual history is of recurrent episodes of obstructive jaundice, fever and rigors, as a result of ascending cholangitis. After many months of recurrent symptoms progressive fibrosis and secondary biliary cirrhosis may develop.

Diagnosis and management Good-quality cholangiography is essential for diagnosis and planning treatment. Antibiotic cover for these procedures is a sensible precaution. Surgical relief of the obstruction offers the best chance of long-term relief of symptoms, but it is often difficult to find suitable sites in the biliary tree for an adequate bypass anastomosis. Fat malabsorption is common and should be treated with replacement of the fatsoluble vitamins A, D and K. Sequential use of antibiotics can reduce the number of disabling and potentially dangerous attacks of ascending cholangitis.

PRIMARY SCLEROSING CHOLANGITIS (PSC) In primary sclerosing cholangitis, short strictures from in the intrahepatic and extrahepatic biliary system. There is a close association with inflammatory bowel disease, especially ulcerative colitis. About 70% of patients with PSC have ulcerative colitis and up to 10% of patients with ulcerative colitis develop PSC. The precise pathogenetic mechanism is unknown, but it is probably (at least in part) autoimmune. Episodes of ascending cholangitis and jaundice are common. ERCP shows a characteristic beaded appearance of the bile ducts, and liver biopsy typically shows concentric, onion-skin fibrosis around the bile ducts. 1 Drainage procedures are rarely possible. Ursodeoxycholic acid may improve the biochemical abnormalities and some of the clinical symptoms, but it has not been shown to prolong survival. Corticosteroids and penicillamine have been tried, but are of no proven benefit. The disease is slowly progressive and secondary biliary cirrhosis usually develops. Cholangiocarcinoma (bile duct cancer) develops in 10-15% of cases. There is no reliable method of detecting this malignant transformation which is a contraindication to liver transplantation. Otherwise transplantation is a very successful intervention in advanced disease.

Carcinoma of the pancreas A full discussion is given on page 824.

CANCER OF THE BILE DUCTS AND GALLBLADDER Incidence and aetiology Tumours of the bile duct and gallbladder have an equal sex incidence and occur most frequently over the age of 60

16

FIG. 16.22 CT scan of the liver showing a carcinoma of the gallbladder infiltrating the liver parenchyma FIG. 16.21 The steep rise in the incidence of gallbladder cancer with age

(Fig. 16.21), when they are commoner than hepatoma. Known aetiological factors include liver flukes (Opisthorchis felineus and Clonorchia sinensis} and gallstones. The risk of these tumours in patients with long-standing gallstones is about 2%.

inoperable. With gallbladder cancer, ultrasound and CT (Fig. 16.22) may show a mass but the diagnosis is frequently made at laparotomy. With bile duct cancer PTC is usually carried out before surgery, and will show the dilated proximal ducts and the upper limit of the block. ERCP may be used to demonstrate the distal extent.

Pathology

Management

The distal common bile duct and the hilum are the commonest sites (50% of cases). Of the remainder, 20% arise from the small intrahepatic biliary ducts, and 30% from the proximal extrahepatic duct. Gallbladder cancers usually arise in the body (95% of cases). Bile duct cancers are adenocarcinomas, which usually provoke a fibrous reaction. Gallbladder cancers are usually adenocarcinomas (85%), with occasional squamous and anaplastic forms. The tumours infiltrate locally and metastasize to regional lymph nodes. They spread to involve the liver by both local infiltration and metastases.

Gallbladder cancer often cannot be resected, and a biliary decompression operation is undertaken when technically possible. Resection may involve removal of part of the liver as well as the gallbladder. Only 20% of carcinomas of the biliary tree are operable, and major resections carry a high operative mortality (10-15%). Biliary bypass procedures are effective in palliation and may be accomplished during PTC or ERCP, with the passage of a cannula (stent) through the site of the obstruction. Radiotherapy can be used to reduce tumour size, and may relieve obstruction and pain for a considerable time.

Clinical features Bile duct tumours grow slowly and cause obstructive jaundice with pruritus. The jaundice may fluctuate if the tumour sloughs and obstruction is temporarily diminished. The gallbladder may be distended if the tumour is distal to the junction of hepatic and cystic ducts. Intrahepatic tumours may not cause jaundice until they are very large. Right upper quadrant pain is common. Carcinoma of the gallbladder causes similar pain, with obstructive jaundice occurring later.

Diagnosis and investigation The presentation with pain may lead to a diagnosis of gallstones. A palpable mass usually means the tumours are

CIRRHOSIS

Definition Cirrhosis (Table 16.11) is an irreversible change in liver structure, the essential features being disorganization of the lobular architecture, the presence of regeneration nodules and increased fibrosis. 0 The size of the regeneration nodules can be smaller than of a normal liver lobule (micronodular cirrhosis), or they can embrace several portal tracts and central veins (macronodular cirrhosis). These are not separate varieties. The nodules tend to be small when there is active, continuing liver cell destruction, and larger when there is low-grade inflammation and

859

TABLE 16.11 Classification and distribution of cirrhosis Common causes Alcohol The commonest cause in the western world ' Chronic hepatitis Chronic virus infection (HBV, Delta, HCV) The commonest cause in the Middle and Far East Autoimmune Almost restricted to north European Caucasians

Primary biliary cirrhosis Broad geographical spread, 90% female Schistosomiasis Equatorial and in the Middle and Far East; fibrosis is usual, but cirrhosis is found in S. japanicum

Rare but potentially reversible causes Wilson's disease Drug-induced (hepatolenticular degeneration) Methotrexate, methyldopa, isoniazid, perhexiline Early copper chelation can reverse liver damage Galactosaemia and fructosaemia Genetically determined Haemochromatosis metabolic disorders; diet can Venesection removes iron, prevent cirrhosis improves liver function Biliary atresia Constrictive pericarditis, Early protoenterostomy can congenital venous web prevent cirrhosis Two examples of chronic venous congestion, often completely reversible Rarities Secondary biliary cirrhosis Cystic fibrosis Glycogen storage disease Indian childhood cirrhosis

Sclerosing cholangitis Neonatal hepatitis syndrome deficiency 1antitrypsin Veno-occlusive disease

necrosis. The most important pathological distinction is from portal fibrosis, which is a reversible increase in collagen deposition in and between portal tracts.

Classification Table 16.11 gives a classification of cirrhosis based on aetiology. In the developed countries of the west, alcohol is the commonest cause (about 80% of cases in the UK), but this is relatively less frequent in Africa and Asia. Cirrhosis due to chronic viral infection dominates the picture in the Middle and Far East, and in many places in Africa, but both these main causes are common in southern Europe. Whatever the cause, cirrhosis produces common complications (pp. 871-875).

1

860

Fig. 16.6

Presentation and general clinical features If the process causing the cirrhotic transformation is inactive, presentation is usually delayed until portal hypertension and liver cell failure produce symptoms of ascites, haematemesis from bleeding varices, or encephalopathy. In older males, primary hepatocellular carcinoma (primary hepatoma) is a frequent complication, usually associated with an underlying inactive macronodular cirrhosis. In some cases the disease is detected at an asymptomatic stage during routine physical examination or biochemical screening. Some of the causes of cirrhosis produce more florid liver damage, and may present with jaundice, pruritus or symptoms of other organ involvement before cirrhosis has become established. On examination, cutaneous stigmata of chronic liver disease are usually prominent. Spider naevi are restricted to the upper half of the body. Possible causal mechanisms are inappropriate vasodilatation or high oestrogenic activity, which are normal features of pregnancy and sometimes found in females taking oral contraceptives. Leukonychia is a manifestation of low serum albumin levels and is not specific for liver disease. The term 'liver palms' refers to a blotchy erythema of the thenar and hypothenar eminences, and may be a visible manifestation of a widespread disturbance in vasomotor tone. Dupuytren's contracture is due to a thickening of the palmar fascia and is particularly common in alcoholic liver disease. A flapping tremor of the outstretched hands (asterixis), best seen when the wrists are extended, is a feature of hepatic encephalopathy. Feminization in males, with gynaecomastia, testicular atrophy and loss of body hair, is particularly common in alcoholic liver disease and haemochromatosis. It is not due to high oestrogen levels alone, but also to more complex gonadal and hypothalamic dysfunction. Female patients with cirrhosis may develop amenorrhoea. Examination of the abdomen usually reveals hepatosplenomegaly, although in advanced cirrhosis the liver may become very small. Ascites may be present, and striae may be a sign of previous abdominal distension. Patients with portal hypertension may have dilated veins on the abdominal wall.

FURTHER READING ON CIRRHOSIS Cabrera J, Maynar M, Granados R et al 1996 Transjugular intrahepatic portosystemic shunt versus sclerotherapy in the elective treatment of variceal haemorrhage. Gastroenterology 110(3): 832-839. Fernandez J, Navasa M, Gomez J, Colmenero J, Vila J, Arroyo V, Rodes J 2002 Bacterial infections in cirrhosis: epidemiological changes with invasive procedures and norfloxacin prophylaxes. Hepatology 35:140-148. Henrikson J H 1995 Cirrhosis: ascites and hepatorenal syndrome. Recent advances in pathogenesis. J Hepatol 23 (Suppl 1): 25-30. Laffi G, Foschi M, Masini E et al 1995 Increased production of nitric oxide by neutrophils and monocytes from cirrhotic patients with ascites and hyperdynamic circulation. Hepatology 22(6): 1666-1673.

16

FIG. 16.23 Deaths resulting from chronic alcoholic liver disease in the UK

Wong F, Logan A, Blendis L 1996 Hyperinsulinemia in preascitic cirrhosis: effects on systemic and renal hermodynamics, sodium homeostasis, forearm blood flow, and sympathetic nervous activity. Hepatology 23(3): 414-422.

CHRONIC ALCOHOLIC LIVER DISEASE

Epidemiology Chronic alcoholic liver disease is the commonest cause of cirrhosis in Europe, and is increasing in prevalence in most countries (Fig. 16.23). In district general hospitals in Britain, 80% of all cases of cirrhosis are associated with alcohol abuse. In the USA, deaths from cirrhosis, mostly alcohol-related, increased by 72% between 1950 and 1974, making it the fourth commonest cause of death in white male adults. This was associated with a sharp increase in alcohol consumption over this period, but a reversal in trend saw a reduction of 40% over the subsequent 15 years.

Pathology There is a striking range of histopathological changes produced by alcohol. The most consistent is increased fat deposition (steatosis), 1 which is almost invariably present in those drinking more than 80 g (one bottle of wine or 3 pints of beer - 6 units, p. 248) of alcohol per day. Although a large proportion of the cytoplasm of affected hepatocytes is occupied by a single large triglyceride inclusion, liver function is surprisingly normal. The fatty change is rapidly reversible during abstinence. In most cases isolated steatosis does not progress to cirrhosis, although in some there may be slowly increasing fibrosis around central veins, leading to stellate scarring and incomplete septum formation. It is possible that this could eventually lead to cirrhosis. The particular histological changes seen in alcoholic hepatitis (p. 855) are thought to be of fundamental importance in determining progression to cirrhosis.

The perinuclear eosinophilic inclusion bodies (Mallory's hyaline), one of the most obvious features, are probably condensed and disorganized fragments of the cytoskeletal framework of the hepatocyte. They are not specific to alcoholic liver disease, but may also be seen in Wilson's disease, primary biliary cirrhosis, Indian childhood cirrhosis and perhexiline-induced liver injury. Collagen deposition is another striking feature of alcoholic hepatitis. It can be pericellular, in the space of Disse, and around central veins (central hyaline sclerosis). This latter lesion is particularly associated with rapid progression to cirrhosis, and is commoner and more severe in women than in men. Collagen bridges eventually develop between central veins and portal tracts, isolating groups of hepatocytes which form the regeneration nodules characteristic of cirrhosis.

Pathogenesis The metabolism of ethanol to acetaldehyde and acetate is normally dependent on the enzymes alcohol dehydrogenase and acetaldehyde dehydrogenase (Fig. 16.24). These oxidation reactions are associated with the formation of NADH from NAD and alter the redox state of the cell. This in turn has profound effects on lipid and carbohydrate metabolism, one of which results in steatosis. In habitual drinkers a microsomal mixed-function oxidase, the microsomal ethanol oxidizing system (MEOS), is increased by enzyme induction, and is also responsible for the production of acetaldehyde. Breakdown of acetaldehyde may then become the rate-limiting step in ethanol metabolism, and it is thought that this toxic metabolite may be responsible for the liver cell injury. The direct correlation, in populations, between alcohol intake and risk of cirrhosis obscures the wide individual variation in susceptibility, for which there is at present no satisfactory explanation. Retrospective analyses of lifetime alcohol consumption in patients with cirrhosis have shown that women develop the disease about twice as fast as men, and that individuals who have inherited the histocompatibility antigen HLA B8 have a much shorter history of

861

CASE STUDY 16.3 PROGRESSIVE ABDOMINAL SWELLING IN A 45-YEAR-OLD WOMAN WITH A HISTORY OF DRINKING 5 UNITS OF ALCOHOL EACH DAY A 45-year-old woman attended her doctor with a swollen abdomen after a friend jokingly told her she looked pregnant. She had had amenorrhoea for 5 months but assumed that she was perimenopausal. Her only other complaint was tiredness, which she attributed to a stressful secretarial job in a large business. She had a previous history of deep venous thrombosis while taking the oral contraceptive pill. Her alcohol intake was constant at about 35 units per week. She did not smoke and had no risk factors for hepatitis. Physical examination revealed palmar erythema, facial telangiectasia and 11 spider naevi. The abdomen was distended with fluid, the liver could not be palpated and the spleen was palpable 4cm below the costal margin. The remainder of the examination was normal. The laboratory investigations were: haemoglobin ll.lg/dL, white cell count 11.7 x 109/L, platelets 86 x 109/L, prothrombin time 17s (control 13s), serum bilirubin 34umol/L, aspartate aminotransferase 54IU/L, alkaline phosphatase 99IU/L and serum albumin 26g/L. Question 1, What is the most likely diagnosis? The most likely diagnosis is ascites and portal hypertension secondary to chronic liver disease. The splenomegaly and low platelet count point towards long-standing portal hypertension and the low serum

albumin suggests cirrhosis. The past history of venous thrombosis should prompt consideration of the Budd-Chiari syndrome (hepatic vein thrombosis), but this would normally be associated with hepatomegaly rather than a small liver. The cirrhosis could be due to many causes, although the alcohol intake was sufficiently high to consider it the most likely aetiology. Question 2. What further investigations are required? A range of blood tests should be performed to look for other causes of cirrhosis. Serum immunoglobulin levels could point to autoimmune hepatitis if the IgG is elevated, or alcoholic cirrhosis if the IgA is increased. Of the autoantibodies, the antinuclear factor (ANA), antismooth muscle (ASM) and antiliver kidney microsomal (LKM) antibodies are most suggestive of a diagnosis of autoimmune hepatitis. Chronic viral hepatitis B and C should be excluded by the appropriate serological tests. Other routine investigations include serum ferritin, 1-antitrypsm phenotype and serum caeruloplasmin (in patients under the age of 40 years). An ultrasound of the liver should be performed to assess the size and shape of the liver, confirm the presence of ascites and splenomegaly, screen for hepatocellular carcinoma and assess the patency of the portal vein. A liver biopsy is performed once it can be

heavy drinking. This particular HLA antigen is strongly associated with organ-specific autoimmunity, and there is other evidence suggesting that immune reactions against Mallory's hyaline, or against new antigens on the surface 1 862

MCQ 16.6

2

MCQ 16.7

carried out with minimal risk (absence of ascites, near normal coagulation and platelet count >50 x 109/L). In this case the liver biopsy showed micronodular cirrhosis consistent with a diagnosis of alcoholic liver disease.

Question 3. How should this case be managed? Abstinence from alcohol is extremely important. A low-salt diet should be instituted and fluids restricted to 1.5L per day. Spironolactone is commenced in an initial dose of 50-100 mg daily and the dose titrated to achieve a weight loss of 0.5 kg per day. Frusemide is added if the desired response is not achieved with doses of Spironolactone in the order of 300-400 mg daily. A paracentesis is often performed at the start of therapy to reduce the length of time required to control the ascites. A sample of the ascites should be sent for analysis to exclude infection (white cell count >400/mL) and determine protein content (confirm it is a transudate and initiate antibiotic prophylaxis if the protein level is very low). In this case, abstinence was easily achieved. She was a regular consumer of alcohol but did not exhibit physical or psychiatric signs of dependency. The ascites responded rapidly to dietary manipulation and diuretics, and she remains well 7 years after presentation.

of hepatocytes, induced by acetaldehyde, could be responsible for liver cell damage.

Diagnosis The presenting features of cirrhosis and alcoholic hepatitis have already been discussed. However, some cases with significant liver damage are asymptomatic, and liver biopsy is the only way reliably to assess the degree of liver injury.

TABLE 16.12 Classification of chronic hepatitis Type Autoimmune Classic LKM positive type

Chronic virus infection Hepatitis B Delta hepatitis FIG. 16.24 The metabolism of ethanol Ethanol metabolism takes place largely in the mitochondria. The main product is acetaldehyde (A), which may itself be damaging. Hydrogen ions (B) are an important by-product which can drive many metabolic reactions in the liver cell cytosol, to create harmful effects (C, D).

Recognition of alcohol abuse is crucial, so that disease can be detected at a stage where abstinence can still lead to complete resolution of tissue damage. Important points in the history and examination are discussed on pages 845 and 860. A high mean corpuscular volume and serum -GT are important clues from laboratory tests.

Management Malnutrition is common in those with alcohol addiction, and should be treated with vitamin supplements and a high-protein diet. Fits are best controlled with diazepam, and delirium tremens prevented and treated with clomethiazole (chlormethiazole) or chlordiazepoxide. Other causes of confusion and coma, such as Wernicke's syndrome, subdural haematoma, hypoglycaemia and hepatic encephalopathy, must be excluded. Abstinence is the key to successful long-term management (p. 251). Ninety per cent of patients with alcoholic hepatitis alone will show complete regression of the liver damage, and those with cirrhosis who remain abstinent have a much better prognosis than those who continue to drink. O

CHRONIC HEPATITIS 2 Chronic hepatitis is a syndrome consisting of several different diseases with similar histological appearances on liver biopsy (Table 16.12). The portal tracts are enlarged and contain a chronic inflammatory cell infiltrate of lymphocytes, plasma cells and macrophages. The cells extend beyond the edges of the portal tracts through the limiting plate of periportal hepatocytes, and there is 'interface

Hepatitis C

Drug-induced

Metabolic Wilson's disease

Distribution

Features

Northern Europe, USA, Australia, South Africa Europe

Female/male ratio 7:3; high litre antinuclear antibodies Female/male ratio 1:1; high litre liver/kidney microsomal antibodies

Mediterranean, Middle and Far East Mediterranean and Middle East

Female/male ratio 1:9; low titre antinuclear antibodies Endemic in HBV carriers in Mediterranean; intravenous drug abusers in northern Europe Commonest post-transfusion hepatitis; also transmitted by blood products Examples are methyldopa, isoniazid, nitrofurantoin and dantrolene, usually like autoimmune chronic hepatitis

Wide distribution

A phenomenon of the developed world

Wide distribution

Haemochromatosis

Wide distribution, but not Brittany and Utah Mormons

Alcohol

Wide distribution, but particularly the developed world

16

Can mimic any liver disease, so easily missed Usually inactive fibrosis, but rarely resembles chronic active hepatitis, almost exclusively males affected Controversial as a cause of chronic hepatitis; may be in recovery phase of alcoholic hepatitis

hepatitis' of liver cells in this part of the lobule. Collagen deposition is prominent in areas of liver cell damage, leading to the formation of fibrous septa linking portal tracts to each other and to central veins. If the disease process continues unchecked, cirrhosis is an inevitable consequence.

AUTOIMMUNE Aetiology and pathogenesis Autoimmune chronic hepatitis was the first to be described. It is best considered as one of the organ-specific autoimmune diseases. There are two main subtypes. In the classic variety women are affected more commonly than men and antinuclear and/or smooth muscle antibodies are commonly found in the serum, whereas the other variety is characterized by the findings of a liver/kidney microsomal antibody in serum which reacts with one of the P450 microsomal liver enzymes. In both forms of the disease

863

serum globulin and IgG levels are high, and there is evidence of both cellular and humoral liver-specific autoimmune reactions directed at normal antigens on the surface of hepatocytes. HLA B8 and DR3 antigens are found in a high proportion of cases. Antibodies reacting with liver cell membrane components are almost always present, and probably damage hepatocytes in cooperation with cytotoxic lymphocytes bearing Fc receptors for the antibody molecules. There is a profound disturbance in immunoregulation associated with functionally abnormal suppressor T cells.

Clinical features Two-thirds of patients present with an illness indistinguishable from acute viral hepatitis. The remainder have symptoms due to an already established cirrhosis, such as ascites, haematemesis from bleeding varices, or encephalopathy. Half of those with an initial acute hepatitis remain jaundiced and have biochemical evidence of continued liver necrosis. The others appear to recover normally, but develop a second episode of jaundice some months later. Multisystem involvement is common, and includes arthralgia, ulcerative colitis, autoimmune thyroid disease and haemolytic anaemia. Cutaneous stigmata of chronic liver disease are often present, and more than 70% of patients already have cirrhosis when the diagnosis is first made.

cautiously withdrawn. However, relapse is very common (60-80%) and can be severe, necessitating prompt reintroduction of corticosteroids.

CHRONIC HEPATITIS B VIRUS INFECTION Worldwide, HBV is the commonest type of chronic active hepatitis, being particularly prevalent in the Mediterranean, the Middle and Far East, and tropical Africa. The serology and structure of HBV have been described earlier (pp. 848-850).

Clinical features Many patients are asymptomatic, the chronic virus infection being detected on screening for blood donations or the investigation of other diseases. When symptoms do occur they are often mild and non-specific, malaise being the commonest; some progress from an acute HBV infection. There is a striking male preponderance. The incidence of multisystem disease is lower than that found in the autoimmune variety. The liver function tests usually show a relatively mild hepatitic picture, with only moderately raised levels of serum transaminases.

Serological markers Investigations Liver function tests usually show a hepatitic picture with high serum transaminases. A marked elevation in the total globulin (more than 40g/L), with a high IgG, is a useful pointer to this subgroup. The diagnosis is made on liver biopsy, 1 and the presence of the appropriate non-organ specific autoantibodies indicates the correct classification.

Management The response to corticosteroids is excellent. The survival rate is greatly increased, with more than 60% alive at 10 years, although cirrhosis, if absent initially, may still develop. Azathioprine in a dose of l-2mg/kg is a useful adjunct to steroid therapy and may allow a lower dose of steroid to be used. This can be particularly important in older women who are prone to osteoporosis. Although serum transaminase levels are often used to monitor progress, liver biopsy is useful in establishing the activity of the disease during follow-up. If the disease has remained inactive for more than 2 years, steroid treatment can be

Fig. 16.7

864

Case 16.2

MCQ 16.8

There are two phases to the infection. Initially there is active viral replication, with HBsAg, HBeAg, HBV DNA, and the DNA polymerase of the virus detectable in the serum. In this phase the histological features of chronic hepatitis are most often present. After several years there may be a rather abrupt cessation of viral replication, with loss of serum HBeAg, HBV DNA and DNA polymerase. This change is often preceded by a rise in the levels of serum transaminases, but is frequently followed by a striking decrease in the activity of the disease, with a reduction in the extent of the inflammatory infiltrate in the liver. HBsAg levels in serum decrease because free viral DNA has been lost, but usually remain detectable in serum, probably because there has been integration of the viral genome into the hepatocyte DNA. The defect responsible for the failure to clear the virus after the initial infection has not been identified. Antibodies reacting with whole virions are absent, which may allow continued viral penetration of uninfected hepatocytes. Cellular immunity to HBeAg on the surface of infected liver cells, which is probably responsible for liver cell necrosis in acute HBV infection, is present and may be a component of the inflammatory liver lesion. On the other hand, factors must be present which interfere with effective clearance of infected cells. Autoimmune reactions to normal liver cell membrane antigens, similar to those found in 'autoimmune' chronic hepatitis, can be demonstrated, which could explain the similarity of the periportal histological

changes in two diseases with quite distinct aetiological backgrounds.

Management There is current interest in antiviral agents, particularly the -interferons, which appear to act by stimulating existing immune responses against the virus as well as by suppressing virus replication. Although immediate effects on viral replication can be convincingly demonstrated, only 20-40% of cases show long-lasting benefit. Treatment is particularly effective in those with pre-existing active liver inflammation. New nucleoside antiviral compounds have been developed which show considerable efficacy against HBV. They appear to suppress viral replication almost completely and are well tolerated. Lamivudine is the most widely used, and the main problem experienced to date is the emergence of viral mutants at a rate of about 7% per year of therapy. The clinical significance of these mutants is uncertain. Corticosteroids are of no benefit in the phase of active viral replication, and may be harmful. Liver transplantation has been used to treat patients with end-stage chronic liver disease or small hepatocellular carcinoma. Recurrence of HBV infection was problematic, but strategies have emerged to prevent reinfection in over 80% of cases. These rely on the suppression of viral replication before transplantation, e.g. with lamivudine, and passive immunoprophylaxis with anti-HBs (HBIg) indefinitely after transplantation. The results of transplantation are particularly good in patients with coexisting Delta virus infection because of a natural suppressing effect it has on hepatitis B virus replication, even in the presence of immunosuppressive medication.

CHRONIC DELTA VIRUS INFECTION The histological changes of chronic hepatitis are often present in chronic Delta virus infection (p. 850). The diagnosis is made by finding anti-Delta antibodies in serum or demonstrating Delta antigen in the nuclei of liver cells by immunofluorescence. Progression to cirrhosis seems to occur quite rapidly and the prognosis is poor. Little is known about the mechanisms of tissue damage, and no specific treatment has been shown to be effective.

CHRONIC HEPATITIS C VIRUS INFECTION GO The discovery of hepatitis C virus and tests for its serological and genomic detection have allowed a clear picture to emerge of a common cause of chronic liver disease throughout the world.

Clinical features The majority of carriers of the hepatitis C virus are rela-

tively asymptomatic and their life expectancy is not altered. The commonest symptom in these patients is fatigue, which can be severe even in patients with normal or near-normal blood tests and liver histology. Significant liver disease develops in 20-30% of patients who have necroinflammatory disease, fibrosis and ultimately cirrhosis. This process evolves over 20-30 years and is accelerated by moderate to high alcohol consumption. Once cirrhosis has developed, there is a 1.5-2% incidence of hepatocellular carcinoma per year. In the absence of this complication, 85% of patients are alive 10 years after the diagnosis of cirrhosis has been established. Extrahepatic manifestations and associations include porphyria cutanea tarda and cryoglobulinaemia-associated disease e.g. glomerulonephritis.

16

Serological markers Exposure to hepatitis C is confirmed by the detection of antibodies in serum. About 85% of patients with antibodies have HCV RNA in their serum and this is now the most commonly used confirmatory test. In the absence of HCV RNA testing, a RIBA test which screens for a range of antibodies is used to confirm exposure to hepatitis C. An IgM antibody test has not yet been developed, and therefore recent infections cannot be confirmed by serological testing.

Management Assessment of the activity of the liver disease in carriers can be difficult as the conventional liver function tests may be normal. However, if viral RNA is present, then liver biopsy should be performed to assess the need for therapeutic intervention. a-Interferon is effective in suppressing viral replication and normalizing serum transaminases in about 45% of patients who have precirrhotic disease. However, a sustained virological response occurs in only 9-29% of cases, depending on the dose and duration of interferon therapy. The response rate among patients with cirrhosis is negligible. Combination therapy using interferon 3 million units thrice weekly and ribavirin 1-1.2 g daily induces a sustained virological response in 45% of patients without cirrhosis. A new long-acting preparation of interferon, PEGylated interferon, in clinical trials yielded a virological response rate of 25% in patients with cirrhosis. Liver transplantation is an effective treatment for endstage disease but reinfection of the graft is almost inevitable. There are so far no effective prophylactic or therapeutic measures to modify the rate or impact of reinfection. The latter is very variable, but cirrhosis can develop 10 times more rapidly than in the immunocompetent individual, possibly because of the much higher viral loads seen in immunosuppressed patients. The overall results of liver transplantation are the same as for other aetiologies of cirrhosis up to 8 years after transplantation.

865

DRUG-INDUCED CHRONIC HEPATITIS Although rare, these cases are important as the disease almost always responds to drug withdrawal. All reported cases are examples of idiosyncratic reactions, which have not been reproduced experimentally. The drugs most frequently implicated are isoniazid, a-methyldopa, nitrofurantoin and oxyphenisatin. In many cases the disease is very similar to the 'autoimmune' type, with antinuclear and smooth muscle antibodies present in serum.

OTHER CAUSES The most important of these is Wilson's disease (p. 869), which can mimic the changes seen in viral or autoimmune cases. Because long-term survival on D-penicillamine is possible, it is clearly of great importance to exclude this condition in all cases of chronic active hepatitis in children and young adults. The histological features of chronic hepatitis can also be found in some cases of 1-antitrypsin deficiency, and in alcoholic liver disease.

PRIMARY BILIARY CIRRHOSIS Like autoimmune chronic hepatitis, primary biliary cirrhosis is a condition which is not necessarily associated with cirrhosis at presentation, although this frequently develops later in the course of the disease. The essence of the disease is a slowly progressive destruction of intra-hepatic bile ducts, possibly due to immune-mediated injury.

Incidence and aetiology In most series more than 90% of patients are women, with most being in middle age. In the UK the incidence is between 6 and 12 per million, with a prevalence of 40-80 per million. Familial cases are well described and a significant association with the HLA class II antigen HLA DR8 has been confirmed in a number of reports. This is a relatively rare antigen in both the controls and PBC patients, but the frequency of DR8 in PBC is two to six times higher than in controls. The antibodies in serum reacting with mitochondria, which are such a strong feature of the disease, are also found in family members, but in most cases there is no apparent liver or bile duct injury associated with this finding.

Pathology

866

Four stages in the histopathology of the disease have been described. They are not necessarily sequential, and there is considerable overlap between them.

RECENT ADVANCES HEPATITIS MUTANTS ESCAPE THE IMMUNE ATTACK Both HBV and HCV have proved capable of dodging the immune attack. Both viruses have a replication mechanism which does not faithfully copy their DNA or RNA, creating mutants with amino acid substitutions which can, in theory, alter their immune reactivity. In the case of HBV, the viral nucleocapsid (HBcAg) is one of the most important targets of the cellular immune attack on infected hepatocytes, and in acute viral hepatitis the vigorous polyclonal attack is able to eliminate all the viral species present. In chronic infection the immune tolerance gradually wanes and the T-cell response, when it occurs, is almost certainly oligoclonal, restricted to a small number of peptides. It follows that some of the mutants may be able to escape this restricted attack. Analysis of the HBV genome from patients with chronic active hepatitis B does show a variety of different mutants present in the peripheral blood, and in some experiments it has been possible to show that the mutants have core peptides which do not express amino acids that are key to stimulating existing cytotoxic T cells. In the case of HCV, one of the major epitopes on the cell surface to which neutralizing antibodies are made is in the middle of a hypervariable segment of the RNA genome. Changes in the amino acid sequence of this region alter the conformation of the surface determinant on the virion, and have been shown to escape from the neutralizing effect of the antibodies made to the original structure. These frequent changes in the virus surface conformation will almost certainly create considerable difficulties in designing a suitable vaccine against HCV. In both these instances, these changes in the viral genome are not only a successful ploy by the virus to ensure continued survival, but are also a source of frustration to those designing a therapeutic strategy to eliminate the viruses. 1. First stage. There is focal damage to the larger intrahepatic bile ducts, with a surrounding chronic inflammatory infiltrate of lymphocytes, plasma cells, eosinophils and macrophages. In some areas the infiltrate may be organized into an epithelioid cell granuloma, lying close to a damaged duct. 2. Second stage. Some of the larger ducts are replaced by lymphoid aggregates, and there is proliferation of smaller ductules. Fibrosis is present. 3. Third stage. Fibrosis extends beyond the edges of the portal tracts. There is often periportal cholestasis and accumulation of copper-binding protein. 4. Final stage. The features are those of an established cirrhosis. Ill-defined collections of lymphocytes and a paucity of bile ducts may be the only clues to the aetiology.

CASE STUDY 16.4 POOR CONCENTRATION AND REDUCED MENTAL ABILITY IN A 44-YEAR-OLD MAN WITH KNOWN HEPATITIS C INFECTION A 44-year-old man with an established diagnosis of cirrhosis secondary to hepatitis C was assessed in the outpatient clinic. He complained of tiredness, constipation, sleeping during the day and then being unable to go to sleep at night. He also said that his concentration was poor and he was no longer able to complete the crossword he did on a daily basis. He had acquired hepatitis C during a 7-month period of intravenous drug use while at university. He had consumed 40-60 units of alcohol per week consistently until the diagnosis of hepatitis C was made 3 years previously, but was now totally abstinent. His medication was spironolactone 300 mg, frusemide 40 mg and dihydrocodeine as required for analgesia. Physical examination revealed that he was not jaundiced but had palmar erythema, facial telangiectasia, spider naevi and gynaecomastia. A liver flap was not present. The liver was not palpable but the spleen was enlarged and extended 5 cm below the costal margin. There was no evidence of ascites. A number connection test was performed and this was completed in 54 seconds. The laboratory investigations were: haemoglobin 13.1 g/dL, white cell count 6.1 x 109/L, platelets 55 x 109/L, prothrombin time 19s (control 13s), serum bilirubin 24umol/L, aspartate aminotransferase 113IU/L, alkaline phosphatase 102IU/L and serum albumin 27 g/L. The afetoprotein was 34 lU/mL. An ultrasound examination showed a small liver with no focal lesion, a patent portal vein and splenomegaly.

Question 1. What is the clinical assessment? He has cirrhosis with impaired synthetic function (low albumin and prolonged prothrombin time). There is no evidence of a complicating hepatocellular carcinoma, as evidenced by the a-fetoprotein level and the ultrasound examination. The tiredness is characteristic of both cirrhosis and infection with the hepatitis C virus. The reversal of day/night sleep pattern is characteristic of early hepatic encephalopathy, and this diagnosis was further supported by the length of time required to complete the number connection test (normal less than 25 seconds). Constipation and dihydrocodeine were identified as two possible causes of the encephalopathy. The former was treated with lactulose and dihydrocodeine was discontinued. Three months later the patient was admitted to hospital in a coma. He responded only to painful stimuli. The other change on clinical examination was the presence of ascites. A therapeutic paracentesis was performed and the white cell count in the ascites was 650 per mL, with 90% neutrophils. An ultrasound examination of the liver showed no changes other than thrombosis of the portal vein. The only significant alteration in the blood tests was a haemoglobin of 9.3 g/dL, which represented a fall of 3.7 g/dL.

Immunopathogenesis Antimitochondrial antibodies are directed at an antigen on the inner membrane of the mitochondrium, and are unlikely to be of pathogenetic importance. However, they

16

Question 2. What is the cause of the coma and what precipitated it? Coma is typical of advanced hepatic encephalopathy and this is the most likely diagnosis, given the background history. He now has two of the most frequent causes of severe encephalopathy, infection and gastrointestinal haemorrhage. The white cell count in the ascites is diagnostic of spontaneous bacterial peritonitis, as it is above 400 per mL and is composed predominantly of neutrophils. A positive bacterial culture is achieved in no more than 70% of cases and is not necessary to substantiate the diagnosis. Spontaneous bacterial peritonitis does not usually present with classic signs of peritonitis, and a diagnostic paracentesis is mandatory to screen for this complication in all encephalopathic patients who have ascites. The possibility of gastrointestinal haemorrhage is suggested by the falling haemoglobin despite the absence of a history of haematemesis or melaena. Bleeding from oesophageal varices was confirmed and treated endoscopically. Bleeding oesophageal varices and spontaneous bacterial peritonitis are not random events occurring coincidentally, but are a common combination probably caused by bacterial translocation across the bowel wall into ascitic fluid. The thrombosis of the portal vein documented by ultrasound is a well recognized complication of cirrhosis and was the precipitant for the variceal bleed and the development of ascites.

are clearly an important feature of the disease. An attractive but unproven hypothesis is that they are cross-reactive antibodies generated by an immune response to an as yet unidentified bacterium which is in some way responsible

867

for the bile duct destruction. This hypothesis has been reinforced by the finding that the antimitochondrial antibodies are directed at the pyruvate dehydrogenase complex and cross-react extensively with the corresponding bacterial enzyme complex in Escherichia coli. Granulomas are usually indicative of a cellular immune reaction, but it is not clear what the antigenic stimulus could be. T cells sensitized to antigenic components in bile ducts may be responsible for their destruction. The disease in other organs may be an indication of a generalized disturbance in immunoregulation, but could also be due to an observed antigenic cross-reactivity between bile duct antigens and secretory duct epithelium in other organs.

Clinical features The disease is much commoner (9:1) in females than in males, and usually presents in the fifth and sixth decades. Pruritus is often the first symptom, and may lead to a dermatological consultation. Weight is usually well maintained and abdominal pain is unusual. Jaundice is a late feature. On examination there is often quite marked hepatosplenomegaly and skin pigmentation. Skin xanthomas are common. 1 With the increasing availability of automated biochemical profiles and screening tests for non-organ specific autoantibodies in serum, asymptomatic cases are being seen more frequently. In these patients minor hepatomegaly may be the only abnormality. The most striking serological feature is the presence of antimitochondrial antibodies in serum. These are of considerable help in the differential diagnosis, as they are found in 95% of patients with primary biliary cirrhosis but are usually absent in prolonged extrahepatic obstruction. 2 Associated diseases Disease in other organs is very common, and includes one or more of the components of the CRST syndrome (calcinosis, Raynaud's phenomenon, sclerodactyly, telangiectasia), autoimmune thyroiditis, renal tubular acidosis, rheumatoid arthritis and the 'sicca complex' (dry eyes and dry mouth).

Course and management In symptomatic cases average survival is 6 years, but it is difficult to give an accurate prognosis in individual cases. The bile duct destruction is slowly progressive, but bilirubin levels usually remain stable for several years before rising steadily in the last year or two of the illness. Steatorrhoea with malabsorption of fat-soluble vitamins can be a problem, and it is important to give supplements of

868

1

Fig. 16.8

4

MCQ 16.10

2

MCQ 16.9

0 Fig. 16.9

calcium and vitamins D and K. Cholestyramine usually controls pruritus. Corticosteroids are not usually given because of the likely exacerbation of osteoporosis in postmenopausal women. Ursodeoxycholic acid may alleviate symptoms and improve the biochemical profile, and consequently, is frequently prescribed. However, there is no evidence of an associated improvement in histological appearances or survival. Several other agents have been tried but all have failed to provide evidence, in controlled clinical trials, of a clear and clinically significant benefit. These include ciclosporin, an immunosuppressive drug which acts by interfering with the production of interleukin-2, an important stimulator of lymphocyte proliferation; colchicine, an antifibrotic agent; and D-penicillamine, which chelates copper, inhibits the synthesis of collagen and immunosuppresses. In advanced cases liver transplantation is a very successful therapeutic option, with survival rates of 90-95 % at 1 year. Patients are selected for transplantation if the prognostic models predict a survival time of 12-18 months, complications of end-stage liver disease develop and are difficult to control, or occasionally when the quality of life is very poor owing to severe lethargy or intractable pruritus. The latter is relieved immediately by liver transplantation and the overall quality of life in all patients is remarkably improved.

HAEMOCHROMATOSIS Increased iron deposition in hepatocytes is the hallmark of haemochromatosis. In idiopathic haemochromatosis this is due to a genetically determined increase in iron absorption. The inheritance is autosomal recessive and the responsible gene, called the HFE gene, is located on chromosome 6. The major mutation associated with clinical haemochromatosis is C282Y and a second minor mutation is H63D. The exact mechanism of the defect in control of iron absorption, however, is still not fully understood. The frequency of the haemochromatosis allele is very high (1:10) in European populations, and the frequency of homozygotes (1:100) makes this one of the commonest inherited disorders. Haemosiderosis - secondary iron overload - is usually due to a chronic haemolytic anaemia such as thalassaemia. The increased liver iron content is due to both increased iron release and repeated blood transfusions (p. 1231).

Clinical features There is a striking male preponderance, almost certainly because menstruation in women prevents a significant rise in iron stores. The disease usually presents in middle age. 'Bronzed diabetes' is a useful reminder of two of the important clinical features. The pigmentation is due to increased melanin in the skin, and is often seen best over the shins, where it gives a slate-grey appearance. Patients may present with diabetes, sometimes well before the liver

disease becomes symptomatic. It is associated with iron deposition in the pancreas. An arthritis, particularly affecting the knees and metacarpophalangeal joints, can also be a presenting feature, and is associated with the deposition of calcium pyrophosphate crystals in the affected joints. On examination, gynaecomastia, testicular atrophy and loss of body hair are often present, and seem to be due to a combination of partial testicular and pituitary failure, associated with iron deposition in these areas. Hepatomegaly is invariably present, but splenomegaly is uncommon. In younger patients congestive cardiomyopathy can be an unusual but troublesome complication. The rate of hepatocellular carcinoma is among the highest seen for the different aetiologies of cirrhosis.

Diagnosis Liver function tests can be entirely normal and haemochromatosis is a good example of a disease that must be considered before it can be diagnosed. The definitive diagnosis is established by detection of the genetic mutations, one of which is present in over 95% of cases. Liver biopsy allows demonstration of the iron deposition in liver cells, usually in a periportal distribution. 0 Extensive portal fibrosis develops before there is progression to a true cirrhosis, with regeneration nodules. Serum iron levels are high, with greater than 90% saturation of the total iron-binding capacity. Serum ferritin levels are a good reflection of total body iron content, but do not correlate precisely with the concentration of liver iron.

Management and prognosis The gradual depletion of body iron stores is best achieved by regular venesection, initially 1 unit weekly. Serum ferritin and a repeat liver biopsy are the most useful markers of complete removal of the excessive iron deposits. The severity of the associated diabetes tends to decrease, and there is a general improvement in malaise and a decrease in liver size. The changes of cirrhosis in the liver are not reversible, and one of the long-term risks is the development of primary hepatocellular carcinoma. Venesection is most effective when started early in the course of the disease, before the development of diabetes or cirrhosis, and investigation of the patient's relatives is an essential part of management. In 1935 the mean survival after diagnosis was 4.4 years, in 1969 the 5-year survival rate was 89%, and as more asymptomatic family members are picked up and treated it is becoming clear that the disease can be prevented if excess iron is removed at an early stage and the future accumulation of iron prevented. 4

WILSON'S DISEASE In this disease, genetically determined accumulation of copper in the liver, basal ganglia, kidneys and corneas can

16

FIG. 16.25 Wilson's disease The postulated block in copper metabolism in Wilson's disease is a failure to excrete copper trom lysosomes into bile. This leads to a build-up of the toxic metal in the liver and other tissues, particularly the basal ganglia in the brain.

produce a wide variety of liver lesions with or without associated extrapyramidal dysfunction. It is an autosomal recessive condition with an incidence of about 1 in 50000. The responsible gene, ATP7B, is located on chromosome 13 and codes for a copper carrier protein. A large number of mutations have been described in patients with Wilson's disease. Heterozygous individuals remain in virtually zero copper balance throughout their life and show no clinical features of the disease. Radiocopper studies indicate that the inherited defect in copper metabolism principally affects copper excretion from hepatic lysosomes into bile (Fig. 16.25). Pathologically and clinically the liver is often the first organ to be affected. The appearance of lipid droplets in the liver cell cytoplasm is the first abnormality, followed by increasing fibrosis and cirrhosis.

Clinical features Wilson's disease is a great mimic of many other liver diseases, and because of the excellent response to Dpenicillamine in the early stages it must always be considered and excluded in any child or young adult with chronic liver disease. In some cases there is a combination of fatty change, Mallory's hyaline and fibrosis, which may be wrongly attributed to the effects of social drinking. In others there is a mononuclear infiltrate in the portal tracts, with piecemeal necrosis of periportal hepatocytes, which can lead to a diagnosis of non-A, non-B or autoimmune chronic hepatitis. In children an 'acute hepatitis', sometimes progressing to acute liver failure, may be a presenting feature. This is usually accompanied by acute haemolysis. At the other extreme the disease may remain silent until a slowly progressive, inactive cirrhosis presents with haematemesis, ascites or encephalopathy. Neurological abnormalities may be absent in younger patients; when present, they are often subtle. They include deteriorating school performance, changes in personality, slurring of

869

speech, clumsiness, and difficulties with fine movements. 1 In adults, dysarthria and bulbar palsy may accompany more classic extrapyramidal signs and symptoms.

Diagnosis Brown deposits of copper in the margins of the cornea, called Kayser-Fleischer rings, are diagnostic. They are always present in patients with neurological symptoms, but can be absent in young patients presenting with liver disease. They are best seen on slit-lamp examination. Serum caeruloplasmin levels are typically low, but may also be decreased in chronic hepatitis and acute liver failure as well as in about 10% of healthy heterozygotes. Urinary copper excretion is high, and rises dramatically after a test dose of D-penicillamine. Liver copper levels are between 1.5 and 25 times the upper limit of normal, but are also increased in some conditions with chronic cholestasis, such as primary biliary cirrhosis. Because of the large number of genetic mutations identified, a clinically applicable genetic test has not yet been developed.

PAS-positive granules in periportal hepatocytes. Heterozygotes with an MZ phenotype have about half the normal circulating levels of 1-antitrypsin, but none can be detected in ZZ homozygotes. It is these ZZ individuals who are susceptible to the development of emphysema (p. 661) and cirrhosis, but the mechanisms responsible are not understood. The liver disease usually presents in infancy, with symptoms of the neonatal hepatitis syndrome (conjugated hyperbilirubinaemia with pale stools, dark urine and hepatomegaly). At this stage cirrhosis may already be present on liver biopsy, although varying degrees of hepatocellular damage and fibrosis are more usual. In those who survive cirrhosis usually develops, and about a quarter will die in early childhood. The clinical course in the remainder may be relatively long, although decompensation can occur in adolescence or early adult life. There is no specific treatment. Some adults with 1 -antitrypsin deficiency can present with cirrhosis without a history of jaundice in infancy. In these patients primary hepatoma is a recognized complication.

Management

CYSTIC FIBROSIS

The chelating agent D-penicillamine gradually depletes the copper stores and is the mainstay of treatment. In cases picked up early, and in those with a relatively inactive wellcompensated liver lesion, there is usually some improvement in liver function and no further progression of the disease. Patients who present with acute liver failure or decompensated cirrhosis show little response to chelation therapy, and have a poor prognosis. Interruption of maintenance therapy for 6-12 months, usually by adolescents or young adults questioning the need to take regular medication, can lead to liver failure which does not respond to the resumption of D-penicillamine. ©

Cystic fibrosis is considered in detail in Chapter 13. There is a generalized disorder of exocrine secretory glands, whose ducts become blocked with viscid, tenacious secretions. The liver can be involved in this process, with mild focal biliary obstruction due to inspissated bile being found in almost all cases. For unknown reasons, in some this process slowly progresses to a relatively inactive cirrhosis. This is present in about 10-15% of adolescents with cystic fibrosis, and can present with the complications of portal hypertension.

SECONDARY BILIARY CIRRHOSIS OTHER CAUSES OF CIRRHOSIS 1-ANTITRYPSIN

DEFICIENCY

1 -Antitrypsin, a protease inhibitor, is remarkably polymorphic, with more than 30 different variants now recognized. The inheritance of these alleles is autosomal and codominant, with the normal phenotype being designated MM. The clinically important variant is called Z. This codes for a defective protein core which has altered carbohydrate side chains and cannot be excreted from hepatocytes. It can be seen in liver biopsies as diastase-resistant

1 870

Case 16.3

2

MCQ 16.11

Secondary biliary cirrhosis can be a complication of any disease in which there is prolonged biliary obstruction. For example, it may arise as a long-term complication of biliary stricture owing to operative damage to the bile duct, or in sclerosing cholangitis (p. 858). There is slowly progressive portal fibrosis, often for several years before the development of true cirrhosis. If adequate biliary drainage can be restored during the earlier stages there may be substantial regression of the portal fibrosis and a fall in portal pressure.

HEPATIC VENOUS CONGESTION 'Cardiac cirrhosis' occurs in patients with chronic right heart failure. It is a progressive liver fibrosis caused by chronic centrilobular congestion. An important cause is constrictive pericarditis. Treatment, by surgical removal of

the pericardium, usually leads to resolution of the ascites and regression of the liver fibrosis. A congenital web in the inferior vena cava at or above the level of the diaphragm is another rare but correctable cause. A more dramatic cause of hepatic venous congestion is sudden occlusion of the main hepatic veins: the Budd-Chiari syndrome. The presenting features, of pain and tenderness over a large liver and tense ascites, are due to the marked elevation in portal pressure, the severe hepatic congestion and liver cell necrosis. In some cases the onset is more gradual, with refractory ascites being the dominant clinical feature. The occlusion of the hepatic veins is occasionally due to obstruction by tumour tissue extending up the vena cava from a hypernephroma, but is usually thrombotic. This may be initiated by trauma, or be due to a thrombotic tendency in diseases such as polycythaemia rubra vera, primary thrombocythaemia, paroxysmal nocturnal haemoglobinuria and chronic myeloid leukaemia. A thrombophilia screen may reveal deficiency of anti-thrombin III, proteins C or S, Factor V Leiden, or the antiphospholipid syndrome as the cause of the procoagulant state. A number of cases have been reported in young women taking the oral contraceptive pill, but since its oestrogen content has been reduced this association seems to have become less frequent. The patency of the main hepatic veins may be investigated by ultrasound, but radiological contrast studies are more definitive. Axial imaging of the liver usually demonstrates sparing and hypertrophy of segment I of the liver, also known as the caudate lobe. The treatment options include creation of a shunt (TIPSS, mesocaval or mesoatrial shunt), hepatic venoplasty or liver transplantation. The choice is influenced by the balance of acute versus chronic disease and the degree of hepatic reserve. Liver transplantation is the only option in patients presenting with acute liver failure.

temic shunt operation is usually performed to reduce portal vein pressure. The prognosis is good. Some cases are associated with congenital dilatation of the biliary tree, and the combination is known as Caroli's syndrome.

16

COMPLICATIONS OF CIRRHOSIS Although there is no specific therapy for many types of cirrhosis, skilled management of the complications can often provide prolonged relief from disabling symptoms.

ASCITES Pathophysiology There are two main hypotheses which seek to explain the formation of ascitic fluid in cirrhosis (Fig. 16.26). The first, sometimes referred to as the 'underfill theory', suggests that the reduction in plasma colloid osmotic pressure, due to a fall in serum albumin concentration together with raised portal pressure, leads to a net extravasation of capillary filtrate from the splanchnic circulation. This in turn reduces the circulating plasma volume and promotes secondary hyperaldosteronism, with retention of sodium and water in the kidneys. This theory predicts a normal or reduced cardiac output, decreased plasma volume and peripheral vasoconstriction. The observations that cardiac output is usually increased, that plasma volume is high and that there is peripheral vasodilatation in most patients with cirrhosis and ascites, prompted the development of an alter-

CHILDHOOD METABOLIC DISEASES Type IV glycogen storage disease, galactosaemia, fructosaemia and tyrosinaemia are all rare causes of cirrhosis in infancy and childhood (see Ch. 19).

CONGENITAL HEPATIC FIBROSIS Although this inherited condition is not a cause of cirrhosis, the presentation, with the symptoms and signs of portal hypertension, places it firmly in this context with respect to differential diagnosis. The basic hepatic architecture is preserved, but portal tracts are linked by broad bands of fibrous tissue. Increased numbers of bile ductules are seen in the portal areas, and some may appear cystic. The presentation is usually in childhood, with hepatosplenomegaly and haematemesis from bleeding oesophageal varices. Liver function tends to be well preserved, and a portosys-

FIG. 16.26 The pathophysiology of ascites formation in cirrhosis In addition to direct effects of the cirrhotic process leading to ascites, such as increased pressure in the'portal vein and decreased albumin lowering the plasma osmotic pressure, there is also an aldosterone-driven increase in sodium retention.

871

native view, sometimes known as the 'overflow theory'. This suggests that the primary abnormality is increased renin and aldosterone levels producing inappropriate sodium and water retention. The plasma volume and cardiac output are increased. The formation of oedema is a secondary event which is encouraged by the low colloid osmotic pressure in plasma, and tends to be localized to the ascitic compartment because of the portal hypertension. Although this theory fits better with the clinical measurements, it does not explain why there should be a primary increase in renin and aldosterone levels in some patients with cirrhosis. Renal haemodynamics are certainly altered, with a tendency for redistribution of blood flow from cortical to juxtamedullary nephrons, a tendency which is opposed in many patients by an increase in renal prostaglandin synthesis. These changes in renal haemodynamics may be part of a more general disturbance in vascular tone. There is evidence in cirrhosis of widespread and inappropriate vasodilatation. Although the cause of this is not understood it does provide a reason for the increased renin and aldosterone levels, driven by the necessity to fill the overexpanded vascular compartment.

Clinical features 1 Fullness in the flanks, with dullness on percussion which shifts with a change of position, is the earliest sign of ascites and is often asymptomatic. Ultrasound can reveal small amounts of fluid. Later the abdomen becomes tense and uncomfortable. At this stage a fluid thrill is more useful than shifting dullness in confirming the presence of fluid. It can be difficult to palpate the liver and spleen, although ballotting through the fluid may reveal them. A pleural effusion, particularly on the right side, may be present, and peripheral oedema often develops. Spontaneous bacterial peritonitis is a serious complication, with a high mortality. It can be asymptomatic, although fever, abdominal pain and a general deterioration in wellbeing, with encephalopathy and renal failure, may be accompanying features.

Management A diagnostic paracentesis is first performed to exclude bacterial peritonitis. The fluid should be clear or strawcoloured and any cloudiness should lead to suspicions of infection. Bacterial culture may be negative, and the most frequent abnormal finding is a polymorph count of more than 500 per mm3. The causative organisms are usually coliforms or streptococci, including anaerobes, and broadspectrum antibiotics should be started as early as possible.

1

872

Fig. 16.10

SUMMARY 5 Management of ascites Treatment Spironolactone Other diuretics

Salt restriction

Fluid restriction Paracentesis

Indication Main treatment Not alone; give with spironolactone if necessary Varies depending on response to diuretics and serum sodium 1.5Lor less if hyponatraemic 3-10L amounts with i.v. albumin

Problems Gynaecomastia K+ depletion; renal failure; encephalopathy Unpalatable diet leads to weight loss; Compliance Renal failure Hypotension

Tuberculous peritonitis is rare but should be considered, particularly in a malnourished alcoholic and when the ascitic leukocytes are predominately mononuclear. Diuretics Diuretics are the mainstay of treatment, but care is necessary as overdiuresis precipitates renal failure, electrolyte abnormalities and encephalopathy, which must be avoided. The diuretic of choice is the aldosterone antagonist spironolactone in a starting dose of l00mg per day. In addition to being a logical choice to combat the high aldosterone levels, it is also potassium sparing. Total body potassium levels are usually low, and hypokalaemia not only increases the risk of a cardiac arrhythmia, but also exacerbates encephalopathy. The rate of fluid loss is best assessed by changes in daily weight. A figure of 0.5kg per day is ideal, as the maximum rate of transfer of fluid from the ascitic to the vascular compartment is only about 700 mL per day. If a satisfactory response is not forthcoming in a few days, the daily dose of spironolactone should be increased in 100 mg increments to a maximum of 400 mg per day. Combination therapy in conjunction with a loop diuretic, e.g. furosemide (frusemide) 40-160 mg daily, is often effective. Measurement of urinary sodium concentration can be helpful in assessing the effectiveness of diuretic therapy. Strict control of sodium intake is not always necessary, but does allow lower doses of spironolactone to be effective. Large-volume paracentesis is once again in vogue in the management of severe ascites. Protocols that administer colloid intravenously at rates of 20-25 % of the volume of ascites drained have dramatically improved the safety of this approach. TIPSS or liver transplantation should be considered in intractable cases. The role for peritoneovenous shunts is now very limited. Hyponatraemia can be a problem, and is usually due to impaired renal free water clearance. Restriction of fluid intake is an appropriate therapeutic measure; sodium infusion simply increases fluid retention.

VARICEAL HAEMORRHAGE

Distortion of the lobular architecture in cirrhosis almost always produces raised portal venous pressure. Collateral veins open up between the portal and systemic systems. The most important clinically are those that run beneath the oesophageal mucosa, as they frequently rupture, producing life-threatening haematemesis (variceal haemorrhage).

Clinical features Evidence of portal hypertension may be detected on examination. Splenomegaly is almost invariable. Dilated collateral veins may be seen in the abdominal wall, and may radiate outwards from the umbilicus. In rare cases these may be especially prominent, forming a caput medusa. The obliterated umbilical vein may recanalize, and flow through it sometimes produces a venous hum heard best in the epigastrium or around the umbilicus.

Management Immediate measures The first priority is resuscitation. Blood transfusion is usually necessary, often with large quantities. The next step is to establish the source of the bleeding. Peptic ulceration is common in cirrhosis, and acute erosive gastritis and oesophagitis are often present, particularly in alcoholics. Early endoscopy is therefore helpful in demonstrating the likely bleeding site. The endoscopic signs of a recent bleed from a varix are a venous spurt of blood, or an adherent blood clot. Four treatment options are available: 1. Endoscopic ablation therapy Variceal banding has now largely replaced injection sclerotherapy as the treatment of choice for variceal ablation. The varices are individually sucked into a cylinder attached to the endoscope and bands are triggered to ligate the base of each varix. Five or six bands can be applied with each passage of the scope. Injection sclerotherapy, e.g. with ethanolamine, may be effective if brisk bleeding precludes good visualization of the varices. Bleeding gastric varices are injected with fibrin. 2. Pharmacology A range of drugs that cause splanchnic vasoconstriction have been used to control variceal bleeding. Terlipressin and octreotide (a somatostatin analogue) are now the drugs of choice, replacing the combination of vasopressin and glyceryl trinitrate. 3. Tamponade This is an effective short-term measure while other management strategies are put in place. A tube is introduced, e.g. Sengstaken-Blakemore balloon, which can be inflated to compress the varices, and a gastric balloon which anchors the lower end in the fundus of the stomach, and which may assist in reducing variceal flow by compressing gastric varices. It is very effective, but skill is needed in its introduction

TABLE 16.13 Child's classification of severity of cirrhosis

16

Points scored for increasing abnormalities Feature

1

2

3

Encephalopathy (grade) None 1 and 2 3 and 4 Ascites None Mild Moderate/severe Plasma bilirubin ( mol/L) less than 25 25-40 more than 40 Plasma albumin (g/L) more than 35 28-35 less than 28 Prothrombin time (sees 1-4 4-6 more than 6 prolonged) Total score: 5-6 = grade A; 7-9 = grade B; 10-15 = grade C

and in maintaining it in a safe position. Potential complications include perforation of the oesophagus and ischaemic damage to the oesophageal mucosa. 4. Shunts These may be fashioned either percutaneously in the form of a TIPSS or surgically. The former is preferred if liver transplantation is an option in the future management of the individual patient. The shunt approach is generally less popular, as several controlled trials have shown no benefit in terms of mortality at any stage. Reduced deaths from rebleeding are offset by an increased incidence of liver failure and encephalopathy, presumably owing to the reduction in liver blood flow and the increased amount of blood bypassing the liver after the shunt procedure. The mortality with each bleeding episode is about 35%. An episode of variceal bleeding is considered resolved once bleeding has stopped and oesophageal varices have been obliterated. The risk of rebleeding depends on the underlying aetiology, Child's classification, and the secondary prophylactic measures adopted. Patients with alcoholic liver disease who abstain are considerably less likely to rebleed than those who continue to consume alcohol. Rebleeding is also more frequent in patients with Childs C rather Child's A disease (Table 16.13). The 'state of the art' management of secondary prophylaxis involves the measurement of the portal pressure before and after the introduction of drugs to reduce the pressure, e.g. with propranalol and/or isosorbide mononitrate. The dose is then titrated to achieve an adequate reduction in the absolute portal pressures.

EXTRAHEPATIC PORTAL VEIN OBSTRUCTION

Although relatively uncommon, this is a condition with a good prognosis which often presents with variceal haemorrhage, and must be distinguished from cirrhosis. The cause of the portal vein obstruction is often not apparent and in many cases is thought to be due to portal vein thrombosis in infancy, possibly secondary to umbilical sepsis. Patients first present with haematemesis from

873

SUMMARY 6 Management of haematemesis in cirrhosis 1. Resuscitate: blood, blood substitutes 2. Identify the source of bleeding by endoscopy (haemorrhagic gastritis and ulcers common) 3. If varices bleeding, either: • Treat endoscopically - banding or sclerotherapy • Sengstaken-Blakemore tube (difficult to pass; can be dangerous) • Terlipressin 4. Long-term: • Banding or sclerotherapy • Shunts - TIPSS, or occasionally surgical • Liver transplantation

bleeding oesophageal varices in childhood, adolescence or early adult life. Clinically there are no cutaneous stigmata of chronic liver disease, and although splenomegaly is often marked the liver is usually impalpable. Liver function tests are normal and ascites does not occur. Liver biopsy is helpful in excluding cirrhosis, and ultrasound can often determine the patency of the portal vein. Direct confirmation of the diagnosis is best obtained from the venous phase of selective coeliac axis and superior mesenteric angiograms. These usually show a leash of collateral vessels in the porta hepatis, where the portal vein would normally be seen. In adults portal vein thrombosis can be a complication of intra-abdominal sepsis or carcinoma of the pancreas. Treatment is aided by the excellent liver function. Portosystemic shunts can be technically difficult, although if successful the risk of encephalopathy is much less than in patients with cirrhosis because liver cell function is not impaired. Endoscopic ablation is a suitable alternative.

ENCEPHALOPATHY This complication of cirrhosis is a sign of severe liver cell failure and usually has a poor prognosis. However, a treatable precipitating factor is often present, and its identification and correction is an important clinical objective.

Pathophysiology A reduction in functioning liver cell mass and the diversion of portal blood past the liver via portosystemic collaterals are the two important ingredients that predispose to the disturbance in cerebral function, but the precise

1

874

MCQ 16.12

mechanisms involved are not known. There are three main theories: 1. Toxic substances normally detoxified by the liver are thought to gain access to the systemic circulation and thus the brain, where they alter cerebral metabolism. 2. An alteration in the balance between cerebral neurotransmitters is postulated, with the formation of false neurotransmitters, possibly because of altered amino acid levels in the circulation, which in turn alter the levels of these neurotransmitter substrates in the brain. 3. The inhibitory GAB A receptor is upregulated by the benzodiazepine receptor, which is closely aligned to it in the form of a macroreceptor. Possible candidates as toxins in the first theory include ammonia, mercaptans, amines and indoles. All can be shown to induce encephalopathy in experimental animals, but in concentrations higher than those found in hepatic encephalopathy. Increased serum levels of the aromatic amino acids, and decreased levels of branched-chain amino acids, have promoted the concepts outlined in the second theory. There have been attempts to treat the encephalopathy by correcting the balance between the aromatic and branched-chain amino acids, but the place of this therapy in clinical practice is not established. Recently changes in the relative proportions of highand low-affinity receptors for some of the cerebral neurotransmitters have been demonstrated in animal models of hepatic encephalopathy, and this seems to be a fruitful area for further research. A functional increase in GABAergic tone has been demonstrated in experimental hepatic encephalopathy, and increased levels of benzodiazepine agonists, including diazepam, have been found in the body fluids and brain of patients with hepatic encephalopathy. Experimental use of the benzodiazepine antagonist flumazenil in individual patients with acute or chronic hepatic encephalopathy appears to be able to induce a transient and incomplete clinical and EEG improvement in about two-thirds of cases.

Clinical features The main difference between the encephalopathy of chronic liver disease and that of acute liver failure is the relapsing acute-on-chronic course and the consequent emphasis on the search for, and correction of, precipitating factors. The prodromal phase of an acute-on-chronic episode may be very gradual. Impairment of judgment can be an early sign. Anorexia is common and the reversal of sleep pattern may be troublesome. Fatigue, withdrawal, apathy and slowness of response often precede a more profound disturbance in consciousness (see Table 16.7). There may be spatial disorientation, with confusion as to place, time and person, and a flapping tremor of the outstretched, extended hands. A distinctive sweet smell of the breath, called fetor hepaticus, is detected with moderate to severe encephalopathy in patients with chronic liver disease.

Although not specific to liver disease, changes in the EEG can be helpful in identification and management. There is generalized slowing of the main components, with a resulting fall in the mean dominant frequency. Serial recordings are one objective way of following progress during treatment, although simple bedside tests are sufficiently reliable and sensitive in practice. A standardized number connection trail is a simple and semiquantitative method, and is now preferred to the classic assessment of constructional apraxia which involved copying a fivepointed star.

Management An important objective is to search carefully for a possible precipitating cause (Table 16.14). Intercurrent infections are the most common, and probably interfere with liver function in two ways: by exposing the liver to endotoxin and by reducing liver blood flow secondary to hypotensive effects in the systemic circuit. Spontaneous bacterial peritonitis is a common causative infection that is difficult to diagnose unless routine diagnostic paracentesis is performed. Empirical antibiotics are advised if the cause of the encephalopathy is uncertain. Increased absorption of nitrogen from the gastrointestinal tract occurs with excessive dietary protein, gastrointestinal bleeding and constipation. Diuretics may precipitate encephalopathy by causing dehydration or electrolyte imbalance. The other common precipitant of

encephalopathy is medication, especially narcotic drugs and benzodiazepines. Treatment of the encephalopathy itself is directed mainly at removing sources of ammonia. Ammonia production in the liver by deamination of amino acids can continue even though its further metabolic transformation to urea is impaired, and lowering the protein levels in the diet can sometimes be effective in improving the encephalopathy. However, care must be taken not to increase the catabolic state and thus exacerbate the long-term malnutrition which is often a feature of chronic liver disease. Ammonia is produced in the bowel by bacteria, and to reduce this the colon is emptied using a magnesium sulphate enema, and ammonia-producing colonic bacteria are reduced in number by giving oral lactulose. Lactulose is a synthetic disaccharide which is broken down into its constituent monosaccharides by bacteria in the colon, and then into lactic acid. This not only acts as a purgative, but also discourages the growth of ammonia-producing bacteria. Because of the change in pH, ammonia may also be fixed in the colon as ammonium ions. 1

16

HEPATOCELLULAR CARCINOMA Often referred to as hepatoma, hepatocellular carcinoma arises from the hepatic parenchymal cells.

Incidence, epidemiology and aetiology TABLE 16.14 Factors precipitating encephalopathy in cirrhosis Intercurrent infections Urinary tract Chest Upper respiratory tract Genitourinary tract Spontaneous bacterial peritonitis Septicaemia Dietary indiscretion High-protein meal Alcohol binge

In Africa and the Far East hepatocellular carcinoma is one of the commonest tumours, but is unusual in Europe and the USA (Fig. 16.27). Hepatocellular carcinoma usually arises in the cirrhotic liver, and in this setting is more common in males than in females. In Europe and North America the peak incidence is from 50 to 70 years of age, but in high-incidence areas patients are often in their 20s or 30s. When there is no

Gastrointestinal haemorrhage Blood acts as protein load, and hypotension impairs liver function (may be occult; melaena on rectal examination) Electrolyte imbalance Hyponatraemia or hypokalaemia (possibly due to vomiting, reduced free water clearance or effect of diuretics) Drugs Usual candidates are sedatives (increased end-organ sensitivity and reduced clearance) and diuretics (electrolyte imbalance) Primary hepatocellular carcinoma Worsening of encephalopathy may be the first indication

FIG. 16.27 The distribution of hepatocellular carcinoma in the world It almost parallels the distribution of HBV infection (Fig. 16.15).

875

underlying cirrhosis the sex ratio is equal, and the tumour usually presents in teenagers and young adults if of the fibrolamellar variety, and later if of standard morphology. Chronic viral hepatitis is closely associated with the development of hepatocellular carcinoma. In the west this is seen most commonly with hepatitis C, and this would appear to be the basis for the increasing incidence there of hepatocellular carcinoma. Chronic HBV infection is also strongly associated with the development of the tumour, particularly in areas with a high HBsAg carrier rate. The finding of subgenomic fragments of HBV DNA in liver cell and tumour DNA, even in some patients without serological markers of previous HBV infection, has led to the suggestion that the virus may be directly oncogenic. However, the pattern of integration of viral DNA appears to be random, and molecular mechanisms that could be responsible for oncogenesis have not yet been identified. Careful prospective studies of hepatoma development in patients with cirrhosis in the UK have shown that previous HBV infection, as detected serologically, is not a significant independent risk factor. Aflatoxin, a product of the fungal mould Aspergillus flavus, can induce liver damage and hepatocellular carcinoma in animals, and is a food contaminant in many tropical countries. Very rarely, hepatocellular carcinoma has been described following prolonged use of the oral contraceptive pill, and in individuals exposed to Thorotrast, an early radioactive X-ray contrast material.

Pathology The commonest microscopic pattern of hepatocellular carcinoma is the characteristic trabecular or sinusoidal pattern, which reproduces the normal relationship between clusters of liver cells and a flat endothelial cell covering (see Fig. 16.3). Fibrolamellar carcinoma is a distinct variant with a better prognosis. The tumour cells are large and polygonal and are arranged in trabeculae separated by parallel bundles of collagen which merge in places to form broader bands of fibrous tissue.

Clinical features In the UK there is usually a slow deterioration in wellbeing, with anorexia, weight loss and abdominal pain in a patient known to have cirrhosis. There may be no specific signs on examination, but hepatomegaly is usually present and there may be an arterial bruit and a friction rub over the tumour. In high-incidence areas the presentation is often more dramatic, with pain over the tumour, fever and weight loss. The deterioration in liver function is more profound in the patient with pre-existing cirrhosis. 1

876

Case 16.4

0 Fig. 16.11

Diagnosis 1 The most important biochemical marker of the tumour is oc-fetoprotein (AFP). This is a glycoprotein which can be thought of as a fetal albumin. It is synthesized by the fetal liver, but levels fall rapidly after birth. In adults it is present in serum in a concentration of less than lOng/mL. About 80% of patients with hepatocellular carcinoma arising in a cirrhotic liver, and 50% of those without an underlying cirrhosis, will show elevated AFP levels in serum. In a cirrhotic, patient levels above 500ng/mL are diagnostic of hepatocellular carcinoma. AFP is also greatly elevated in patients with germ cell tumours of the testis and ovary, but this is not a diagnostic problem. Some patients with hepatic metastases can also have elevated AFP levels, but not as high as those found in hepatoma. In viral cirrhosis levels of AFP can rise above the normal adult range, probably reflecting increased liver cell regeneration, but a rapid sustained rise in the serum levels on serial estimations is typical of hepatoma. Ultrasound, CT and MRI are all capable of visualizing hepatocellular carcinoma. Selective hepatic arteriography is a very useful diagnostic technique, and also provides information on the distribution of the tumour within the liver. There is usually a well-defined tumour circulation, with new vessel formation. Although the definitive diagnosis can only be made by histological examination of tumour tissue, in practice liver biopsy may not always be possible because of a prolonged prothrombin time. In these cases a working diagnosis is established from a combination of elevated AFP levels in serum and the demonstration of a tumour circulation on arteriography.

Management Hepatocellular carcinoma is usually a rapidly growing tumour which presents late. The prognosis is poor, with a mean survival of about 4 months. Resection is potentially curative but is only possible in those cases where the tumour is limited to one lobe of the liver. Most patients undergoing resection have normal unaffected liver tissue, but some progress is being made resecting tumours from patients with well compensated cirrhosis. The long-term outlook following successful resection is good, with local recurrence being much more common than distant metastases. On this basis it might be expected that liver transplantation would be another effective treatment, also applicable to those with cirrhosis. However, both tumour recurrence in the liver graft and distant metastases are commonly observed within 1-2 years of the procedure. It is not clear whether this is due to immunosuppressive therapy, or whether the tumours are often more advanced than initially suspected. Consequently, liver transplantation is reserved for patients with one or more tumours that do not exceed 4cm in diameter, and in this setting the outcome is excellent. The only cytotoxic drug to show any benefit so far is adriamycin (doxorubicin). About one-third of those

treated show a response, but in most of these survival is only prolonged by a few months. TACE (transcutaneous arterial chemoembolization) is effective in a minority of patients, and the mechanism of action is probably a combination of the effect of the cytotoxic drug and tumour ischaemia induced by the occlusion of small arterial vessels by the lipiodol. Direct injection of alcohol into the tumour may also temporize the progression of the disease.

LIVER METASTASES Many tumours metastasize to the liver, particularly those arising in the abdomen; in western societies this is the commonest cause of malignancy within the liver.

Clinical features Although the primary tumour may already have been identified, in some cases the metastatic tumour may be the cause of the presentation. In these cases, malaise, weight loss, pain over the liver and abdominal distension are the likely symptoms. Examination often confirms substantial weight loss and usually demonstrates a hard, enlarged liver with an irregular surface and edge.

Investigations The serum alkaline phosphatase is often elevated. Jaundice is usually a late occurrence and has a poor prognosis. Ultrasound and CT frequently confirm the presence of lesions in the liver, © and the definitive diagnosis can then be made on liver biopsy. Tumour tissue can be expected on liver biopsy in more than 70% of cases, and the success rate can be further enhanced by using ultrasound to indicate where the biopsy should be taken.

Differential diagnosis Because of the importance of the diagnosis of metastatic liver cancer, it is usually necessary to obtain histological proof of malignancy. The conditions most likely to be misdiagnosed are cysts, abscesses and granulomas. In areas where hepatoma is common, the pain, fever and rapidly enlarging liver make amoebic abscess and hydatid disease important treatable conditions to be considered in the differential diagnosis.

Management Treatment depends on the nature of the primary tumour. For certain tumours, such as small cell carcinoma of the lung, breast cancer and lymphoma, chemotherapy may be highly effective in producing regression of metastases

and symptomatic improvement. With other tumours, such as metastatic gastrointestinal and pancreatic cancer, chemotherapy is usually ineffective, although infusion of cytotoxic drugs such as 5-fluorouracil into the hepatic artery occasionally relieves pain and produces regression. It is questionable whether this complex procedure is more effective than intravenous administration. Pain in the liver is usually controlled by analgesics, but severe pain may be helped by external irradiation and even hepatic artery ligation or tumour embolization. Rarely, resection of an isolated metastasis may be feasible; colorectal metastases and carcinoid tumour (p. 796) are examples where this approach may be valuable. Symptoms of pain, nausea and anorexia are best treated by palliative measures, including analgesics, antiemetics such as metaclopramide, and corticosteroids.

16

LYMPHOMA The liver is often directly involved with infiltration of lymphoma cells. Liver biopsy can therefore be helpful in diagnosis. Documenting liver involvement is also important in staging the disease before deciding on appropriate therapy. Serum alkaline phosphatase levels may be increased and jaundice can occur. In Hodgkin's disease a cholestatic jaundice may appear before there is obvious lymph node involvement, and occasionally in the absence of any evidence of direct hepatic involvement.

THE LIVER IN SYSTEMIC DISEASE RHEUMATOID ARTHRITIS Increased serum alkaline phosphatase levels from the liver are frequently seen in rheumatoid arthritis. Histological changes on liver biopsy are often non-specific and unremarkable. Amyloidosis and primary biliary cirrhosis are well defined but infrequent associations. Nodular regenerative hyperplasia is sometimes found in Felty's syndrome (p. 1137).

SYSTEMIC LUPUS ERYTHEMATOSUS The early finding of lupus erythematosus (LE) cells in about 10% of cases of autoimmune chronic hepatitis, and the use of the term 'lupoid hepatitis' to describe these patients, created considerable confusion concerning the relationship between chronic active hepatitis and systemic lupus erythematosus (SLE). Although anti-DNA antibodies are frequently present in autoimmune chronic hepatitis, it is now clear that the two conditions

877

are quite distinct. In particular, clinically significant renal involvement, an important feature in SLE, is very rare in chronic autoimmune hepatitis. Furthermore, careful studies of liver involvement in SLE have in most cases shown only minor portal tract inflammation.

POLYARTERITIS NODOSA Chronic HBV infection has been reported in up to 40% of cases of polyarteritis nodosa, but this is not a consistent finding. The associated liver disease in these cases is usually very mild, with minor histological changes on liver biopsy. The appearance of characteristic arterial lesions in biopsy material can be of help in diagnosis.

SCLERODERMA There is a well-recognized association between scleroderma and primary biliary cirrhosis (PBC). In some cases of PBC the associated sclerosis is truly systemic, but in most it is limited to the skin of the terminal phalanges (sclerodactyly), and may be part of the CRST syndrome.

AMYLOIDOSIS The liver is often involved in systemic amyloidosis (p. 1263) but this rarely produces clinically significant liver disease. Rectal biopsy is safer than liver biopsy in establishing the diagnosis.

PORPHYRIA Porphyria cutanea tarda (Ch. 19) may be associated with alcoholic liver disease, and sometimes hepatoma; the accumulation of protoporphyrin in the liver in erythropoietic protoporphyria may eventually lead to cirrhosis and liver failure.

GRANULOMATOUS CONDITIONS The finding of granulomata in liver biopsy specimens is relatively common. Although they are a feature of some primary liver diseases, particularly primary biliary cirrhosis, they are more often a reflection of a systemic condition (Table 16.15). In the UK the most common causes are sar-

1 MCQ 16.13 878

TABLE 16.15 Causes of granulomatous liver disease Infections Bacterial Usually chronic intraceliular

and

infections Mycobacterial including tuberculosis* Brucellosis* Leprosy Listerosis Fungal Histoplasmosis Nocardia Candidiasis Actinomycosis Parasitic Schistosomiasis* Ascariasis Toxocara Amoebiasis Strongyloides Giardiasis Viral Cytomegalovirus Lymphogranuloma venereum Infectious mononucleosis Other Syphilis Q fever

Drug-induced Sulphonamides and derivatives Sulphonamide-containing antibiotics* antidiabetics such as chlorpropamide* Others e.g. Phenylbutazone* Allopurinol Hydralazine Carbamazepine Quinidine Multisystem diseases Collagen-vascular disease Polymyalgia rheumatica* Systemic lupus erythematosus Others Sarcoidosis* Whipple's disease Wegener's granulomatosis Granulomatous disease of childhood Miscellaneous Primary biliary cirrhosis* Crohn's disease* i.v. drug abusers Patients on haemodialysis Haemophilia After ileal bypass surgery Erythema nodosum

Neoplasms Hodgkin's lymphoma* Adenocarcinoma Hepatocellular carcinoma * Common causes of liver granulomata.

coidosis, tuberculosis and drug reactions. There is usually hepatomegaly, with a raised serum alkaline phosphatase but a normal serum bilirubin level. In up to 30% of cases the cause of the hepatic granulomata cannot be determined, in spite of rigorous investigation. Occasionally portal hypertension may develop, either because of a direct effect of the granulomata on liver blood flow or in association with increasing liver fibrosis. In these cases a trial of steroids is justified to try to halt the progress of the disease. Granulomatous hepatitis denotes the presence of focal parenchymal liver damage in association with multiple granulomata. In some cases of unknown aetiology there may be quite striking symptoms of a systemic illness,

including fever, malaise and weight loss. There is often a dramatic improvement on corticosteroid therapy. 1

NON-VIRAL INFECTIONS INVOLVING THE LIVER SCHISTOSOMIASIS Worldwide, schistosomiasis (p. 370) is a common cause of chronic liver disease. Adult worms live in small veins of the colon (Schistosoma mansoni) and the small intestine and colon (S. japonicum}. The liver damage is due to granulomatous reactions to eggs which have embolized to the liver and impacted in presinusoidal branches of the portal vein. Resolution of granulomata leads to severe periportal fibrosis in patients with heavy infections. Portal hypertension usually accompanies the developing fibrosis, but liver cell function is well preserved unless there is coexisting cirrhosis from another cause. Patients often present with haematemesis from ruptured oesophageal varices; gross splenomegaly is common and ascites occurs. The prognosis is fairly good. Diagnosis and management are discussed in Chapter 9.

LIVER FLUKES Fasciola hepatica is a flatworm which is common in sheep and, to a lesser extent, cattle worldwide (p. 375). It lives in the biliary tree and sheds its ova into the stools. The liver damage is caused by worms in the biliary system. In severe infections there is intermittent obstructive jaundice, with high fever, painful hepatomegaly and jaundice. Eosinophilia is common, and a complement fixation test is usually positive in high litre. The definitive diagnosis is made by finding ova in stool concentrates. The infection responds to treatment with praziquantel. Clonorchis sinensis is found in the Far East (p. 373). The worms inhabit the biliary system. Symptoms are usually mild, except in heavy infections, when there may be upper abdominal pain, tender hepatomegaly and diarrhoea. Praziquantel is an effective treatment. Late complications include obstructive jaundice, cholangitis, stones in the common bile duct and cholangiocarcinoma (pp. 858-859).

HYDATID CYSTS Hydatid cysts (p. 377) are formed in the liver by the larval stages of two tapeworms, Echinococcus granulosus and E. multilocularis, which normally live in dogs and wild carnivores, respectively. They may be asymptomatic or may present with pain in the right hypochondrium. If they are

very large they can cause abdominal distension and discomfort. Secondary infection in a cyst is a serious complication, presenting with fever and pain like other pyogenic liver abscesses. Ultrasound and CT are the most useful investigations and can be diagnostic if 'daughter' cysts are identified. Complement fixation and haemagglutination tests for antibodies to hydatid material can also be helpful in making a specific diagnosis, although negative results can occur, particularly in asymptomatic cases. Mebendazole and, more recently, albendazole have been used for treatment and have become the first line of therapy. Surgical removal of the cysts still offers the best chance of cure, but great care must be taken to prevent rupturing of the cysts, which leads to the dissemination of daughter cysts into the peritoneal cavity. This risk is minimized by injecting formalin, alcohol or hypertonic saline into the cyst before attempting removal. It is wise to leave asymptomatic cysts alone.

16

AMOEBIASIS In some cases of amoebic colitis, and in others without bowel symptoms, the organisms reach the liver through the portal venous system and form intrahepatic amoebic abscesses. Pain in the right hypochondrium, with tender hepatomegaly and fever, are the usual manifestations. A right-sided pleural effusion can also occur if an abscess has ruptured through the diaphragm. Ultrasound and CT scans are very useful investigations, and antibody tests are almost always positive. A week's course of metronidazole is a very effective treatment.

WEIL'S DISEASE The spirochaete Leptospira icterohaemorrhagica can be transmitted to humans from infected rats' urine. Those in contact with sewage and rat-infested water are at highest risk. The incubation period is 1-2 weeks. At first the illness may resemble the prodromal stage of viral hepatitis, with mild fever and malaise, but it is often more severe, with myalgia and high fever. This may progress to prostration, jaundice and renal failure. Headache, confusion and meningism are often present. Epistaxis, haematuria and petechiae occur. There is often a prominent leukocytosis, and the urine contains red cells and albumin. Identification of the causative organism in blood or urine is the definitive diagnostic test, although there is a rise in serum IgM antibody at the end of the acute phase of the disease. Although the jaundice may be deep, liver failure does not occur, and general toxaemia and renal failure are the usual problems in management. The organism is sensitive to benzyl penicillin and large doses are usually given. However, there is no convincing evidence that this contributes to the resolution of the disease unless adminis-

879

tered early, and skilled intensive care is essential. The reported mortality varies from 4 to 40%.

SEPTICAEMIA Cholestasis is not uncommon in severe sepsis. A rise in serum bilirubin occurs in 25-50% of cases, but serum transaminases and alkaline phosphatase are often normal. The pathogenesis is not understood, but there is no evidence of gross liver damage and there is usually rapid resolution with appropriate antibiotic therapy. In jaundice due to staphylococcal sepsis, treatment with fusidic acid should be avoided as this interferes with bilirubin excretion.

PYOGENIC LIVER ABSCESS O

Biliary tract disease Ascending cholangitis is the commonest cause and may be a complication of any obstruction in the biliary tract; it can also follow cholecystitis Portal phlebitis Any primary septic focus within the abdomen (diverticulitis, appendicitis, infected haemorrhoids, etc.) can cause thrombophlebitis in the local branches of the portal vein, and infect the liver Septicaemia Severe generalized sepsis can infect the liver via the hepatic artery Trauma Secondary infection of a damaged area of liver or a haematoma can lead to abscess formation Malignancy Some tumours develop areas of central necrosis, and pyogenic infection can then follow Extension of local infection e.g. subphrenic abscess and empyema of the gallbladder

The commonest cause of liver abscess worldwide is amoebiasis, but in the developed world pyogenic causes are of increasing importance. In the past pyogenic liver abscess was most commonly seen in children or young adults in association with appendicitis. However, as antibiotics have become more widely prescribed the condition is now most often seen in the elderly, owing to incomplete treatment of biliary tract disease or diverticulitis. In these patients the presenting symptoms are often non-specific and the diagnosis delayed.

Classically there is upper abdominal pain, swinging pyrexia, anorexia, malaise and weight loss. However, in many older patients the symptoms and signs may be much less dramatic, and pain and hepatomegaly may not be present. This is one of the causes of pyrexia of unknown origin.

Aetiology

Investigations

Possible causes of pyogenic liver abscess are listed in Table 16.16. Infection in any site drained by the portal vein can cause portal pyaemia, which can then seed the liver with abscesses. These are usually multiple and often concentrated in the right lobe. The primary site, particularly in diverticular disease, may be relatively silent. Direct spread of infection from the biliary tree is the presumed route of entry of the bacteria in those cases where abscess complicates biliary tract disease. Trauma and hepatic malignancy are uncommon causes.

A polymorphonuclear leukocytosis is usually present and the routine liver function tests often show a high serum alkaline phosphatase. The most useful investigation is an ultrasound scan. If there has been no previous antibiotic treatment, aerobic and anaerobic cultures of blood or direct aspirates of pus under ultrasound control will usually reveal the causative organism.

Bacteriology The dominant organisms isolated from hepatic abscesses have changed over recent years. Whereas aerobic Gramnegative organisms, especially E. coli, used to predominate, anaerobes, particularly Streptococcus milleri, are now found in up to 50% of abscesses. Mixed infections are identified in about one-third of cases.

1

880

TABLE 16,16 Causes of pyogenic liver abscess

MCQ 16.14

Presentation

Management Antibiotics are always needed and are best chosen according to the bacteriological findings. However, a combination of gentamicin (or a broad-spectrum cephalosporin), metronidazole and flucloxacillin will cover most of the likely organisms. In cases with very small multiple abscesses appropriate antibiotic treatment alone may be adequate, but in most cases drainage is also required. In a solitary abscess, or where the number is small, it may be possible to achieve satisfactory percutaneous drainage by repeated needle aspiration. In elderly patients the first line of treatment is usually closed aspiration combined with antibiotics instilled into the abscess cavity and given intravenously. When the abscesses are difficult to approach, open surgical drainage is used. Soft pliable drains are placed in the largest cavities and intensive broad-spectrum

antibiotic treatment is started. Whichever approach to drainage is used, the antibiotics need to be continued for 6 weeks to ensure resolution of the infection.

Prognosis Unfortunately, the diagnosis of pyogenic liver abscess is often delayed, and under these circumstances the mortality may be as high as 80%. With a high index of suspicion and early diagnosis the mortality has been as low as 10% in some series. Coexistent biliary tract disease, old age and multiple abscesses are associated with a poor prognosis.

CONGENITAL BILIARY AND HEPATIC DISORDERS CAROLI'S DISEASE Caroli's disease is a rare condition in which saccular dilations of the intrahepatic biliary tree predispose to ascending cholangitis. Symptoms usually begin in childhood or early adult life. Long-term antibiotic treatment reduces the frequency of attacks.

Large cysts, similar to those seen in polycystic kidney disease and often associated with them, may be found throughout the liver. Gross nodular hepatomegaly, often asymptomatic, is usual. Although pain, due to haemorrhage or infection in a cyst, can occur, the prognosis with respect to the liver lesions is excellent.

16

LIVER TRANSPLANTATION Liver transplantation is now an accepted and integrated component of the management of most of the conditions discussed in this chapter. However, the majority of patients with the common liver diseases, particularly alcoholic liver disease, are not transplanted, and most have rare conditions such as primary biliary cirrhosis. Guidelines have been developed by the British Society of Gastroenterology for the referral of adult patients for consideration for liver transplantation, and the salient points are outlined in Table 16.17. In chronic liver disease patients are considered for transplantation if there are major clinical complications that are not responding to standard therapy, available prognostic factors indicate that survival beyond 12-18

TABLE 16.17 Indications for liver transplantation in adults* CHOLEDOCHAL CYST Choledochal cyst is a gross dilatation or diverticulum of the common bile duct which obstructs normal bile flow and presents as cholestatic jaundice in childhood. The diagnosis is made on ultrasound and cholangiography. Treatment is by surgical resection and drainage, usually via a Rouxen-Y anastomosis.

Category Chronic liver disease Primary biliary cirrhosis Primary sclerosing cholangitis Autoimmune hepatitis Alcoholic cirrhosis Hepatitis C

BILIARY ATRESIA Originally thought to be a congenital anomaly, with failure of development of the extrahepatic biliary system, biliary atresia is now recognized as an acquired condition in many cases. There is gradual destruction of the common bile duct soon after birth, leading to deepening jaundice and secondary biliary cirrhosis. Early surgical intervention, using the Kasai technique to fashion a bypass anastamosis high in the porta hepatis, has proved remarkably successful in some cases. Liver transplantation has also produced promising results when used as primary therapy and in salvaging patients at various times after Kasai operations.

POLYCYSTIC DISEASE OF THE LIVER Polycystic disease of the liver may be a congenital anomaly related to biliary atresia, but rarely presents with jaundice.

Hepatitis B

Acute liver failure Paracetamol Other aetiologies Malignancy Hepatocellular carcinoma Neuroendocrine tumours Epithelioid haemangioendothelioma

Comment Bilirubin > 100 mol /L, liver failure Consider early to pre-empt cholangiocarcinoma Liver failure without reversible component with immunosuppression Liver failure after at least 6 months' abstinence Liver failure or hepatocellular carcinoma (see below) Liver failure if HBV DNA negative (spontaneously or with antiviral therapy), associated hepatitis D is favourable Medical criteria indicating a poor prognosis combined with favourable psychosocial assessment Medical criteria indicating a poor prognosis under the age of 60 years 1-2 nodules <5cm in diameter on radiological assessment Consider selected cases for palliation Must differentiate from angiosarcoma

* Modified from British Society of Gastroenterology Guidelines.

881

FIG. 16.28 MR cholangiography showing a tight stricture at the site of the biliary anastomosis in a liver transplant recipient

months is unlikely, or if quality of life is severely impaired by non-life threatening symptoms, e.g. fatigue, pruritus. Emergency liver transplantation is an essential component of the management plans for acute liver failure, and these patients should always be managed in specialist units with access to transplantation. Small hepatocellular carcinomas detected through routine surveillance in patients with cirrhosis account for the vast majority of cases undergoing liver transplantation for malignant disease. Liver transplantation continues to have a small role in the management of metabolic disease and normal liver morphology, e.g. Crigler-Najjar syndrome, hyperoxaluria, familial amyloid. There are remarkably few absolute contraindications to liver transplantation. Extrahepatic malignancy or infection and AIDS are the most obvious. Patients have been successfully transplanted as young as 5 days and over the age of 70 years. Each patient undergoes an assessment protocol to determine the risk/benefit profile, technical difficulties and comorbidity that may alter the outcome after transplantation. Most operations use whole organs from cadaveric donors. However, the liver may be reduced in size to match the recipient, may be split between two recipients, or only part of the liver be transplanted into the orthotopic position beside the native liver, having the potential to recover;

1 MCQ 16.15

882

this is known as auxiliary transplantation. The latter is used in the treatment of selected patients with acute liver failure and some metabolic disorders. Living related donation of either the right or the left lobe has now been established as an option for adult and paediatric recipients, respectively. The postoperative problems include technical complications, rejection, infections, complications of immunosuppressive therapy and disease recurrence. Arterial thrombosis and biliary complications, either leaks or strictures, are the most frequent technical problems. Acute rejection is usually easy to manage with increases in immunosuppression, and chronic rejection of the liver requiring retransplantation has now become rare. A very wide range of infections can be observed, including bacterial, fungal, viral and opportunistic. The need for long-term immunosuppression results in complications from the immunosuppression per se (e.g. lymphoma, increased risk of some other malignancies, opportunistic infections), as well as the side-effects of the constitutive drugs (e.g. nephrotoxicity, hypertension). The range of immunosuppressive drugs is now expanding (tacrolimus, ciclosporin, mycophenolate, sirolimus), giving more scope to customize immunosuppression regimens and minimize the attendant toxicity. The possibility of disease recurrence exists for the majority of transplantation indications, and ranges from being a significant problem (viral hepatitis, alcohol, malignancy) to one of academic interest (primary biliary cirrhosis). The results of liver transplantation are now generally excellent. Patients with primary biliary cirrhosis can expect a 90-95% 1-year survival rate, followed by a near-normal life-expectancy. The overall results for elective transplantation are 80-85% 1-year and 70-75% 5-year survival rates; individual programmes produce even higher rates. The equivalent figures for emergency transplantation are 70-75% and 55-60% 1- and 5-year survival rates, respectively. 1

FURTHER READING ON LIVER TRANSPLANTATION Chenard-Neu M P et al 1996 Auxiliary liver transplantation: regeneration of the native liver and outcome in 30 patients with fulminant hepatic failure - a multicenter European Study. Hepatology 23(5):1119-1127. Devlin J, O'Grady J G 1999 Indications for referral and assessment in adult liver transplantation: a clinical guideline. Gut 45 (Suppl VI); VI1-VI22. Devlin J, Wendon J, Heaton N, Tan K C, Williams R 1995 Pretransplantation clinical status and outcome of emergency transplantation for acute liver failure. Hepatology 21(4): 1018-1024.

17

Endocrine Diseases Anthony P Weetman

Historical introduction 883

The posterior pituitary gland 910

The physiology of hormone synthesis and action 883

The thyroid gland 912

Classification of endocrine diseases 888 The aetiology of endocrine diseases 888 Systemic effects of endocrine disorders 890 Endocrine function testing 891 The pituitary gland 892

The adrenal glands 925 Diseases of the adrenal cortex 929 Diseases of the adrenal medulla 938 The testes 939 The ovaries 945 Disorders of sex differentiation 949

The anterior pituitary gland 893

HISTORICAL INTRODUCTION Endocrinology emerged as a medical science in the 19th century. Berthold of Gottingen, in 1849, was the first to provide 'experimental proof for the existence of an internal secretion, by demonstrating that transplantation of the testis to another part of the bird's body prevented atrophy of the cock's comb, known to be caused by castration. In 1855, Thomas Addison described the clinical picture that resulted from destruction of the adrenal glands in humans. The first clear demonstration that an extract of a gland could correct a deficiency disease was provided by Murray and others, who showed that the administration of sheep thyroid extract would cure myxoedema and its childhood equivalent, cretinism, and the effects of thyroidectomy. In 1915 Kendall isolated the active principle and named it thyroxine. By then, the first demonstration of a hormone had been made, in 1902 by Baylis and Starling, who extracted the hormone they termed secretin from the duodenum, and found that it promoted pancreatic enzyme secretion. Adrenaline (epinephrine) was extracted from the adrenal glands at about the same time. By the mid19308, German chemists had isolated and later synthesized

testosterone and the main oestrogen, oestradiol, followed in the 1940s by the adrenal glucocorticoids, cortisone and cortisol. Because of its unusual chemical structure and high potency, the main mineralocorticoid, aldosterone, was not discovered until the mid-1950s. The major pathways of steroid synthesis were not completely identified, nor the inborn errors of steroid production recognized, until this time. From an early stage, clinical observation has had a major impact on the development of the science of endocrinology and the recognition of different hormones. This is particularly true of the pituitary gland and its subservient endocrine glands. Early this century, Harvey Gushing, the father of pituitary surgery, recognized that some features of pituitary disease were due to deficiencies, and others (for example in acromegaly) to hormone excess (in this case, growth hormone). Geoffrey Harris, working in Cambridge, was responsible for the concept that the anterior pituitary gland was controlled by secretions coming to it through a portal blood supply from the hypothalamus. Since the late 1960s we have seen the isolation and synthesis of several of these releasing hormones and their entry into clinical practice. Recently, the pace of development of our understanding of basic and clinical endocrinology has been truly staggering. Much has come from advances in methods for measuring the minute concentrations of hormones in body fluids. In many instances the concentration of active hormones in the circulation is in the range of picograms (10-12g) per mL. In the 1960s, initial bioassay and chemical methods gave way to protein-binding assays, of which the most widely used have been radioimmunoassays. These depend upon developing highly specific antibodies to hormones, coupled if necessary (e.g. for steroids) to a protein. A further essential development was the capacity to make radiolabelled hormones. Such methods have demonstrated, among other phenomena, the episodic nature of the release of certain hormones. Thus, administration of gonadotrophin-releasing hormone (GnRH) (initially to monkeys) showed the importance of pulsatile secretion of GnRH in transmitting the correct message. Continuous administration was found to switch of secretion of the hormones that it stimulates when given episodically. Recently, very sensitive bioassay methods have been developed, such as those using isolated cells; these are especially important for polypeptide hormones. Furthermore, immunoradiometric assays for many polypeptide hormones have now become available, with even greater sensitivity and specificity than radioimmunoassays.

THE PHYSIOLOGY OF HORMONE SYNTHESIS AND ACTION Hormones act in a classic endocrine fashion by exerting effects on a target organ at a distance from the endocrine

883

cell by secretion into the bloodstream. Some hormones have additional biological actions. They may have paracrine effects on adjacent cells, usually within the endocrine organ itself, or may act as neurotransmitters. Most hormones are polypeptides, and general features of their synthesis and secretion are detailed below; the synthesis of thyroid hormones, steroid hormones and catecholamines depends on multiple enzymes and is described in the appropriate sections of this chapter.

HORMONE ACTION THROUGH CELL SURFACE RECEPTORS There are four main classes of cell surface receptors (Table 17.1). All endocrine receptors have a high-affinity hormone-binding region: binding of the hormone to this site alters the conformation of the receptor. They also have a region which couples to an intracellular signalling mechanism that mediates the function of the hormone.

Actions mediated by adenylate cyclase POLYPEPTIDE HORMONE SYNTHESIS

884

After transcription from DNA, the mRNA precursor for a typical polypeptide hormone is subjected to posttranslational processing, particularly the excision of introns to form mRNA, which is translated into a precursor hormone. One type of precursor, a pre-hormone, consists of an aminoterminal signal sequence joined to the bioactive hormone (apoprotein), which requires no further proteolytic processing. The second type of precursor, the preprohormone, comprises an aminoterminal signal sequence followed by a prohormone (polyprotein) with one (or more) bioactive domains, which becomes available after post-translational modification of the prohormone. Such modification includes the removal of spacer sequences or cryptic peptides with no known function from prohormones, and glycosylation, sulphation or other modifications of amino acids. The signal sequence of polypeptide hormones is important in directing their transport into the cell's secretory pathway. Typically this pathway consists of the formation of transition elements from the rough endoplasmic reticulum, which are incorporated into the Golgi apparatus, leading to the formation of secretory granules. Such granules then fuse with the extracellular membrane to release the hormone. The regulation of polypeptide hormone gene expression can occur at various levels: RNA transcription or posttranscriptionalprocessing,translationand post-translational processing. As well as determining the amount of any hormone produced, these steps can create diversity of expression in a tissue-specific fashion. For instance, alternative exon splicing during post-transcriptional processing can generate different polypeptides; calcitonin and calcitonin gene-related peptide are generated in this fashion from the same RNA precursor in the thyroid and nervous systems, respectively. Post-translational processing of the pro-opiomelanocortin molecule generates a number of different peptides (see below) in different regions of the anterior pituitary, and these can be further modified by phosphorylation. Hormones in general exert their effects on the target organ in one of two ways. They may act on cell surface receptors, leading to intracellular signalling, or as transcription regulatory factors, by binding to intracellular receptors that usually cause gene transcription.

The first such effector mechanism to be recognized was the adenylate cyclase system, which, when activated, generates cyclic AMP (from ATP) as second messenger (Fig. 17.1). It is now recognized that hormones can either activate or inhibit this enzyme, depending upon whether they are linked to it via a stimulatory (Gs) or an inhibitory (G;) guanosine triphosphate (GTP)-containing coupling

TABLE 17.1 Classes of membrane receptor Receptor

Ligand

Signal system

G-protein coupled

Polypeptide hormones Neurotransmitters Prostaglandins

cAMP, cGMP Inositol triphosphate Ca2+

Receptor kinase

Insulin Growth factors

Protein tyrosine kinase Protein serine kinase

Receptor-linked kinase

Growth hormone Prolactin Cytokines

Receptor-associated tyrosine kinase

Ligand gated ion channel

Neurotransmitters Amino acids

Ion transport

FIG. 17.1 Mechanism of action of hormones that activate adenylate cyclase LTl The hormone (H) binds to the receptor (R), which alters the conformation of the regulatory coupling protein (Gs), allowing GTP to bind. B The binding of GTP causes the coupling protein to bind and activate adenylate cyclase (AC) in the membrane, generating cAMP. GTP breaks down to GDP and the activation then ceases.

protein. G protein consists of three subunits, , and . When the receptor is activated by the hormone, it associates with a Gs coupling protein; the -subunit dissociates and activates adenylate cyclase. The -subunit is also a guanosine triphosphatase (GTPase), and its activation leads to a simultaneous breakdown of GTP to GDP, switching off adenylate cyclase. Cyclic AMP, in turn, binds to and activates certain intracellular protein kinases, which then activate other enzymes and proteins by phosphorylation. The concentration of cyclic AMP is determined by the combined activity of adenylate cyclase and the enzyme that destroys cyclic AMP, phosphodiesterase. Hormones which act wholly or in part via the secondmessenger cyclic AMP include many polypeptide hormones such as adrenocorticotrophic hormone (ACTH), luteinizing hormone (LH), follicle-stimulating hormone (FSH), thyroid-stimulating hormone (TSH), parathyroid hormone, calcitonin and vasopressin. The effects of the catecholamine hormone adrenaline (epinephrine), mediated by the -receptor, are also produced by activation of adenylate cyclase. On the other hand, its -adrenergic effects are mediated, at least in part, by inhibiting adeny-

late cyclase (via the Gj coupling protein discussed above). This accounts in part for the mutually antagonistic a and (3 effects of adrenaline, at least in tissues that possess both types of receptors. Many of the effects of prostaglandins, which are not systemic hormones but do act via cell surface receptors, are also mediated by adenylate cyclase. A related G-protein coupled receptor system uses guanylate cyclase to generate a second messenger; atrial natriuretic peptide and nitric oxide signal through this type of receptor.

17

Actions mediated by phosphatidyl inositols and calcium transport Many hormones act through cell surface receptors independently of adenylate cyclase, by inducing the transport or release of calcium and/or cleavage of cell membrane phospholipids to generate two further second messengers, inositol 1,3,4-trisphosphate (IP3) and diacylglyceroi (DG) (Fig. 17.2). These hormones include luteinizing hormonereleasing hormone, thyrotrophin-releasing hormone and vasopressin. In general, these ligands interact with G-

FIG. 17.2 Mechanisms of hormone actions mediated by calcium transport or release H Hormones that promote calcium transport into the cell (1). (2) Receptor (activated) hormone (H) binds to receptor (R) opens up Ca2+ channel; Ca2+ binds to calmodulin, which binds to secretory granule. (3) Secretory granule fuses with cell surface and releases product. E Hormones that act via cleavage of phosphatidyl inositol biphosphate (PIP2). (1) Hormone binds to receptor. (2) Receptor (activated) cleaves PIP2 to diacylglyceroi (DG) and inositol trisphosphate (IP3). (3) IP3 causes Ca2+ release from cytosolic pool; Ca2+ binds to calmodulin. DG binds and activates protein kinase C (PKC). PKC and Ca2+/calmodulin activate protein kinase and then phosphorylate proteins (often in a synergistic manner).

885

FIG. 17.3 Mechanism of action of insulin Binding to subunit(s) alters conformation of the -subunit and switches on its protein kinase activity.

protein coupled receptors but instead of adenylate cyclase, the receptors are coupled to phospholipase C, which catalyses the generation of the second signals IP3 and DG. IP3 is believed to act principally by stimulating calcium release from a cytosolic pool of calcium and an influx of extracellular calcium. A transient rise in intracellular calcium activates intracellular proteins, notably calmodulin, which has four binding sites for calcium. Activated calmodulin stimulates certain protein kinases which, using ATP as a substrate, will then phosphorylate other proteins, thus activating them. Analogues of DG (the phorbol esters) will stimulate prolonged cell growth and proliferation and promote tumour growth. Some hormones act through more than one of these mechanisms, particularly depending on their concentration. There is a wide variety of possible combinations of different effects involving these simple organic compounds and calcium. An appreciation of the general structure of the system is of increasing importance, not only in endocrinology, but also in our understanding of tumour growth and development.

Actions mediated by tyrosine kinase activity A typical member of the receptor kinase family of receptors comprises an extracellular ligand-binding domain, a single transmembrane spanning domain, and an intracellular domain with intrinsic kinase activity. The insulin receptor is a well-characterized receptor kinase; it consists of four subunits, two a and two . The -subunit acts as a tyrosine kinase, which is capable of phosphorylating and thereby activating a number of intracellular enzymes (Fig. 17.3). This is also the signalling pathway used by a number of growth factors, including insulin-like growth factor-1 (IGF-1). Once insulin activates receptor kinase, insulin receptor substrate 1 is phosphorylated and this binds to

1

886

MCQ 17.1

a family of proteins with SH2 (src homology 2) domains. These in turn activate a cascade of intracellular enzymes, resulting in stimulation of glycogen synthesis and the other effects of insulin. It seems likely that the action of insulin in promoting glucose uptake depends upon a similar effect of altering the phosphorylation state of the glucose transporter molecule. The receptors for growth hormone, prolactin, leptin and certain cytokines are similar to the receptor kinases but have no integral kinase activity. Instead, the receptor interacts with membrane or intracellular kinases, such as those of the Janus (JAK) family.

HORMONE ACTION THROUGH INTRACELLULAR RECEPTORS Receptors for steroid hormones (including vitamin D) and thyroid hormones are located principally in the nucleus. In the presence of its ligand, the receptor binds with high affinity to specific promoter sites on DNA. Specific regions of DNA are then 'read' by the enzyme RNA polymerase, with synthesis of messenger RNA and specific important proteins (Fig. 17.4). Some hormones are made from the conversion of less active precursors in target tissues. The clearest example of this is dihydrotestosterone, which is synthesized from testosterone (the circulating form) by the enzyme 5ocreductase. This transformation is essential for many of the actions of testosterone.

CIRCULATION AND METABOLISM OF HORMONES Most polypeptide hormones have a relatively short halflife (minutes) in the circulation before they are degraded by proteases in the liver, kidneys, lungs and elsewhere. As implied above, some of the hormone is also internalized in target tissues. Exceptions to the short half-life are the glycosylated peptides - TSH, LH, FSH and human chorionic gonadotrophin (hCG) - where a coating of sugar residues partially protects the protein from degradation without apparently affecting its binding to the receptor. In some instances, such as FSH, the hormone is secreted episodically, but this is difficult to demonstrate because of its long half-life (several hours). LH, on the other hand, has a much shorter half-life and so episodic secretion is easily detected, with 'pulses' occurring up to once every 90-120 minutes. The steroid and thyroid hormones are partially protected from degradation by the liver and from renal excretion by circulating binding proteins, the most important of which are corticosteroid-binding globulin (CBG), sexhormone-binding globulin (SHBG), vitamin D transport globulin (or transcalciferin), and thyroxine-binding globulin (TBG). These proteins are less specific than the receptors; indeed, no general rules can be made for the relation

17

FIG. 17.4 Mechanism of action of steroid and thyroid hormones A Hormone (H) binds to nuclear receptor (R). B Activated receptor now binds to specific promoter on DNA, allowing access of RNA polymerase. C and D RNA transcript is processed by removal of introns and splicing. E Messenger RNA (mRNA) is translated in cytoplasm by ribosomes which synthesize proteins.

between the affinity of a compound for its binding protein and its biological activity, except to say that these proteins generally bind biologically active compounds. In addition, a substantial fraction of many of these hormones bind to albumin and, in some cases, to other proteins with low affinity. Because the concentration of albumin is high, albumin binding is in some cases (e.g. oestrogens) quantitatively important. Hormone-binding proteins serve a damping role, reducing the fluctuations produced by the episodic secretion of hormones and reducing their rate of degradation. The small fraction (0.1-5%) of total hormone in free solution (unbound) is believed to be the 'biologically active' fraction, i.e. the amount available to activate cellular receptors. In the case of steroid hormones, it is probably also the unbound plasma steroid that is monitored as part of the feedback control mechanism. In the case of the IGFs there is a whole family of binding proteins, whose control is complex but which clearly has a major influence on the local availability and action of the growth factors.

INTERACTIONS BETWEEN DIFFERENT HORMONES Two hormones acting together may transmit information differing both qualitatively and quantitatively from that conveyed by either separately. For example, glucocorticoids and adrenaline (epinephrine) interact at the levels of both production and action. Glucocorticoids induce the last enzyme necessary for adrenaline (epinephrine) synthesis; they also promote production of the -adrenergic receptor and so enhance some target hormone responses to -adrenergic agonists (Fig. 17.5). The two hormones together enhance fetal lung maturation at birth. This may have therapeutic importance in treating respiratory distress syndrome and, in later life, bronchial asthma, when analogues of glucocorticoids and adrenaline act synergistically. Many of the symptoms and signs of thyroid hormone excess result from an augmented sensitivity to adrenaline from enhanced responsiveness to -adrenergic effects. The

FIG. 17.5 Interactions between glucocorticoids and adrenaline (epinephrine) at levels of production and action

mechanism which enhances (3 (and probably diminishes a) agonist activity is not yet clear. Many of these effects (excess sweating, tachycardia and tremor) can be blocked acutely by administering a -adrenergic blocking agent while the thyroid hormone excess persists. A third example of hormone-hormone interaction is that between oestrogens and progestogens. One effect of oestradiol (and its synthetic counterparts) is to induce production of the progesterone receptor, without which progesterone has little if any effect. The presence of progesterone receptors in breast cancer is among the best available evidence that the tumour is oestrogen dependent and amenable to an anti-oestrogen, such as tamoxifen. 1

FURTHER READING ON THE PHYSIOLOGY OF HORMONE ACTION Bidey S P, Best L C, Ray D W, Davis J R E 1998 Hormone receptors and intracellular signalling. In: Grossman A, ed. Clinical endocrinology. Oxford, Blackwell Science, 57-80. Postel-Vinay M-C, Kelly P A 1996 Growth hormone receptor signalling. Bailliere's Clin Endocrinol Metab 10: 323-336. Spiegel A M 1996 Mutations in G proteins and G protein-coupled receptors in endocrine disease. J Clin Endocrinol Metab 81:2434-2422.

887

White A, Gibson S 1998 ACTH precursors: biological significance and clinical relevance. Clin Endocrinol 48: 251-255.

CLASSIFICATION OF ENDOCRINE DISEASES

THE AETIOLOGY OF ENDOCRINE DISEASES

HYPER-, HYPO- AND DYSFUNCTION

TUMOURS

Although excess and deficiency diseases form the basis of endocrine practice, it is an oversimplification to think of endocrine disease simply in terms of 'too much or too little hormone'. It is true that with some systems, e.g. the thyroid gland, normal production of the hormone(s) is relatively steady and the concept of pure excess or deficiency approximates to the truth. However, in other systems, e.g. the adult female reproductive system, there is an inbuilt and complex rhythm. Each of the major components of this system (hypothalamus, pituitary and ovary) secretes hormones and responds to those produced by other components. The interactions depend on both the circulating levels of hormone (often more than one) and the very important dimension of time. Here, the capacity for responses to become out of phase with one another is considerable, and many of the syndromes that we see with this system have features which suggest dysfunction rather than simple hyper- or hypofunction. As one abnormality may lead to another, with endocrine dysfunction we are, in some instances, still unable to dissect out the primary abnormality. Our only hope of doing so is to use the model of the normal system as a starting point.

Neoplasms of endocrine tissue are extremely common; they are usually benign and may secrete one or more hormones, or be apparently functionless. Such tumours cause problems for one of three reasons:

PRIMARY VS SECONDARY ABNORMALITIES It is essential to identify whether an abnormality of hormone production is primary or secondary. Adrenal or thyroid insufficiency, for example, may be due either to intrinsic disease of these glands, or to a failure of secretion of the trophic hormone (ACTH and TSH, respectively) by the pituitary. Pituitary deficiency in turn may also be intrinsic, or due to damage to the hypothalamus. The distinction may be important from the point of view of treatment and in predicting and treating accompanying features, including other hormone deficiencies.

TARGET ORGAN DISEASE

888

quite common. Only recently have methods become available for examining such defects.

There are many instances of 'target organ' endocrine diseases. These generally involve receptor proteins with a binding site for the hormone, linked to an effector system. Although major disorders due to problems at this level are rare they are well documented and minor problems are

• They may present as a local space-occupying lesion. This is particularly true of tumours of the pituitary gland, because there is very little room for expansion without producing pressure effects on important surrounding structures. • By secreting a hormone, or hormones, they may produce the syndrome associated with excess of that hormone or of the gland that it controls. • They may destroy or compress the normal gland and lead to underproduction of other hormones (again, this applies almost exclusively to the pituitary gland). Malignant tumours of endocrine tissue share many of the properties common to other cancers. With some (such as the pituitary), distant metastases are rare. With endocrine tumours, hormone production may continue or be abnormal; for example, carcinomas of the adrenal cortex may secrete the normal product cortisol, or precursor steroids, or no steroids at all. The control mechanisms of hormone secretion for malignant tumours are generally more disturbed than are those of benign tumours. Occasionally, the properties of the normal gland can be used to treat the tumour; thus, for example, the uptake of radioactive iodine by the tumour can be used very effectively to treat metastases in papillary and follicular carcinomas of the thyroid gland. In some familial syndromes, in which tumours occur in more than one endocrine gland (so-called multiple endocrine neoplasia, MEN), an underlying fundamental defect in cell regulation is suspected (p. 938).

PRIMARY, SECONDARY AND TERTIARY HYPERFUNCTION Primary, secondary and tertiary hyperfunction can be readily appreciated in the case of the parathyroid glands (Fig. 17.6). Primary hyperparathyroidism exists when excess secretion of parathyroid hormone, not apparently secondary to any other process, leads to hypercalcaemia and other metabolic changes. It may be due to parathyroid adenoma or to 'primary hyperplasia'.

commonest inborn error of adrenal cortex function is 21hydroxylase deficiency (Fig. 17.32, p. 933), which is inherited as an autosomal recessive condition. In its extreme (salt-losing) form it is associated with blocks in the conversion of progesterone to aldosterone, and 17-hydroxyprogesterone to cortisol, leading to mineralocorticoid and glucocorticoid deficiency. Clinical disease results in part from deficiencies of these two hormones, and in part from the build-up of precursors and their androgenic by-products. Finally, excess ACTH secretion (in response to cortisol deficiency) and excess renin and angiotensin secretion (in response to aldosterone deficiency) also result. These tend to restore a new steady state, but at the expense of marked adrenal hyperplasia and even greater secretion of androgenic by-products. Recently, increasing numbers of 'target organ' endocrine diseases have been characterized, e.g. the androgen receptor defect (which causes testicular feminization) and 5a-reductase deficiency (which causes a particular form of male pseudohermaphroditism, see p. 949).

17

FIG. 17.6 Primary, secondary and tertiary hyperparathyroidism.

AUTOIMMUNE ENDOCRINE DISEASE Secondary hyperparathyroidism is a physiological adaptation to hypocalcaemia. For example, in vitamin D deficiency, calcium malabsorption leads to hypocalcaemia and secondarily increased parathyroid hormone (PTH) secretion. This exerts effects on kidneys and bone which restore the plasma calcium level to normal while increasing phosphate excretion. In chronic renal failure other factors make this process less efficient (p. 1053), and so the parathyroid hyperplasia and PTH excess tends to be more prolonged and more extreme. If this secondary hyperparathyroidism leads to the development of an autonomous parathyroid tumour or tumours, tertiary hyperparathyroidism is said to have developed. After renal transplantation temporary hypercalcaemia is quite common. Although often referred to as tertiary hyperparathyroidism, it is usually due to the increased bulk of hyperplastic parathyroid tissue which has developed during prolonged renal failure, rather than to true autonomous tumour development.

ABNORMALITIES OF HORMONE BIOSYNTHESIS AND ACTION As the pathways of production and action of the major hormones have been defined, so defects in these pathways leading to specific clinical syndromes have been discovered. Often, logical and effective treatment of the condition has resulted. The defects may be inborn, due to single point mutations, or produced by pharmacologists developing new drugs. Most inborn errors occur in the peripheral rather than the central (hypothalamus and pituitary) endocrine glands. A single example illustrates some general principles. The

Autoimmunity is a frequent cause of destruction and malfunction of parts of the endocrine system. It is evident that both antibody-mediated and cell-mediated immunity may be involved, but the former is better understood. A wide range of antibodies may be involved, which can be directed at cell surface components, or at microsomes, mitochondria and other intracellular organelles. In many cases it is not clear which, if any, of the antibodies cause tissue damage and which are produced in response to cell destruction. The immune disturbance leads to: • Destruction of the gland, e.g. in autoimmune Addison's disease (where adrenal antibodies are generally present in plasma) or in primary myxoedema; • Lymphocytic infiltration of the gland, as in Hashimoto's thyroiditis (where a high titre of thyroid peroxidase (TPO) and thyroglobulin (TG) antibodies is invariably present); • Uncontrolled stimulation of the gland. In thyrotoxic Graves' disease antibodies are produced that mimic TSH sufficiently closely to bind to, and activate, the TSH receptor; • Blocking of the stimulatory hormone, e.g. some cases of autoimmune hypothyroidism. Different individuals vary greatly in their inherited tendency to develop autoimmune endocrine disease. The full basis for individual susceptibility is not known, but one factor is the histocompatibility complex HLA. Certain HLA alleles, for example HLA-DR3, confer susceptibility. Gender appears to be important, as these disorders are 3-10 times more common in women. Other genetic and environmental factors remain poorly defined. The term organ-specific autoimmune disease encompasses both endocrine and non-endocrine diseases, and

889

includes autoimmune Addison's disease, Graves' disease, Hashimoto's disease, autoimmune hypothyroidism, myasthenia gravis, hypoparathyroidism, true autoimmune diabetes mellitus, pernicious anaemia, autoimmune ovarian failure and vitiligo (autoimmune skin depigmentation). As a general rule, individuals with one such disease are more likely than normal individuals to develop another.

FURTHER READING ON THE AETIOLOGY OF ENDOCRINE DISORDERS Weetman A P 1998 Endocrine autoimmunity and related conditions. Kluwer Academic Publishers, Dordrecht, 292pp

SYSTEMIC EFFECTS OF ENDOCRINE DISORDERS CENTRAL NERVOUS SYSTEM Local expansion of pituitary tumours commonly causes headache (due probably to tension on the diaphragma sellae) and damage to adjacent structures. 1 The optic nerves are particularly vulnerable to pressure, especially at the optic chiasm, and bitemporal hemianopia may result. Rarely, pituitary or suprapituitary tumours may cause hydrocephalus or damage in the region of the fourth ventricle, resulting in a syndrome of recent memory loss similar to Korsakoff s psychosis (p. 1439). Both hyper- and hypocalcaemia may precipitate any psychosis to which that individual is predisposed, or cause neuropsychiatric symptoms de novo. Hypocalcaemia may be associated with raised intracranial pressure, epileptic fits and, of course, tetany. Hyperfunction of both the adrenal cortex and the adrenal medulla may present with psychiatric manifestations. In Cushing's syndrome, the individual may present with depression or with a psychotic episode. Excess cortisol secretion may occur in severe depression per se, so it can be very difficult to establish whether or not the patient has Cushing's syndrome. Many of the symptoms and signs of emotional or physical stress result from the release of hormones, notably adrenaline, from the adrenal medulla. Tumours of the adrenal medulla (phaeochromocytomas) usually secrete noradrenaline (nonepinephrine), thus causing hypertension, but sweating and anxiety attacks are often an accompanying feature. Thyroid disease may also be accompanied by a variety of neuropsychiatric features. Thyrotoxicosis may present as anxiety, depression or behavioural disturbance, although

1 Figs17.1-17.5

890

warm sweaty hands and a rapid pulse usually distinguish the condition from a simple anxiety state. Occasionally the patient appears unduly apathetic. Proximal muscle weakness is another common feature of thyrotoxicosis. Accompanying exophthalmos due to extraocular muscle infiltration may cause diplopia. Hypothyroidism may also present with depression. Its other neurological manifestations include carpal tunnel syndrome, myotonia and cerebellar dysfunction. Coma may be the presenting sign of a number of endocrine disorders. These (and other) manifestations of hyper- and hypoglycaemia are considered in Chapter 19. In profound hypothyroidism coma is likely to be caused by conditions that affect respiratory drive or by hypothermia. In severe hypopituitarism stupor leading to coma commonly follows a relatively minor disturbance (such as a trivial accident or infection). It is due in part to profound cortisol deficiency, in part to hypothyroidism, and is usually associated with hyponatraemia due to excess secretion of antidiuretic hormone (ADH) as a consequence of cortisol deficiency. In diabetes insipidus (DI) hypernatraemia may occur rapidly and produce convulsions or irreversible brain damage if access to water is restricted (e.g. postoperatively). Conversely, in the treated patient with DI, unlimited drinking may lead to cerebral dysfunction and fits from hyponatraemia (water intoxication).

CARDIOVASCULAR SYSTEM Hypotension, especially in the upright posture, commonly results from either glucocorticoid or mineralocorticoid deficiency, whereas hypertension, usually of mild degree, may result from mineralocorticoid excess. Phaeochromocytoma also causes hypertension, which can be episodic or sustained. Endocrine diseases leading to electrolyte disturbances such as hyper- or hypocalcaemia or hypokalaemia may present with dysrhythmias and/or characteristic ECG abnormalities. Prolonged hypercalcaemia and prolonged glucocorticoid excess are associated with an increased incidence of arterial calcification. Patients with primary hypothyroidism and angina are particularly vulnerable to worsening of their symptoms and even myocardial infarction when treated with thyroxine. In the elderly, thyrotoxicosis commonly leads to atrial fibrillation. Acromegaly is associated with cardiac enlargement, hypertension and a form of cardiomyopathy.

SKELETAL SYSTEM Skeletal manifestations of endocrine disease are commonly caused by osteoporosis, as in Cushing's syndrome, sex hormone deficiency or prolonged hyperthyroidism. Primary hyperparathyroidism may lead to calcification of articular cartilage, and present with calcium pyrophos-

phate crystal arthropathy (pseudogout, see p. 1171). The classic bone disease described by von Recklinghausen, osteitis fibrosa cystica, with giant cell tumours of bone due to primary hyperparathyroidism, is now very rare. Occasionally, Graves' disease is associated with finger clubbing and periosteal new bone growth (so-called thyroid acropachy). In acromegaly, periarticular overgrowth of new bone and increased cartilage quite commonly lead to osteoarthritis. In severe acromegaly and gigantism, skeletal abnormalities may be gross.

GENITOURINARY AND

REPRODUCTIVE SYSTEMS Conditions that affect sex hormone secretion in utero produce malformation of the external and internal genitalia (p. 939). Renal function may be disturbed by a number of metabolic disorders, some of which are endocrine in origin. For example, both hypercalcaemia and hypokalaemia may present with a form of (nephrogenic) diabetes insipidus and polyuria due to a water diuresis.

INTEGUMENTARY SYSTEM

Hirsutism and acne are common manifestations of androgen excess in the female. Mild hirsutism, skin atrophy, bruising and poor wound healing are features of Cushing's syndrome. Hyperpigmentation is common in Addison's disease. Vitiligo and mucocutaneous candidiasis occur in some autoimmune endocrine diseases. Of other endocrine disorders presenting with skin manifestations, the glucagonoma syndrome, which causes migratory necrolytic erythema, is probably the most striking (p. 833).

limitations of the assay method, including interference by drugs. The minimum number of tests should be made to answer a particular question. Old tests are rapidly being superseded by new and better methods.

17

BASAL MEASUREMENTS Much can be learnt by basal measurements of so-called routine biochemistry (Table 17.2). Where endocrine disease involves hormones that control the electrolytes, these measurements are the mainstay. Thus, a parathyroid tumour is most unlikely unless the serum calcium level is raised, and a low serum phosphate further suggests the diagnosis. Aldosterone deficiency in Addison's disease is suggested by the finding of a high serum potassium level, usually associated with a low sodium and elevated urea owing to extracellular fluid depletion. In assessing posterior pituitary function, the most widely used test (fluid deprivation) depends upon assessing the renal response, mediated by ADH, to deliberate but controlled withholding of fluids. In most cases, and certainly to exclude diabetes insipidus, a simple study of the relationships of fluid loss to serum and urine osmolality suffices, without the need for difficult assays of ADH. Basal measurements of hormone levels or metabolites are most practical when they can be conducted on serum or urine samples collected at random. Nowadays, the widespread development and application of radioimmuno- and immunoradiometric assays have helped enormously, especially for hormones that are relatively stable in body fluids. This applies to the steroid hormones, thyroid hormones and most of the pituitary hormones. In most cases, serum is preferred to plasma; clotted blood can be left for 1-2 hours before centrifugation, and serum stored indefinitely

ALIMENTARY SYSTEM Thyrotoxicosis commonly causes diarrhoea and sometimes malabsorption. Hypothyroidism causes constipation. Conditions such as Addison's disease, hypopituitarism and diabetic ketoacidosis may present as apparent upper abdominal emergencies with epigastric pain and rigidity. Hypercalcaemia commonly causes vomiting, constipation and abdominal pain.

TABLE 17.2 'Routine' biochemical pointers in blood plasma to endocrine disorders Pointers

Condition

Hyperkaiaemia, hyponatraemia Hypokalaemia, metabolic alkalosis (high serum bicarbonate)

Addison's disease Conn's syndrome Cushing's syndrome (severe) Inappropriate ADH secretion Hypopituitarism Severe hypothyroidism Diabetes insipidus Diabetic ketoacidosis

Hyponatraemia, hypochloraemia

ENDOCRINE FUNCTION TESTING Endocrine function tests are used to establish or to refine the diagnosis, to monitor the response to treatment, to assess the prognosis, and to establish the pathogenesis of the condition, often in combination with endocrine gland imaging. Clearly, it is important to understand both the underlying physiology on which the test is based and the

Hypernatraemia, hyperchloraemia Hyperglycaemia, hyperchloraemia, metabolic acidosis (low serum bicarbonate) Hyperglycaemia, uraemia, hypernatraemia, hyperchloraemia Hypercalcaemia, hypophosphataemia, hyperchloraemia Hypocalcaemia, hyperphosphataemia

Diabetic hyperosmolar coma Primary hyperparathyroidism Hypoparathyroidism

891

at -20°C. Economic and methodological considerations dictate that samples are assayed batchwise, so that the result is often not available at once. It may therefore be necessary to start treatment while the result is awaited. Urine sometimes has the advantage of reflecting integrated secretion and the 'free' (unbound) level (e.g. with cortisol). However, the major steroids in urine are metabolites several steps away from the biologically active compound.

FURTHER READING ON ENDOCRINE FUNCTION TESTING Jenkins R C, Ross R J M 1998 Protocols for common endocrine tests. In: Grossman A, ed. Clinical Endocrinology. Blackwell Science, Oxford, 1117-1134.

THE PITUITARY GLAND DYNAMIC FUNCTION TESTS The static nature of information obtained by basal measurements has led to the development of a whole range of dynamic function tests, where the system under study is deliberately driven, usually by interrupting negative feedback or by stimulation at one or more levels (Fig. 17.7). In testing anterior pituitary function there is a bewildering array of possible tests, not all of which are useful. For example, stimulation tests for ACTH (and so cortisol) with insulin hypoglycaemia, glucagon, vasopressin and CRH have all been devised, in addition to direct adrenal stimulation with synthetic ACTH (synacthen). With TSH and the thyroid hormones, marked episodic and diurnal variation - a problem inherent in assessing the ACTH-cortisol axis - is missing and, especially with new sensitive TSH assays, dynamic testing with TRH is not needed. Sometimes two or more tests can be carried out simultaneously, e.g. insulin, TRH and GnRH can all be given together to assess the extent of hypopituitarism. An example of a test used to monitor or predict response to treatment is the measurement of serum prolactin or growth hormone levels after a single dose of bromocriptine in patients with prolactinoma or acromegaly, respectively. In complex endocrine systems such as the ovulatory cycle, repeated samples over the cycle of more than one hormone may be necessary, and may be much more useful than contrived 'dynamic function' tests.

892

FIG. 17.7 Principles of dynamic endocrine function testing Hypothalamic hormone A stimulates pituitary to produce pituitary hormone B, which then stimulates peripheral gland to produce C. Hormone C (or a synthetic analogue) inhibits production of A and/or B by negative feedback.

The pituitary gland is composed of two functionally separate lobes, the anterior and the posterior (pars nervosa), with an intermediate lobe which in humans is vestigial. The gland is derived from a downgrowth from the floor of the third ventricle and from Rathke's pouch, which is formed by invagination of part of the roof of the primitive nasopharynx. The pituitary gland is almost entirely enclosed in a bony box, the sella turcica; this has a floor and an anterior and posterior wall, the lateral extremities of which become the anterior and posterior clinoid processes. Immediately lateral to the pituitary gland are the two cavernous sinuses, with the third, fourth, fifth and sixth cranial nerves sandwiched between: the carotid syphon courses through the cavernous sinus (Fig. 17.8). Overlying the pituitary is a fibrous sheet, the diaphragma sellae, which is pierced by the pituitary stalk; this slender structure carries the nerve fibres to the posterior pituitary gland and the blood supply to both lobes, including the portal blood vessels draining from the median eminence of the hypothalamus.

FIG. 17.8 Gross anatomy and relations of the pituitary gland A View of pituitary gland, anterior clinoids, dorsum sellae and optic chiasm from above. B Schematic coronal cross-section through pituitary gland to show relation to important structures nearby.

17

TABLE 17.3 Terminology of hypothalamic and anterior pituitary hormones Hypothalamic hormone

Effect (stimulation or inhibition)

Gonadotrophin releasing hormone (GnRH) LH-releasing hormone/factor, LH-FSH releasing hormone (LHRH) Corticotrophin-releasing hormone (CRH) Arginine vasopressin (or antidiuretic hormone; ADH) Somatostatin (growth hormone/somatotrophin releaseinhibiting hormone) Growth hormone-releasing hormone (GHRH) Thyrotrophin-releasing hormone (TRH) Dopamine

The main hypothalamic and pituitary hormones are shown in Table 17.3.

+ +

SYNTHESIS AND SECRETION OF THE MAIN HORMONES

Luteinizing hormone (LH) Follicle-stimulating hormone (FSH) Adrenocorticotrophic hormone (ACTH, corticotrophin)

+

Growth hormone (GH, somatotrophin) + +

Thyroid-stimulating hormone (TSH, thyrotrophin) Prolactin

TABLE 17.4 Factors directly affecting anterior pituitary hormone synthesis and release Pituitary hormone

THE ANTERIOR PITUITARY GLAND

Pituitary hormone

ACTH

Stimuli

Inhibitors

CRH

Cortisol

Vasopressin (ADH) lnterleukin-1 (IL-1)* GH

GHRH

Somatostatin Insulin-like growth factor 1(IGF-1) Cortisol

TSH

TRH

T 3 ,T 4 Somatostatin

LH

GnRH Oestradiol (+ve feedback)

Oestradiol Progesterone Testosterone (male)

FSH

GnRH

Inhibin Oestradiol Progesterone Testosterone (male)

Prolactin

TRH Oestradiol**

Dopamine

Much is now known about how the synthesis and secretion of anterior pituitary hormones is controlled, and this is summarized in Table 17.4.

Adrenocorticotrophic hormone ACTH is synthesized in specific cells (the corticotrophs) from a large precursor molecule called proopiomelanocortin (POMC, Fig. 17.9). It is produced in both the anterior and the intermediate lobe of the pituitary, but the final processing differs in the two. Of the other ACTH-related peptides, B-lipotrophin (LPH), B-endorphin and a-melanocyte-stimulating hormone (MSH) are important. LPH and B-endorphin are synthesized and secreted with ACTH; the latter may have a role as an opiate in modifying the response to pain at times of stress. The synthesis and release of ACTH is controlled by corticotrophin-releasing hormone (CRH), which acts synergistically with vasopressin. There are at least two basic rhythms of ACTH secretion: episodes of secretion every 1-3 hours and at meal times are superimposed on a 24-hour rhythm, with lowest levels during the earliest hours of sleep and peak secretion just before wakening (Fig. 17.10). At any time, exposure to sudden physical or mental stress will overcome this rhythm and lead to a pulse of ACTH secretion. The adrenal cortex responds in turn with a pulse of cortisol secretion within minutes, so that the 24-hour plasma profile of cortisol is a somewhat dampened version of the 24-hour ACTH profile.

* Effect probably via CRH. **Increases prolactin synthesis, produces lactotroph hyperplasia but inhibits milk secretion.

The most important role of ACTH is to control the synthesis and secretion of the glucocorticoid cortisol. As would be expected from its relationship to MSH (see Fig. 17.9), ACTH possesses MSH activity and will directly stimulate the melanocytes of the skin and buccal mucosa when secreted in excessive amounts. This accounts for the increase in pigmentation to above normal (for that indi-

893

FIG. 17.9 Relationship between pro-opiomelanocortin (POMC) and the various intermediates and end-products of its processing in the pituitary gland LPH = lipotrophin; NT = NH2 terminal peptide; MSH = melanocyte-stimulating hormone; CLIP = corticotrophin-like intermediate lobe peptide. FIG. 17.11 Control of growth hormone and IGF-1 secretion

FIG. 17.10 Plasma cortisol levels through the 24-hour cycle to show circadian variation, pulsatile secretion and influence of meals and stress

vidual) seen in Addison's disease (p. 930) and Nelson's syndrome (pp. 906 and 936). 1

relatively recently in evolution; their genes are located on chromosome 17. PRL is much less closely related and is encoded on chromosome 6. GH is synthesized in specific cells known as somatotrophs, where it is stored in granules and released episodically in response to peaks of hypothalamic secretion of GHRH (Fig. 17.11). The precise physiological interactions of GHRH with somatostatin, which inhibits its release, are complex. Much of the GH is secreted at night and episodically. Its secretion is inhibited by a rise in blood glucose (the basis of a useful suppression test) and stimulated by hypoglycaemic stress and by amino acids such as arginine (the basis of stimulation tests). GH controls the secretion of insulin-like growth factor-1 (IGF-1, somatomedin C), which functions as a hormone controlling the growth of cartilage and soft tissues. During puberty, the secretion of sex hormones appears to increase the pituitary secretion of GH, which is probably important in promoting growth before the sex hormones lead to closure of the epiphyses. GH also acts directly on tissues to control local production of IGF-1. The effects of IGF-1 are modulated by specific binding proteins, which are also under hormonal control.

Prolactin Growth hormone Growth hormone (GH), human placental lactogen (HPL) and prolactin (PRL) belong to a family of polypeptide hormones. GH and HPL both contain 191 amino acids, and PRL 199. GH and HPL are closely homologous (with 85% identical residues) and probably arose by gene duplication

1MCQ17.2

894

PRL is produced in the lactotrophs, under stimulation by TRH and, more importantly, under tonic inhibition by dopamine. PRL has a basically nocturnal pattern of secretion and is also released in response to stress. Oestrogens probably act directly on the lactotrophs, increasing their number and activity. In pregnancy, PRL levels are high and so lead to hypertrophy of the breast epithelium (Table 17.5). However, high oestrogen and progesterone levels inhibit milk production until after delivery of the baby. During lactation, suckling stimulates the secretion of PRL via a reflex arc. High PRL levels have a contraceptive effect, probably by inhibiting cyclical gonadotrophin release.

TABLE 17.5 Action of hormones on the human breast Hormone Polypeptide hormones Prolactin Oxytocin Human placental lactogen Growth hormone Insulin Steroid hormones Oestrogens Progestogens Cortisol Androgens

Action

Milk protein synthesis and excretion in ducts and lobules Milk ejection. Contraction of myoepithelial cells Similar to prolactin Permissive' effects 'Permissive' effects

Increase duct and breast fat growth Inhibit prolactin-induced milk secretion Inhibit lactation (effect requires oestrogen) 'Permissive' effects Inhibitory effects on ?stroma

Luteinizing hormone and follicle-stimulating hormone LH and FSH are controlled by the same releasing hormone and are produced in the same cell type (the gonadotroph). They resemble TSH and hCG in being two-subunit polypeptides, and in possessing a sugar 'coat'. The ocsubunits of these peptides are identical, but the B-subunits differ greatly and confer the specificity. The ratio and absolute amount of LH and FSH secreted in response to a single pulse of GnRH vary greatly under different circumstances, being modified by age, sex hormone secretion by the gonads, and the preceding pattern of discharge of GnRH. Low-frequency discharge seems to favour FSH synthesis and release. In females, a rise in oestrogen level differentially reduces FSH and eventually promotes LH release ('positive feedback', see Fig. 17.37A). The gonads also produce a polypeptide hormone (inhibin) which feeds back to the pituitary to switch off FSH synthesis and secretion selectively.

Thyroid-stimulating hormone TSH promotes the synthesis and secretion of thyroid hormones, and the growth of the thyroid gland. The principal hypothalamic control is from thyrotrophin-releasing hormone (TRH), whose secretion, as with GnRH, is apparently pulsatile. The thyroid hormones feed back on both the hypothalamus, to regulate the synthesis of TRH, and the pituitary gland, inhibiting TSH release.

ANTERIOR PITUITARY HYPOFUNCTION Anterior pituitary hypofunction may develop rapidly or insidiously, and may affect all the pituitary hormones, a few, or only one. In progressive lesions LH and FSH are usually

lost first, followed by GH,TSH and, lastly, ACTH. We consider here those causes associated with partial or complete loss of all the anterior pituitary hormones, causing a state of so-called (anterior) panhypopituitarism. This will in some cases be associated with diabetes insipidus, where ADH secretion from the posterior pituitary is also impaired.

17

Aetiology The causes of panhypopituitarism are listed in Table 17.6. The commonest vascular cause is Sheehan's syndrome, in which hypopituitarism occurs following a pregnancy complicated by a period of shock and hypotension, commonly due to postpartum haemorrhage. The pituitary normally enlarges during pregnancy under the action of oestrogens on the lactotrophs, but the blood supply does not; it is therefore particularly vulnerable to a fall in blood pressure. There are usually no symptoms (such as headache) directly related to the infarction, but lactation does not occur and the symptoms of hypopituitarism soon follow. Menstrual periods do not return. Rarely, prolonged shock in other situations may precede the development of panhypopituitarism. Tumours, both within the pituitary (functioning or nonfunctioning adenomas) and suprasellar (e.g. the craniopharyngioma, which develops from Rathke's pouch), are the most common causes of panhypopituitarism. The clinical features include pressure effects on surrounding structures, including the optic chiasm and the hypothalamus. In the case of intrasellar tumours, panhypopituitarism is usually a late development, as some normal pituitary tissue remains. If the tumour is functional there may be additional features other than those from pressure on the remaining normal pituitary. A large prolactinoma, for example, may cause hypopituitarism and hyperprolactinaemia. Suprasellar tumours are more likely to affect the pituitary stalk, and thereby the posterior pituitary gland. If a large pituitary tumour becomes infarcted it will cause the syndrome of 'pituitary apoplexy' (severe headache, neck stiffness, vomiting), which may be followed by hypopituitarism if the normal tissue is compromised (see figs 17.14 and 17.15). Granulomatous conditions that may occasionally be responsible include eosinophilic granuloma (histiocytosis X, p. 1257) and sarcoidosis. Carotid aneurysms may occasionally damage the pituitary. Panhypopituitarism may develop long after pituitary irradiation (e.g. for a growthhormone-secreting tumour). This is usually a consequence of hypothalamic damage, and commonly affects growth hormone and gonadotrophin secretion before other hormones. In many cases of panhypopituitarism the aetiology is obscure; some may be autoimmune. In some cases there is an empty sella turcica and a defective diaphragma. It is still undecided whether the so-called 'empty sella syndrome' is a primary cause of hypopituitarism. Most subjects with an

895

TABLE 17.6 Causes of panhypopituitarism Cause

Mechanism

Associated features

Developmental abnormalities

Failure of hypothalamic or pituitary development Perinatal asphyxia

Severe cranial abnormalities Midline facial abnormalities GH deficiency predominant

Infection Encephalitis Bacterial meningitis

Hypothalamus or stalk damage, or pituitary destruction

History of encephalitis or bacterial meningitis

Granulomata Eosinophilic granuloma Sarcoidosis

Infiltration of pituitary/hypothalamus

Histiocytosis X (see p. 1257) Sarcoidosis (p. 682).

Hypotension, especially when the pituitary is enlarged, e.g. by pregnancy or tumour

Sheehan's syndrome from antepartum or postpartum haemorrhage Hypotension (surgical or traumatic) Pituitary enlargement/tumour Cranial nerve lesions Apoplexy simulates subarachnoid haemorrhage Ballooning of pituitary fossa

Vascular Infarction Haemorrhage

Pituitary apoplexy Subarachnoid haemorrhage

Aneurysms

Pressure

Tumour Suprapituitary (craniopharyngioma)

Pressure on stalk

Pituitary

Pressure or invasion of normal gland

Empty sella

Mechanism uncertain

Ballooning of pituitary fossa

Autoimmune

Antibodies or T-cell-mediated damage to pituicytes

Other organ-specific autoimmune disease

Trauma

Severe head injury causing stalk damage

Brain and cranial nerve damage

apparently empty sella have normal pituitary function and are detected by chance on CT scanning.

Local pressure effects of pituitary and suprasellar tumours The principal pressure symptoms are headache, which may be severe and persistent, and visual field disturbances. Bitemporal hemianopia is characteristic but may be preceded by loss of the upper outer quadrant (if the tumour is below the chiasm). The patient may notice loss of the lateral part of the visual fields; the defect will not at first be detectable by confrontation, but can be shown by formal perimetry with a Bjerrum's screen, and may be most obvious if a red object is used. Fundoscopy may reveal bi- or unilateral optic atrophy. Other field defects are also

1

896

Figs 17.1-17.5

2MCQ17.3

Pressure on optic chiasm or brainstem Suprasellar calcification (craniopharyngioma) Large pituitary fossa Hormone excess if tumour functional

SUMMARY 1 Clinical features of hormone-producing pituitary tumours • Local symptoms are headache and visual disturbance; large tumours of any type may cause hypopituitarism. • In acromegaly the tumour is usually large, but is often small in Cushing's disease. • Incipient visual loss merits an immediate trial of bromocriptine if the prolactin is very high. • Macroprolactinomas occur equally in men and women. Microprolactinomas are generally detected in women.

quite common. More rarely, other cranial nerves (III, IV, V and VI) are involved. With extensive lesions a syndrome reminiscent of Korsakoffs syndrome with recent memory loss may occur if there is pressure on the floor of the third ventricle. Disturbances of appetite (causing hyperphagia) and thirst may also be seen. The latter causes particular problems if diabetes insipidus is also present, as the patient does not

SUMMARY 2 Summary of the Royal College of Physicians of London Working Party guidelines: recommendations for service provision and management of pituitary tumours

standard rates of growth. Panhypopituitarism in childhood is also associated with failure of pubertal development. Other common features of hypopituitarism in childhood include a tendency to hypoglycaemia and other general features discussed below.

Recommendations for service provision

Adult life

• All patients with pituitary tumours require evaluation and followup at a specialist centre. • Management plans should be agreed jointly between endocrinologist, neurosurgeon and radiotherapist. • A diagnostic register should be held in the specialist centre. • Patients should receive detailed education and be informed about self-help groups. • MRI scanning is preferable to CT.

In panhypopituitarism developing in adult life, patients usually feel unwell in a non-specific way. They appear pale without being anaemic, and the skin has a 'wax doll' appearance. They notice the cold, although features of hypothyroidism are often not as gross or obvious as in myxoedema. Women have amenorrhoea and men lose their sex drive and potency. Loss of secondary sexual hair and a decline in other secondary sexual characteristics is a feature in both sexes. The patients may present with a 'hypopituitary crisis' when challenged by trivial stress, which would normally be met by increased secretion of ACTH and cortisol. Acute abdominal pain may mimic an acute abdominal emergency. This syndrome responds rapidly to cortisol administration. Patients commonly develop hypoglycaemia, but because of an associated lack of adrenaline secretion (see Fig. 17.5, p. 887), they do not develop symptoms of 'classic' hypoglycaemia, i.e. a sympathetic response to the low blood sugar. They may therefore present with coma due to hypoglycaemia. Coma may also be due to hyponatraemia, as there is commonly an excess of ADH secretion (cortisol inhibits this secretion and the renal response to ADH). Manifestations of growth hormone deficiency include pallor, thin skin and hypoglycaemia. Those of ACTH (and thus cortisol) deficiency include malaise, somnolence, hyponatraemia, poor response to stress, hypoglycaemia, confusion, coma, hypotension and fever. Effects of sex hormone deficiency include delayed puberty, sexual infantilism, amenorrhoea and lack of breast development. Men show an abnormal 'puppy fat' distribution of body fat and lack of muscular development. Loss of secondary sexual hair is much more marked in panhypopituitarism than in conditions where there is only a lack of gonadal sex hormones. As with other causes of androgen deficiency in men, the patient's partner may complain more than he does. Loss of sex drive is also common in hypopituitary women. Features due to TSH deficiency include cold intolerance and dry skin, although hypothyroidism may be masked by other deficiencies. Diabetes insipidus is discussed on page 911; its effects are often less marked when cortisol is also deficient, as the latter is necessary for a water diuresis in ADH deficiency. PRL deficiency leads to failure of milk production, which is important only to the woman who is breastfeeding. ©

Objectives of treatment for all types of pituitary tumour: • • • •

Restoration of wellbeing Relief of pressure symptoms Treatment of any syndrome of hormone excess Recognition of or substitution for specific hormone system defects • Prevention of tumour regrowth • Reversal of increased long-term mortality. Comment The evaluation and management of diverse abnormalities of anterior pituitary function are not standardized. Alternative strategies are discussed in detail in this document for each disorder. Reference Clayton RN, Wass JAH 1997 Pituitary tumours. Recommendations for service provision and guidelines for management of patients. Royal College of Physicians, London.

drink enough to prevent hyperosmolarity. The occurrence of pressure effects depends mainly on the size of the tumour: for example, GH-secreting pituitary tumours (in acromegaly) are generally much larger than ACTHsecreting tumours (in Cushing's disease). Non-functioning tumours and prolactinomas are the two other intrasellar lesions that commonly cause pressure symptoms. Although pituitary tumours may be locally invasive, malignancy is very rare. 1

Clinical features of anterior pituitary hypofunction The clinical features of anterior pituitary hypofunction depend in part upon the age of onset.

Childhood If the onset of panhypopituitarism is before growth is completed, the patient will have the features of hypopituitary dwarfism, i.e. normal body proportions but failure to grow. The latter is detected and documented by careful measurements of height, using a stadiometer, recorded on standard growth charts (Fig. 17.12). From serial measurements, growth velocity (cm/year) can be calculated and related to

17

Investigation of suspected pituitary hypofunction If hypopituitarism is suspected in an ill patient, a clotted blood sample is taken for hormone analysis and glucocorticoid treatment commenced at once, most conveniently by

897

FIG. 17.12 Hypopituitarism in childhood A Child aged 5 1 years (before GH therapy) and B aged 6 1 years, after 9 months' GH therapy, when he had grown 11 cm. C Height measurements of the boy illustrated in A and B before (black line) and during (blue line) GH treatment. Normal values for each centile are also shown (courtesy of Professor S M Shalet). [D] Height measurements in a boy with delayed puberty treated with androgens. Normal values for each centile are also shown (courtesy of Professor S M Shalet).

giving l00mg hydrocortisone intravenously, followed by 50 mg 6 hours later and then a normal regular replacement dose. If the patient does not have pituitary or adrenal insufficiency such treatment is unlikely to be harmful, and if they do it may be life-saving. There is a risk of precipating an adrenal crisis by starting thyroid replacement first (as this increases the metabolic clearance of cortisol). Much of value can be learnt from basal measurements of serum urea and electrolytes, LH and FSH, testosterone (in men), progesterone and oestradiol (in women), cortisol, thyroxine (T4), GH and PRL. These measurements help considerably to establish the diagnosis and possibly its cause. Dynamic hormone tests are often used as the 'gold standard' for assessing pituitary function (Fig. 17.13), although responses to TRH and GnRH actually add little to basal hormone measurement. For convenience, GnRH, TRH and insulin can be given simultaneously (Fig. 17.13). Insulin should be given in a low dose (0.15 units/kg) with close observation by a doctor throughout. An ECG should be done first, and this test should not be performed in the elderly or those with heart disease or epilepsy. Glucose for

1

898

MCQ 17.4

intravenous injection, or glucagon, should be available, and given early rather than late. GH and cortisol measurements are made serially over 2 hours. In many centres the short synacthen test is used to assess the cortisol response instead (see p. 929). Glucagon can also be used in adults to assess GH and cortisol reserve.

Investigations to establish the underlying cause of hypopituitarism The history and physical examination may point to a particular cause of hypopituitarism, such as previous infarction, meningitis or radiotherapy. Imaging Lateral and anteroposterior X-rays of the skull may reveal expansion of the pituitary fossa and erosion of the dorsum sellae and anterior or posterior clinoids (Fig. 17.14), although such changes do not indicate that the tumour responsible still exists - the fossa may be empty. For this reason, a CT or MR scan is essential (Fig. 17.15), and has now superseded other procedures. Angiography may be

17

FIG. 17.13 Responses in two patients to combined pituitary stimulation with insulin (0.15U/kg), GnRH (100ug) and TRH ( 2 0 u g ) given i.v. as a bolus at time 0 Patient 1 is a woman with amenorrhoea. She has normal responses despite previous operation for a functionless pituitary adenoma. Patient 2 is a man, also with a functionless pituitary tumour. He has absent responses for all except cortisol, which is blunted. The diagnosis is panhypopituitarism, requiring replacement therapy.

helpful if an aneurysm adjacent to the pituitary is suspected from the scans. The 'empty sella' syndrome, in which CT or MRI shows an absent pituitary gland, is usually a diagnosis of exclusion after the relevant radiological studies. In some cases the syndrome follows infarction of a tumour, whereas others are possibly due to the transmission of cerebrospinal fluid pressure to the pituitary fossa as a result of a defective diaphragma sellae. Hormone measurements Hormone measurements (PRL, GH, gonadotrophins, TSH) will indicate the nature of a functioning pituitary tumour, but there may still be major difficulties in diagnosis. For example, a modest elevation of PRL may be due simply to stalk damage, and not to prolactinoma. Levels more than about 10 times the upper limit of normal are, however, always due to a prolactinoma. Similarly, gonadotrophin levels are normally elevated after the menopause. Raised TSH is far more likely to be due to primary myxoedema than to a TSH-producing pituitary tumour. In the former, serum T4 levels will be low, whereas levels will be high with a TSH-producing tumour. A further clue is that there is an elevated level of the free a-subunit with TSH-producing tumours. 1

FIG. 17.14 Lateral skull X-rays in patients with pituitary tumours A Expansion of pituitary fossa and a 'double floor' due to asymmetrical enlargement. B Complete destruction of pituitary fossa and dorsum sellae.

Management of anterior hypopituitarism Although patients with hypopituitarism may have been ACTH- (and therefore cortisol-) deficient for many months or years without severe illness, they are nevertheless extremely vulnerable to even minor stress. A minor infection, illness or trauma may precipitate acute glucocorticoid deficiency, as may ill-advised prior treatment of hypothyroidism. As well as necessitating replacement therapy to deal with hormone deficiencies, pituitary tumours causing hypopituitarism may require treatment to relieve compressive signs and symptoms. Immediate The patient with pituitary collapse or coma should be given intravenous hydrocortisone (100mg repeated in 4 hours), which will generally lead to rapid improvement in the clinical condition, a rise in blood pressure and a fall in temperature. In some cases volume expansion may be necessary and, if the serum sodium level is very low (below 115 mmol/L) because of inappropriate ADH secretion (p. 911), slow correction to around 120mmol/L may be indicated by means of small amounts of hypertonic saline. Steroids should always be replaced before thyroid hormones, which can generally be given orally from the outset.

899

Sex hormone replacement therapy, particularly testosterone therapy in men, should be considered and generally instituted without delay, as this will greatly hasten the return of strength and wellbeing. Where occult posterior pituitary hormone deficiency coexists, it may be uncovered by glucocorticoid therapy. This should be suspected when there is marked and continuing water diuresis in the recovery phase, with low urine and rising serum osmolalities. It is best corrected with the synthetic ADH analogue desmopressin (DDAVP, p. 911). Long-term Steroid therapy There is no convenient way of mimicking the episodic secretion of glucocorticoids. Fortunately, however, this does not seem to matter, and oral treatment with hydrocortisone (cortisol) in a total daily dose of 20-40 mg given in two or, ideally, three divided doses (to approximate the normal diurnal variation) is usually satisfactory. Half the dose is given on waking in the morning and the last dose should be given no later than early evening (reversed for anyone working on a night shift). Cortisone acetate is generally equally effective. However, a possible disadvantage of cortisone over cortisol is that the plasma cortisol level rises to a peak more gradually on treatment with the former. In addition, some individuals, especially those with liver disease, have impaired conversion of cortisone to cortisol, and hydrocortisone is therefore now generally used. Prednisolone is a suitable alternative (5mg being roughly equivalent to 20mg hydrocortisone), but dexamethasone an even more potent halogenated glucocorticoid - should probably be avoided for long-term treatment because its much longer half-life means that glucocorticoid levels do not fall at night. Low levels then may be important to allow collagen repair. The patient should be given a card and wear a bracelet indicating the details of glucocorticoid replacement. As with Addison's disease (p. 931), dosage should be doubled temporarily to cover any stressful situation, and parenteral hydrocortisone will be required if oral steroids cannot be taken (e.g. during surgery, vomiting). Thyroid hormones Thyroid hormones are best replaced as oral thyroxine (100-200ug once daily). This is converted to the more active triiodothyronine (T3) in the tissues. Replacement can only be monitored by checking free T4 levels; TSH levels are obviously of no value, unlike in primary hypothyroidism.

1

900

Figs 17.1-17.5

2

MCQ 17.5

FIG. 17.15 MR scans of pituitary tumours

A Microadenoma (arrowed). B Macroadenoma with lateral and suprasellar expansion (arrowed) causing compression of the optic chiasm (courtesy of DrT Powell). 1

Sex hormone replacement In men testosterone is given in one of four ways, none of which is ideal. It may be implanted subcutaneously in the abdominal wall as fused pellets (600-800 mg every 4-6 months). More commonly, but with wider fluctuations in plasma testosterone levels, it may be given as testosterone esters - Sustanon (mixed testosterone esters) or Primoteston depot (testosterone enanthate) 250 mg i.m. every 2-4

weeks. Less effective is orally absorbed ester testosterone undecanoate (40-80mg twice daily), which also has the disadvantage of high price, and the capsules may cause gastrointestinal symptoms, perhaps because of the oleic acid vehicle. Transdermal delivery of testosterone via patches (e.g. Andropatch, each patch releasing 2.5 mg testosterone per 24 hours) has recently been introduced, but compliance is variable and the patches may cause skin irritation. However, these produce the most consistent physiological levels of testosterone. In women, where the uterus is present oestrogen therapy is indicated, combined with a cyclical progestogen to prevent endometrial hyperplasia (p. 948). In women who have had a hysterectomy oestrogen alone can be used (e.g. Premarin, which is an oral form of conjugated oestrogens, or Estraderm, which is a patch for transdermal delivery of oestradiol; gel preparations are available in some countries). Restoration of fertility In hypopituitary patients of either sex, restoration of fertility can usually be achieved by administering gonadotrophins. In men hCG (1500-2000IU i.m. three times weekly) is given to stimulate Leydig cell testosterone production, and after some 3 months or more this is combined with three times weekly i.m. injections of human menopausal gonadotrophins containing both LH and FSH (e.g. Pergonal 1 -l ampoule). Spermatogenesis should be evident after 3 months, but may take up to 12 months. Gonadotrophin therapy in women is much more complex, as FSH must be given early in the cycle and the follicular response followed by ultrasound scanning and serum or urinary oestrogen measurements (p. 948). When one or two follicles have achieved a diameter of 2cm, ovulation is produced by giving intramuscular hCG (5000 units). This treatment should only be undertaken in units with facilities for ultrasound and rapid oestrogen assays, and probably with access to egg retrieval and in vitro fertilization. An alternative treatment, effective in some cases of isolated gonadotrophin deficiency, is to administer GnRH in a pulsatile fashion through a pump connected to a subcutaneous needle. 2 Posterior pituitary hormone replacement Posterior pituitary hormone replacement is discussed on pages 910-912.

SUMMARY 3 Management of anterior hypopituitarism • Glucocorticoid replacement: hydrocortisone 10mg on waking, 5-10mg lunchtime and 5-10mg late afternoon; advice on steroid replacement in illness • Levothyroxine replacement: thyroxine 100-200ug daily • Sex hormone replacement: testosterone esters or patches in men (e.g. Sustanon 250mg i.m. every 3 weeks); oestrogen plus progestogen (e.g. Prempak-C) • Fertility: specialist treatment with gonadotrophins • Consider growth hormone replacement

Ablative treatment for pituitary tumours

17

Transfrontal surgery inevitably damages the normal pituitary and is generally ineffective in producing cure of large tumours. Transsphenoidal surgery, in good hands and performed via the midline approach, is effective in debulking large tumours and removing small ones, effecting a cure in more than half of the latter. Conventional external radiotherapy requires careful planning and dosing to avoid damage to the optic nerves; it is slowly effective but useful to contain aggressive tumours. The so-called 'gamma knife', coupled with sophisticated imaging, now offers the prospect of applying high-dose local radiation with great precision to pituitary tumours, but is not widely available. Specific types of hormone-secreting tumour and their management are considered below. It is important to note that compressive symptoms due to a prolactinoma respond rapidly to dopaminergic agonists, and an urgent prolactin should be requested before attempting surgery on a newly diagnosed pituitary tumour.

ISOLATED GROWTH HORMONE DEFICIENCY AND RELATED SYNDROMES GH deficiency only produces clinically obvious manifestations when it occurs in childhood. Although GH is secreted from early fetal life, it is not until infancy that its deficiency leads to impaired growth. The prevalence is about 1 in 10000.

Aetiology In most cases of isolated, non-iatrogenic GH deficiency the cause is unknown, but is believed to be due to a hypothalamic abnormality. There are, however, several genetic syndromes resulting in pituitary aplasia or hypoplasia, isolated GH deficiency or deletion of the GH gene. Patients in the latter group in particular may respond to injected GH as if it were a foreign protein and develop antibodies to it. In some cases of GH deficiency it may become apparent after GH therapy is initiated that TSH synthesis is also impaired; GH therapy accelerates the metabolism of TSH. Laron dwarfism is the term given to a syndrome that mimics GH deficiency, but in which GH levels are high and IGF-1 levels low. It is caused by a mutation in the GH receptor, which becomes unresponsive to GH. Some cases of short stature may be due to failure to respond to IGF1. In Pygmies the short stature and other somatic features appear to be due in part to low levels of IGF-1.

Clinical features Abnormally low rates of overall growth, leading to short stature, are associated with typical somatic abnormalities. In particular, the face appears baby-like, with persistence of a small nose, poorly developed nasal bridge and small chin.

901

The dentition may be delayed, with later crowding of teeth. Muscle development is poor, with excess subcutaneous fat, thin hair and skin, a small larynx and a high-pitched voice. Puberty is typically delayed until 18-20 years, when there is a small growth spurt. Even in patients with complete GH deficiency longitudinal growth occurs, but at only about one-third of the normal rate. There is a tendency to hypoglycaemia, which may be the most obvious feature in the neonate. GH deficiency developing in adult life may be associated with increased mortality and the value of GH therapy is becoming established (see below).

Management The diagnosis of GH deficiency is established by a combination of clinical observation (including serial measurements of growth and assessment of growth velocity) and provocative testing with clonidine, glucagon, arginine or, less frequently, insulin injection. Measurement of IGF-1, especially when combined with IGF-binding protein (IGFBP3) estimation, and assessment of the GH response to exercise or sleep are useful screening tests. If GH deficiency is found, CT or MR imaging of the pituitary is needed to exclude a pituitary or hypothalamic tumour (particularly craniopharyngioma), and the remainder of anterior pituitary function should be assessed. GH therapy is then initiated. GH isolated from cadaver pituitaries has been associated with some cases of Creutzfeldt-Jakob disease, leading to its withdrawal; biosynthetic GH is now available, although expensive. Treatment is with 0.5-0.7 units/kg/week (14-20 units/m2 body surface area weekly) as daily s.c. injections. Biosynthetic IGF-1 is necessary in Laron dwarfism. The response is monitored by frequent measurements of height, and calculation of growth velocity (Fig. 17.12, p. 898). There is an initial 'catch-up' period, followed by more steady growth. GH treatment improves growth in children with chronic renal failure and Turner's syndrome, but its use in idiopathic short stature (with GH deficiency) or in other conditions causing short stature is not generally recommended.

Growth hormone deficiency in adults The recent availability of recombinant GH has opened up the possibility of GH replacement in adults with GH deficiency. It seems likely that adult GH deficiency contributes to the increased risk of premature cardiovascular disease in hypopituitarism and causes increased fat mass, reduced muscle mass, low bone density, abnormal lipoprotein metabolism and impaired psychological wellbeing. GH deficiency should be tested for in patients with hypothalamic-pituitary disease, including GH deficiency in childhood, or in patients who have had cranial irradiation. 1

902

MCQ 17.6

2

Figs 17.6-17.10

Partial GH deficiency exists but can be difficult to distinguish from other causes of low GH levels, such as ageing. Short-term trials of GH therapy have shown modest improvements in body composition, exercise capacity and quality of life, although not all patients appear to benefit. Transient oedema is a frequent side-effect, but can be minimized by reducing GH dosage. Patients show variable sensitivity to GH; the usual starting dose is 0.15-0.3mg/day. Monitoring is performed using IGF-1 levels and simple anthropomorphic measurements. Absolute contraindications include active malignancy, intracranial hypertension and proliferative diabetic retinopathy. GH replacement is currently expensive and further work is required to establish the optimum treatment regimen, to identify which patients will benefit most from GH and to ensure that adverse side-effects (e.g. glucose intolerance) are avoided.

Growth retardation in other conditions It is often difficult to distinguish partial idiopathic GH deficiency from other states in which GH or IGF-1 secretion or action is secondarily impaired. These include psychological deprivation, hypothyroidism, growth failure due to inadequate nutrition (e.g. malabsorption due to coeliac disease) and chronic liver and kidney disease. Simple delayed puberty may also present with growth failure. In teenage boys with delayed puberty and short stature it is often justified to give a 3-6-month course of treatment with androgens (p. 943). The resulting growth spurt results in part from the androgens and in part from the augmentation of nocturnal GH secretion that they cause. 1

GROWTH HORMONE EXCESS The term acromegaly refers to the syndrome of GH excess when it develops in adult life. Gigantism is when its onset in childhood also leads to excessive height. At the turn of this century, Harvey Cushing recognized that acromegaly and gigantism were due to acidophil tumours of the pituitary gland that produce GH. He was the first clinician to distinguish features due to hyperpituitarism (hormone excess) from those due to hypopituitarism. Clinical GH excess is almost always due to a pituitary adenoma but, rarely, ectopic secretion of GHRH by a range of neuroendocrine tumours may lead to acromegaly by stimulating the pituitary gland to produce GH. Indeed, it was recognition of this syndrome that led recently to the extraction and purification of pancreatic-tumour-derived GHRH, which was rapidly shown to be present in identical form in the hypothalamus. Acromegaly is rare, with an annual incidence of 3 per million; gigantism occurs in less than 5 % of patients with acromegaly.

Aetiology The factors leading to the development of a GH-secreting adenoma are still uncertain. Most such tumours are mono-

clonal, suggesting a pituitary origin rather than excessive stimulation by the hypothalamus as their cause. Occasionally, mixed tumours secrete PRL and other hormones besides GH. Normally GH release is stimulated by dopamine, acting via GHRH stimulation. Paradoxically, the dopaminergic agent bromocriptine usually has a modest but significant inhibitory effect on GH secretion. In some cases, especially in individuals with mixed GH- and PRL-secreting tumours, this effect may be of therapeutic benefit. Somatostatin inhibits GH secretion, a phenomenon which has led to the use of the synthetic analogue octreotide.

17

Clinical features The clinical features of acromegaly and gigantism are due to a combination of GH excess, a local space-occupying lesion and other hormone deficiencies. Except where the tumour is large and growing rapidly, the latter usually occur late in the disease as a result of attempted treatment (radiotherapy or surgery), especially transfrontal surgery. Gigantism Gigantism is the most obvious and serious consequence of GH excess in childhood and adolescence. Normally, GH secretion occurs at a low level throughout childhood and increases with the rise in sex hormones at the time of puberty. With a GH-secreting tumour in childhood, however, inappropriate GH secretion occurs out of phase with the stage of puberty. The growth of cartilage at the epiphyseal plates is greatly stimulated due to the action of IGF-1 (p. 894). Normally this effect is held in check by an inhibitory action of sex hormones on the epiphyseal cartilage stem cell, ultimately leading to epiphyseal closure. The first feature of gigantism is therefore the onset of an inappropriate, continuing and excessive growth spurt, coupled with increased hand and foot size and general skeletal enlargement. Later, other features (increased skin and soft-tissue growth) become evident. In childhood, as in adult life, the pituitary tumour is often large at the time of presentation and may lead to local pressure effects. These include headache (owing to pressure on the diaphragma sellae) and visual field defects owing to pressure on the optic chiasm (pp. 892, 896) (see Figs 17.14 and 17.15). If pubertal failure results, either as a direct effect of a large tumour or as a result of treatment, epiphyseal closure will be delayed and the ultimate height attained even more extreme. Acromegaly There is a highly variable time between the onset of symptoms and signs and the diagnosis of GH excess in adults. This is dependent, among other things, on chance and the index of suspicion of any doctor who encounters the patient. In general, the greater the experience of the doctor, the more likely he or she is to make a confident and

FIG. 17.16 Typical appearances of acromegaly A Face showing enlarged supraorbital ridges, nose, lips and jaw. B 'Spade-like' hands. 0

correct diagnosis early. Clinical features due to the local tumour include headache and visual field defects (most commonly bitemporal hemianopia). Features due to GH excess may occur in many organs, but the most obvious involve the soft tissues, skin, and skeleton. Together, these give the classic acromegalic fades (Fig. 17.16A), with prominent supraorbital ridges, a large nose, a prominent lower jaw (prognathism) and thickening of the skin and soft tissues, including the lips and tongue. There is often a typical rather deep voice, due to thicken-

903

SUMMARY 4 Clinical features of acromegaly • Symptoms and signs are due to local tumour, growth hormone excess and deficiency of other hormones. • Prognathism, enlarged soft tissues, skeletal deformity and nerve entrapment occur. • Hypertension and diabetes mellitus contribute to mortality. • IGF-1 is a useful screening test. • Diagnosis is confirmed by failure of GH suppression after oral glucose.

ing of the vocal cords, enlargement of the sinuses and changes in other tissues in the nasopharynx and larynx. Dental malocclusion results from protrusion of the lower jaw, and the teeth may become more widely spaced or dentures may need to be changed. The wide 'spade-like' hands are caused by increased thickness and volume of skin and soft tissues and growth of periarticular cartilage (Fig. 17.16B); the feet are similarly affected. Ring, glove, shoe and hat sizes increase. In long-standing and severe acromegaly major skeletal deformities are common. Severe thoracic kyphosis and rib elongation lead to a barrel chest. In adult-onset acromegaly the growth of articular cartilage and soft tissues in and around a joint commonly leads to secondary osteoarthritis, which may be compounded by necrosis of articular cartilage. Neurological problems include carpal tunnel syndrome and muscle weakness. The age-related mortality rate of patients with untreated acromegaly is probably twice normal, largely attributable to cardiovascular disease hypertension, cardiac muscle hypertrophy, cardiomyopathy and ischaemic heart disease. The latter is particularly likely in acromegalics with diabetes mellitus. Overt diabetes due to insulin resistance occurs in about 10% of acromegalics, and about 50% have impaired glucose tolerance. Respiratory problems also contribute to the increased mortality. The prevalence of colonic polyps is increased in acromegaly, but it is unclear yet whether these confer an excess risk of malignancy. A common early complaint is increased sweating, owing to an increase in both number and activity of sweat glands. Macroglossia may lead to sleep apnoea, owing to obstruction of the airway at night, often presenting with daytime somnolence. Enlargement of the liver, spleen, pancreas and kidneys seldom causes problems, although the glomerular filtration rate may be increased because of the latter. About one-quarter of patients have a goitre, which is commonly multinodular. Patients with acromegaly have quite often been labelled as neurotic because many of the symptoms are rather vague. Impotence is a common complaint and only rarely

1

904

MCQ17.7

Minutes FIG. 17.17 GH response to oral glucose load in a normal subject (suppression) and a man with acromegaly (no suppression or paradoxical rise)

appears to be due to hypogonadism. Women commonly have menstrual irregularities, and some are hirsute. Galactorrhoea due to accompanying hyperprolactinaemia occurs in about one acromegalic woman in five. In some this is due to a mixed PRL- and GH-secreting tumour, in others to damage to the pituitary stalk. 1

Investigation Despite classic clinical features biochemical confirmation of the diagnosis is mandatory. Early signs of disease may be subtle and 'acromegaloid' features may occur without GH excess. GH suppression test. Baseline GH levels are of limited value in diagnosis because GH is normally secreted episodically. The definitive test remains the oral glucose load (50-75 g), with accompanying measurements of GH (Fig. 17.17). Normally, GH suppresses to below 4mU/L (2ug/L) within 2 hours after a 50 g oral glucose load, probably as an effect of central inhibition of GHRH secretion. In acromegaly levels either fail to suppress or show a paradoxical rise. About 30-40% of cases also show a paradoxical rise of GH in response to TRH or GnRH given intravenously, presumably because the tumour possesses receptors for these hormones, but these tests are not routine. Bromocriptine suppresses GH in the majority of acromegalics, though it is only clinically useful in a minority. An indication of its potential usefulness in an individual case is obtained from measurement of GH hourly for 4-6 hours after 5mg bromocriptine orally. Octreotide is now commonly used for GH suppression. A test dose (50-100 ug s.c.) with hourly GH measurement is a useful predictor of longer-term response. IGF-1 assay. Measurement of circulating levels of IGF1, especially combined with IGF-BP3, provides a useful screening test for acromegaly and is valuable in monitoring the effects of treatment. Levels do not fluctuate as with

17

CASE STUDY 17.1 DIAGNOSIS OF ACROMEGALY A 56-year-old woman presented with a gradual increase in tiredness over several months, accompanied by sweating and a change in her facial appearance. Friends had commented on the latter, and she had observed that she was developing a bigger nose and jaw. Comparison with photographs that she brought to the clinic confirmed that her face had indeed changed appearance over a 10-year period. Further inquiry revealed that her glove size had also increased, rings which she had previously been able to wear were now unwearable because of the increased size of her fingers, and her shoe size had increased from 6 to 71/2. She had put on 4 stones in weight in 2 years and had left-sided headaches. There was also a history of back pain and general stiffness and aches in the fingers. On examination, apart from the changes to the face and hands, the skin thickness was increased, as assessed by taking a pinch of the skin over the dorsum of the hand. The visual fields were full and the fundi were normal. The cardiovascular examination revealed that the blood pressure was 170/106, but there was no evidence of cardiac failure. There

were no other abnormalities on examination. Question 1, How would you investigate this woman to confirm the diagnosis of acromegaly? Discussion This lady has typical signs and symptoms of acromegaly. The only real differential diagnosis for the soft tissue changes in the extremities and face is pseudo-acromegaly, which can occur in insulin-resistant states, but the other features make this most unlikely. Visual field defects are only present in around 5% of acromegalics, whereas headache can occur in around 60%. Approximately 75% of acromegalics report arthralgia, due to osteoarthrosis secondary to bony overgrowth. Hypertension occurs in 25-50% of patients and is a major contributor to mortality from cardiovascular disease. If left untreated there is ventricular hypertrophy and impaired ventricular function, caused by the effects of growth hormone on the heart and muscle. To confirm the diagnosis of acromegaly, the gold standard is the

GH, and are almost always elevated in acromegaly and gigantism. Physiological increases in IGF-1 occur in pregnancy and puberty. GHRH assay. Ectopic secretion of GHRH is rare and usually causes pituitary hyperplasia, but this is difficult to distinguish from an adenoma radiologically and the diagnosis of a GHRH-secreting tumour is often retrospective. The diagnosis can be confirmed by measuring circulating GHRH and imaging the tumour (usually in the pancreas, lung, thymus or adrenal). Radiological investigations are given on pages 898-899.

use of the glucose tolerance test. This consists of giving 75 g of oral glucose and measuring the circulating growth hormone level over a 2-hour period. Normally growth hormone levels become undetectable (<2 ug/L) during this test, but in acromegaly the growth hormone is not suppressed and may even rise. In this case the basal growth hormone was 54ug/L and rose to 168ug/L 60 minutes after oral glucose. Simultaneous measurement of the glucose during this test will reveal glucose intolerance in around 50% of patients and 5% will have frank diabetes. Measurement of IGF-1 is a useful screening test, as elevated levels are strongly suggestive of acromegaly: in this patient the IGF-1 level was 723 ng/mL (reference range 123-463). The remainder of anterior pituitary function should be evaluated, and in this case was normal. The next step in the diagnosis is an MRI or CT scan of the pituitary to identify the pituitary tumour, which in this case was a 7 mm microadenoma. Treatment consists of surgical removal. Cure rates after transsphenoidal adenomectomy are 70-80%, and indeed this patient was cured by the procedure.

but the most widely used criteria are a mean serum GH level of below 2.5ug/L on repeated (3-5) samples during the day, normal IGF-1 or a normal response to glucose. The lower the GH level after treatment, the lower the mortality.

Management

Surgery Surgery (p. 901) is the usual first-line treatment. With intrasellar microadenomas (diameter < 1cm), the best centres now achieve at least 70% cure rates by transsphenoidal resection, without causing hypopituitarism, although the balance between endocrine benefit and penalty shifts unfavourably with increasing tumour size.

The aim of treatment of GH excess is to arrest and, where possible, reverse the local and systemic effects of the tumour. Except in the growing child, correction of GH excess is seldom urgent. Visual field defects, however, present an immediate and compelling indication for surgery. There is no consensus on what constitutes a cure,

Drug therapy Bromocriptine or cabergoline cause tumour shrinkage in only a minority of cases of acromegaly, but may be helpful when surgery is not thought desirable (e.g. in mild disease or the elderly) or when GH is effectively suppressed by these drugs. The usual doses needed are 10-30 mg of

905

bromocriptine daily in four doses or 0.5-2 mg cabergoline. Additional benefit with higher doses of bromocriptine has been reported. The clinical effect often exceeds the biochemical response, possibly because of a direct suppressive effect on IGF-1 secretion. The somatostatin analogue octreotide is expensive but more effective than bromocriptine. It is given in a dose of 50-100ug s.c. three times daily. It is especially helpful for headache in acromegaly and in patients who are not cured by other treatments. Octreotide causes mild (<20%) tumour shrinkage in about 40% of the cases treated. A long-term complication is the development of gallstones. Two long-acting Somatostatin analogues, lanreotide and Sandostatin-LAR, have recently been introduced which avoid the need for daily injections.

• Surgical removal of the pituitary tumour is first-line treatment, usually via the transsphenoidal route. • If tumour removal is inadequate, consider further surgery (microadenoma) or radiotherapy (macroadenoma). • Somatostatin analogues or, less successfully, dopaminergic agents control GH secretion when surgery is contraindicated or while waiting for radiotherapy to work.

Radiotherapy (p. 901) External radiotherapy can be used to produce tumour regression and is now usually used to treat persistent disease after surgery. It requires careful planning and megavoltage machinery. The treatment is remarkably safe but its effect is delayed, with a gradual and progressive benefit over 2-10 years. Octreotide may be useful while awaiting the full effects of radiotherapy. It is important that the total radiation dose does not exceed 45 Gy and that individual daily doses do not exceed 2 Gy, to avoid damage to surrounding tissues.

occurs as a result of selective enzymatic cleavage within the cells' hormone-containing granules; the relationship of the principal cleavage products is shown in Figure 17.9 (p. ••). It will be apparent that ACTH, B-LPH and B-endorphin are secreted simultaneously in response to CRH and other 'physiological' stimulants, such as insulin-induced hypoglycaemia. The principal hormonal action of ACTH is to stimulate cortisol secretion, particularly from the zona fasciculata. Its action is via cell surface receptors which activate adenylate cyclase. Receptor binding requires extracellular calcium. ACTH secretion is under synergistic control from several hypothalamic factors, shown in Table 17.4 (p. 893). Circadian rhythm (which affects both pulse amplitude and frequency), stress and food all modify ACTH secretion via their effects on these hypothalamic factors.

Complications of treatment

Isolated ACTH deficiency

The complications of surgery and radiotherapy vary greatly with the skill and technique of the operator. After surgery by the transsphenoidal route panhypopituitarism is quite rare, except where radical removal of a large tumour has been attempted, but isolated hormone deficiencies are more frequent. Postoperative diabetes insipidus is usually transient. If anterior hypopituitarism is suspected in the recovery period, full replacement therapy should be initiated and combined pituitary function tests conducted about 6 weeks postoperatively. After external radiotherapy the onset of hypopituitarism should be suspected if the patient complains of lassitude, lethargy and/or impotence. Regular follow-up with pituitary function tests should in any case be undertaken, and replacement therapy initiated as appropriate. 1

Isolated ACTH deficiency is very uncommon and may be difficult to diagnose. Patients are usually middle-aged at the time of presentation, with weakness, lethargy, hypoglycaemia and sometimes hyponatraemia. The diagnosis is confirmed by a low urinary free cortisol excretion, and low plasma ACTH and cortisol levels which do not rise normally in response to insulin hypoglycaemia or a short synacthen test. The defect may result from an autoimmune response against corticotrophs and may be associated with other organ-specific autoimmune disorders. Treatment is with glucocorticoid replacement.

ACTH DISORDERS Physiology ACTH is the most important hormone synthesized and secreted by the corticotrophic cells. Processing of POMC to produce the hormone ACTH (39 amino acid residues)

1

906

SUMMARY 5 Management of GH excess

MCQ 17.8

ACTH excess ACTH excess due to an ACTH-secreting adenoma leads to pituitary-dependent Cushing's syndrome (p. 932). 'Physiological' ACTH excess occurs when the adrenal glands fail, owing either to an inborn error of secretion or to destruction by disease. Under these circumstances the normal episodic and circadian fluctuations in ACTH secretion occur at much higher than normal levels. Rarely, hereditary unresponsiveness to ACTH, due to a mutation in its receptor, may cause glucocorticoid deficiency and very high ACTH levels. Nelson's syndrome is the name given to the hyperpigmentation, ACTH excess and pituitary expansion that may follow bilateral adrenalectomy for Cushing's syndrome due to a pituitary adenoma. It is presumed to result from removal of negative feed-back inhibition of a semi-

autonomous pituitary adenoma that was previously subject to some restraint by the high circulating levels of cortisol. Radiotherapy to the pituitary is used to prevent or treat Nelson's syndrome.

PROLACTIN DISORDERS

Physiology PRL is a polypeptide hormone of 199 amino acid residues, although dimeric and polymeric forms exist whose biological significance is uncertain. Secretion of prolactin PRL is synthesized and secreted by cells of the pituitary (lactotrophs), which are mainly located laterally in the gland (Fig. 17.18). Unlike other anterior pituitary hormones, PRL is secreted in the absence of stimuli, and the predominant influence of the hypothalamus is to restrain this by the secretion of dopamine. Besides causing TSH release,TRH also stimulates PRL secretion. PRL secretion is episodic and increases at night and with stress (e.g. venepuncture or breast examination), which must therefore always be excluded as a simple cause of mild hyperprolactinaemia. In females, PRL levels increase at the time of puberty. Androgens probably have a modest inhibitory effect.

In pregnancy, PRL levels rise about fivefold. Oestrogens, probably together with progesterone, cause hyperplasia of the lactotrophs and enhance PRL secretion while inhibiting its effect on the breast epithelium. The decline in sex hormones after delivery, coupled with the stimulus of suckling, is responsible for postpartum lactation. Postpartum, breastfeeding causes a marked increase in PRL secretion, as a result of a reflex involving sensory fibres from the breast, the dorsal nerve roots and the hypothalamus. This is probably important in maintaining milk production for weeks and months after delivery, when baseline PRL levels have declined to normal. Another reflex arc is responsible for suckling-induced oxytocin secretion, which is important in stimulating myoepithelial cells of the breast and so causing milk ejection from the ducts.

17

Action of prolactin and other hormones on the breast PRL has an action on the breast that is essential for lactation (Table 17.5, p. 895). Oestrogens promote ductal growth, whereas PRL and progesterone promote growth of the lobules and alveoli. Progesterone enhances the synthesis of B-lactalbumin, a non-catalytic subunit that converts the function of the enzyme galactose synthetase to lactose synthesis. Other hormones with important effects on the breast include insulin and cortisol (which are permissive or facilitatory), GH and placental lactogen. The decline in oestrogen and progesterone levels after delivery releases the brake on PRL-stimulated lobuloalveolar milk production.

Pathology PRL deficiency can occur with pituitary damage (Table 17.6, p. 896) and results in the failure of lactation. However, because of the unique predominance of inhibitory control over PRL secretion, damage to the pituitary stalk results in PRL excess as a result of diminished dopamine delivery, which contrasts with the failure of other anterior pituitary hormones, owing to loss of the appropriate hypothalamic releasing hormones. PRL excess. Hyperprolactinaemia and associated disorders are common, particularly in women. The low incidence in men is a reflection of the inhibitory effects of androgens both on the lactotrophs and on the breast.

HYPERPROLACTINAEMIA Aetiology

FIG. 17.18 Regulation of prolactin secretion

Unlike other anterior pituitary hormones, prolactin is secreted in the absence of stimuli and is normally under inhibitory restraint by dopamine.

The causes of hyperprolactinaemia are best considered in relation to PRL physiology. Table 17.7 lists the principal causes, most of which lead to only mild to moderate hyperprolactinaemia (levels up to 5-10 times normal). Pituitary tumours may lead to hyperprolactinaemia for one of two reasons: they may interfere with the pituitary stalk and so produce local dopamine deficiency, or they

907

TABLE 17.7 Classification of hyperprolactinaemia Cause

Example

Pituitary lactotroph excess Physiological Drug-induced Pathological

Pregnancy Oestrogens Tumour microprolactinoma macroprolactinoma mixed GH and prolactin tumour

Inhibition of dopamine secretion Thoracic sensory nerve stimulation Neurogenic Chest wall burns/incisions Suckling/nipple stimulation Pituitary stalk damage

Stalk section Stalk/hypothalamic tumour

Drug-induced

Methyldopa, reserpine, opiates, H2-receptor blockers

Inhibition of dopamine action Neuroleptic drugs Phenothiazines, e.g. chlorpromazine Butyrophenones, e.g. haloperidol Antiemetics

Metoclopramide

Endocrine disorders

TRH-mediated, e.g. primary hypothyroidism Miscellaneous, e.g. adrenal insufficiency, Cushing's syndrome, polycystic ovary syndrome

Impaired excretion

Renal failure

may secrete PRL. Pure prolactinomas are classified into two broad categories: the microprolactinoma, which is less than 1 cm in diameter and is associated with mild to moderate hyperprolactinaemia and no pressure symptoms, and the macroprolactinoma. In many instances, the diagnosis of microprolactinoma is presumptive, based on exclusion of other causes.

Clinical features and management The unwanted secretion of milk (galactorrhoea) is common in women, but in only about one-third of cases is it due to PRL excess. If treatment is required dopamine agonists (see below) may be effective, even when PRL levels are normal. Increased secretion of PRL from any cause has an inhibitory effect on cyclical ovarian function, probably

908

1

MCQ17.9

4

MCQ 17.11

2

MCQ 17.10

3

Case 17.1

owing to direct inhibition of GnRH secretion in hypothalamic neurons. Erratic and anovulatory cycles or amenorrhoea result. Treatment with a dopaminergic drug such as bromocriptine results in a drop in PRL and the resumption of normal ovulatory cycles. This restores normal oestrogen levels, which prevent premature osteoporosis and cardiovascular disease in women. It should be remembered that the commonest cause of hyperprolactinaemic amenorrhoea is, of course, pregnancy. 1

Microprolactinoma Clinical features Around 5-10% of the adult population have a microprolactinoma, but few of these ever come to medical attention. Classic presentations are with erratic menstrual cycles or amenorrhoea, with or without galactorrhoea. In some patients infertility or decreased libido is the main complaint, usually associated with amenorrhoea (primary or secondary). Frequently, presentation follows cessation of the pill (post-pill amenorrhoea) or pregnancy. Patients with microprolactinoma never present with visual field defects or other pressure effects. Indeed, the presence of these effects associated with a modest PRL excess indicates dopamine deficiency due to a non-PRL-secreting tumour in the pituitary stalk or above (e.g. a craniopharyngioma). Investigation Diagnosis depends upon demonstration, in several serum samples, of hyperprolactinaemia in the absence of other causes of PRL excess. The effect of stress can be eliminated by sampling 1 hour after the insertion of a cannula. A CT or (preferably) MR scan is also required; however, not all microadenomas visible on such scans are necessarily the source of PRL excess, and some microprolactinomas may be too small to visualize (see Fig. 17.15). The main purpose of scanning is to rule out stalk compression from a nonPRL-secreting tumour. Management Less than 5% of microprolactinomas enlarge over a 5-year period and repeated annual imaging is all that is required to assess tumour growth. If there has been no progression over 1-2 years, monitoring PRL levels will suffice thereafter, as growth will be indicated by rising PRL levels. Any change in size warrants treatment. Other indications for treatment are to restore fertility, to induce regular menses (and thus prevent the consequences of low oestrogen levels) and to treat decreased libido or dyspareunia (other symptoms of oestrogen deficiency). Medical treatment is with dopaminergic agents. Bromocriptine has been the most widely used but may initially cause nausea and postural hypotension. It should be started at a low dose (1.25 mg at night) with food, and increased gradually over 1-2 weeks to 2.5mg three times daily with food. Cabergoline has fewer side-effects and is now the drug of choice; the usual starting dose is 0.25 mg at night twice a week. In most cases this is highly effective

both in suppressing PRL and in resolving the amenorrhoea and galactorrhoea. If pregnancy is desired it may be sensible to advise contraception for 6 months to allow a reduction in adenoma size. In contrast to the situation with acromegaly, PRL-secreting tumours commonly shrink in size with dopaminergic agents. Although it may be safely continued, bromocriptine is usually stopped once the pregnancy test is positive, and at present cabergoline should be discontinued prior to conception once regular menses are resumed; PRL remains low for several months after treatment has ceased. With a microprolactinoma the risk of tumour expansion during pregnancy, sufficient to cause visual field defects, is probably negligible. It is as yet uncertain for how long dopaminergic therapy should be given in such cases. A reasonable policy, if pregnancy is not desired, is to treat for 1-2 years and then to repeat PRL levels several weeks after stopping the drug. Sustained normal levels would suggest that the tumour has regressed and that treatment can be safely stopped and the patient observed. If contraception is desired by a patient with microprolactinoma, bromocriptine or cabergoline may be combined safely with the contraceptive pill. Surgical removal (p. 901) of microprolactinomas gives a long-term cure rate of 60-70% in the best series. It has been used for patients intolerant of dopaminergic agonists or who do not wish to take a prolonged course of drugs. 2

Macroprolactinoma The definition of macroprolactinoma by size is arbitrary, and any non-pregnant patient with circulating PRL levels in excess of 10 times the upper limit of normal (over 5000 mU/L, 200ng/mL) probably has a more aggressive prolactinoma which is potentially or actually large and expanding. Virtually all detected prolactinomas in men and about one-fifth of those in women fall into this category. Clinical features All features are likely to be more extreme than with a microprolactinoma. These include the endocrine consequences of PRL excess: galactorrhoea, amenorrhoea and, in men, impotence. The latter is due principally to gonadotrophin (and hence testosterone) deficiency, suggested also by lethargy, depression and loss of secondary sexual hair. Rarely, presentation may be with infertility and oligo- or azoospermia. Local pressure effects - headache, visual field defects and other cranial nerve lesions - are often present at the time of diagnosis. As with other pituitary tumours, the occurrence of panhypopituitarism in the untreated patient indicates the presence of a large and destructive tumour. In males, galactorrhoea may be present even in the absence of gynaecomastia. Investigation PRL levels are 10-200 times normal. CT or MRI will show the full extent of the tumour. Formal plotting of the visual fields is required if scans reveal that the optic chiasm is being approached by the tumour.

Management The urgency of treatment depends upon the severity of local pressure effects. A therapeutic trial of bromocriptine (2.5-5 mg three times daily) or cabergoline (0.25-0.5 mg twice weekly), with regular measurement of visual fields and acuity, is the initial treatment of choice, followed by transsphenoidal decompression if visual failure persists and there is no tumour shrinkage on repeat scans. It is impossible to eradicate macroprolactinomas surgically. Lifelong dopamine agonist treatment is needed unless the tumour shows spontaneous remission (owing to infarction). Radiotherapy is only used if medical treatment fails to suppress prolactin or to induce tumour shrinkage over the long term. The response to treatment can be monitored by gradually tailing off the drugs and measuring PRL levels. 0

17

GONADOTROPHIN DISORDERS Gonadotrophin disorders are considered in more detail under the section on gonads (pp. 941-945). Gonadotrophin-secreting tumours are rarely diagnosed clinically. Most published cases have been FSH-secreting tumours. 'Functionless' pituitary tumours often stain for LH and FSH or their subunits, and these non-secreting gonadotrophinomas present with hypopituitarism or pressure effects. Treatment is usually transsphenoidal removal, with radiotherapy for residual tumour. Ectopic secretion of gonadotrophins appears to be confined to hCG, which is a normal product of the placenta and a tumour marker for hydatidiform mole and its malignant counterpart the choriocarcinoma (a tumour which can now be cured by intensive chemotherapy). Germ cell tumours of the testis, ovary and other sites (e.g. central nervous system) often secrete hCG, as may some carcinomas (e.g. adenocarcinoma of the lung). This may lead to excess testicular oestrogen secretion and gynaecomastia. In children, hepatoblastomas may produce pseudoprecocious puberty by hCG secretion. 4

TSH DISORDERS TSH disorders are discussed mainly under panhypopituitarism (p. 897) and thyroid disorders (pp. 914-918). Iso-

SUMMARY 6 Management of hyperprolactinaemia • Microprolactinomas are usually treated with dopaminergic agents, such as cabergoline 0.25-0.5mg twice weekly. • Surgical removal of microadenomas is an alternative in patients intolerant of these drugs or who prefer surgery. • Macroprolactinomas are best treated with dopaminergic agents, even when associated with compressive symptoms.

909

lated TSH deficiency is a very rare cause of secondary hypothyroidism. Primary TSH excess due to a thyrotrophinoma is a rare cause of hyperthyroidism. Surprisingly, TSH levels by immunoassay may be within normal limits in both conditions, which are recognized respectively by the low or high free T4 levels. However, the biological activity of the TSH is abnormal, accounting for the clinical picture. Patients with long-standing severe primary hypothyroidism (myxoedema) often develop pituitary enlargement owing to thyrotroph hyperplasia.

Osmotic and related stimuli Osmotic pressure Atrial natriuretic peptide Angiotensin II Osmoreceptors

Non-osmotic stimuli Arterial/venous pressure 'Stress' Hypoglycaemia ±

Thirst centre

Supraoptic nucleus

+ Volume receptors

Drinking behaviour Paraventricular nucleus

FURTHER READING ON THE ANTERIOR

PITUITARY GLAND Carroll P V, Christ E R and the members of Growth Hormone Research Society Scientific Committee 1998 Growth hormone deficiency in adulthood and the effects of growth hormone replacement: a review. J Clin Endocrinol Metab 83: 382-395. Colao A, Lombardi G 1998 Growth-hormone and prolactin excess. Lancet 352:1455-1461. Consensus guidelines for the diagnosis and treatment of adults with growth hormone deficiency: summary statement of the Growth Hormone Research Society Workshop on adult growth hormone deficiency. 1998 J Clin Endocrinol Metab 83: 379-381. Duncan E, Wass J A H 1999 Investigation protocol: Acromegaly and its investigation. Clin Endocrinol 50: 285-293. Lamberts S W J, de Herder W W, van der Lely A J 1998 Pituitary insufficiency. Lancet 352:127-134. Melmed S, Jackson I, Kleinberg D, Klibanski A 1998 Current treatment guidelines for acromegaly. J Clin Endocrinol Metab 83: 2646-2652. Orme S M, McNally J Q, Cartwright R A, Belchetz P E for the United Kingdom Acromegaly Study Group 1998 Mortality and cancer incidence in acromegaly: a retrospective cohort study. J Clin Endocrinol Metab 83: 2730-2734. Vance M L, Mauras N 1999 Growth hormone therapy in adults and children. N Engl J Med 341:1206-1216.

Magnocellular neurons ADH release (Ca2+dependent) Renal free water reabsorption FIG. 17.19 Factors stimulating ADH secretion by osmotic and non-osmotic mechanisms Osmoreceptors also activate the thirst centre to stimulate drinking.

nuclei as part of a larger precursor, which in each case also includes the sequence of another protein (a neurophysin). Axonal flow carries the normal precursors to the specialized nerve terminals in the posterior pituitary gland, where they are stored in cytoplasmic granules. Neuronal activation leads to the release of the hormone and its neurophysin into the hypophyseal capillaries and hence into the general circulation.

Control mechanisms THE POSTERIOR PITUITARY GLAND The posterior lobe of the pituitary gland consists principally of the secretory terminal extensions of neurons that arise in the preoptic nucleus of the hypothalamus. Its blood supply is derived principally from the hypophyseal arteries, rather than from the hypothalamopituitary portal circulation.

SYNTHESIS AND SECRETION OF THE MAIN HORMONES

910

The posterior pitutary gland secretes two principal octapeptide hormones, vasopressin and oxytocin. In humans the vasopressin produced is arginine vasopressin (AVP), which is also called antidiuretic hormone (ADH). Both hormones are synthesized in specialized (magnocellular) neurons in the supraoptic and paraventricular

ADH secretion is stimulated by an increase in osmolality, and by 'stress' stimuli, such as volume depletion (Fig. 17.19). Glucocorticoids exert an inhibitory effect on its release and also interact with ADH at the renal tubule, so that in the presence of glucocorticoid insufficiency diabetes insipidus may be masked. ADH acts on two different kinds of receptors, those on the renal tubules (type I) and those on blood vessels (type II). The action on the former is mediated by cyclic AMP; ADH opens up pores in the basolateral membrane for water and urea. Oxytocin acts on the smooth muscle of the uterus and mammary ducts (to cause milk ejection). It also causes contraction of myoepithelial cells of the seminiferous tubules.

ADH DEFICIENCY Deficiency of ADH causes the syndrome of diabetes insipidus (see also Ch. 21, pp. 1101-1103).

Aetiology ADH deficiency may be caused by trauma (head injury or neurosurgery); primary or secondary neoplasms of the pituitary, pituitary stalk or hypothalamus; inflammatory disorders such as sarcoidosis; eosinophilic granuloma; and bacterial meningitis and viral encephalitis. It may be part of a familial syndrome of diabetes insipidus, diabetes mellitus, optic atrophy and nerve deafness (DIDMOAD), or occur as an isolated autosomal dominant disorder. ADH deficiency is idiopathic in around a third of cases; autoimmunity may account for some of these. In addition, the condition may be simulated by disorders that lead to renal resistance to ADH. These include X-linked nephrogenic diabetes insipidus (confined to males), and hypercalcaemia and hypokalaemia, both of which lead to a water diuresis from resistance to ADH.

Clinical features The principal symptoms of ADH deficiency are thirst and incidental polyuria, as the patient is forced by the osmotic stimulus to drink large volumes of water or other fluids. It is not uncommon for patients to pass in excess of 10 L of urine a day, of very low specific gravity (1.003 or less, osmolality less than l00mmol/kg). The serum sodium concentration is generally moderately raised, but seldom over 150mmol/L, provided the patient has free access to water. The major danger is when such access is denied, or when the patient is unconscious, e.g. due to an anaesthetic or head injury. Under these circumstances the osmolality and sodium concentration rise rapidly. A serum sodium concentration over 160-170 mmol/L may lead to severe and irreversible brain damage. As with hyponatraemia, correction should be gradual.

Investigation The major differential diagnosis is from psychogenic polydipsia and from other causes of thirst and polyuria (see Table 20.10, p. 1045). The mainstay of diagnosis is measurement of serum and urine osmolality, and the diagnosis is suggested by a urine osmolality of less than 200mmol/kg or specific gravity of 1.005 or less. Plasma osmolality is usually above 287mmol/kg. Confirmation of the diagnosis usually depends on the water deprivation test. Free access to water is allowed up to the start of the test at 8 am, when the patient is weighted and blood and urine samples are taken for osmolality. Timed hourly urine collections, serum osmolality measurements and body weight recordings are made. For safety, the test is discontinued if 3% of body weight is lost or 3 L of urine are passed. At the end of 8 hours, if urine concentration has not occurred 2ug of the analogue DDAVP is given i.m. and further samples are taken; in true diabetes insipidus this leads to prompt urine concentration. Further details are given in Chapter 21, page 1101. In equivocal cases the failure of plasma ADH to rise after

infusion of hypertonic saline may be useful, but assays for this hormone are not widely available.

17

Management of diabetes insipidus The long-acting vasopressin analogue desamino-D-arginine vasopressin (DDAVP, desmopressin) acts almost exclusively on the V2 receptors of the renal tubule, and so does not cause hypertension and constriction of the splanchnic circulation (mediated by smooth muscle V1 receptors). It is also resistant to enzyme degradation in the circulation or peripheral tissues, and has a greatly prolonged half-life. DDAVP is usually given once or twice a day intranasally (10-20 ug) or, for the postoperative patient, intravenously (1-2 ug). Recently an oral formulation of DDAVP has been introduced and many patients prefer this to intranasal DDAVP; the maintenance dose is 100-200 ug three times a day. With the introduction of desmopressin diabetes insipidus is no longer a problem of management, provided the patient has an intact thirst centre and does not drink excessive volumes of water. Chlorpropamide enhances the renal response to ADH, but is only useful in partial forms of diabetes insipidus.

'INAPPROPRIATE' SECRETION OF ADH Aetiology The syndrome of inappropriate secretion of ADH is frequently encountered in a variety of different conditions (Ch. 21, p. 1103). These include acute and chronic pulmonary disease, congestive cardiac failure, small cell carcinoma of bronchus, cerebral disease, liver cirrhosis, cortisol deficiency and hypothyroidism. ADH secretion is a transient accompaniment of surgery for 24-48 hours. Of the above conditions, only small cell carcinoma of bronchus is ever associated with true ectopic tumour synthesis and secretion of vasopressin. Even in this condition, however, hyponatraemia is often caused by excess hypothalamopituitary ADH secretion, rather than an ectopic source.

Clinical features, investigation and management The hallmark is a low serum sodium concentration (often below 125 mmol/L) and serum osmolality (below 270mmol/kg), associated with an inappropriately high urine osmolality (usually greater than the plasma osmolality), with persistent urinary sodium excretion (>20 mmol/L), in a patient who is not clinically volume depleted. Volume depletion from any cause is a stimulus to ADH secretion, so these biochemical abnormalities would be appropriate in, for example, a patient overtreated with diuretics, a common cause of hyponatraemia in the elderly. Similarly, in states of hypervolaemia (e.g. cardiac failure, cirrhosis, nephrotic syndrome, inappropriate i.v. therapy) there is hyponatraemia simply due to the excess total body water, and this does not constitute the syndrome of inap-

911

propriate ADH secretion. In these states urine sodium loss is low (<10mmol/L). The major clinical features are confusion, coma and fits (water intoxication) if the serum sodium falls much below 120mmol/L (osmolality <250mmol/kg). Treatment is with fluid restriction (l000mL/day) and correction of the underlying cause. Cautious treatment with small volumes of hypertonic (1.8%) sodium chloride is indicated in patients with CNS manifestations to restore the serum sodium to 120-125 mmol/L: rapid increases above this risk central pontine myelinolysis and other CNS damage. Mild cases may respond to lithium or demeclocycline, which diminish the renal response to ADH. For further details, see Chapter 21, page 1107. 1

FURTHER READING ON THE POSTERIOR PITUITARY GLAND Baylis P H 1995 Investigation of suspected hypothalamic diabetes insipidus. Clin Endocrinol 43: 507-510. Hirshberg B, Ben-Yehuda A 1997 The syndrome of inappropriate antidiuretic hormone secretion in the elderly. Am J Med 103: 270-273. Kumar S, Berl T 1998 Sodium. Lancet 352: 220-228. McKenna K, Thompson C 1998 Osmoregulation in clinical disorders of thirst appreciation. Clin Endocrinol 49:139-152.

THE THYROID GLAND EMBRYOLOGY, ANATOMY AND PHYSIOLOGY The thyroid gland is derived in the first trimester from an outpouching of the floor of the pharynx which becomes displaced caudally and fuses with the fourth pharyngeal pouch. During its caudal displacement it leaves the thyroglossal duct, which usually becomes fragmented and resorbed, but may persist to form thyroglossal cysts. If migration is incomplete ectopic thyroid tissue may lie at the base of the tongue. In adults the thyroid gland weighs about 20 g and lies in front of the thyroid cartilage, its two lobes connected by a narrow isthmus. It moves upwards with the thyroid cartilage on swallowing. Histologically, the thyroid gland is composed of follicles of varying diameter lined by a single-layered epithelium and arranged into lobules of 20-40 follicles, each served by a common blood supply. The gland is highly vascular, being supplied by two pairs of thyroid arteries which arise from external carotid (superior) and subclavian (inferior) arteries. There is a rich lymphatic drainage and adrenergic and cholinergic innervation. The gland also contains parafol1

912

MCQ 17.12

2

licular or C cells, which lie next to the follicle but do not abut the colloid. These cells, which arise from the last pair of pharyngeal pouches, secrete the calcium-lowering hormone calcitonin. The glycoprotein thyroglobulin (TG), which is rich in thyroxine residues, is the major constituent of colloid. The thyroid epithelial cells synthesize and secrete TG into the colloid. Later, they resorb it, cleave the thyroid hormones from the glycoprotein backbone and secrete them into the circulation. TG is both an integral part of the process of thyroid hormone synthesis and a major storage form of iodine and thyroid hormones within the thyroid.

Synthesis and secretion of thyroid hormones The thyroid gland contains about 90% of the body's total iodine, mainly in organic form. The thyroid cells actively transport the iodide into the cells by a process which is competitively inhibited by other anions of similar size, such as perchlorate. Iodide is then oxidized to iodine and incorporated into mono- and diiodotyrosine (MIT and DIT) residues on TG by the enzyme thyroid peroxidase (TPO). Favoured residues at particular sites on the TG molecule are then converted to thyroxine (T4) and triiodothyronine (T3) residues (principally the former), either by interaction with other DIT and MIT residues or by interaction with an oxidation product of DIT and MIT themselves. This process of 'organification' of iodine (which requires the action of TPO) takes place principally at the apical membrane of the thyroid cell, as TG is being formed and secreted into the colloid (Fig. 17.20). lodination is most active when iodide levels are low. TSH is the principal active regulator and exerts its action on a cell-surface receptor, mainly via adenylate cyclase. TSH accelerates iodoprotein synthesis and the resorption of colloid by a phagocytic process, followed by fusion with lysosomes. TG is then hydrolysed, facilitated by reduction of disulphide bonds in the molecule. lodotyrosine deiodinase removes iodine from liberated MIT and DIT, and this iodine returns to the iodide pool for iodination of more thyronine residues.

MCQ 17.13

FIG. 17.20 Thyroid hormone synthesis and release (see text for details).

17

FIG. 17.21 Structure of T4, T3 and reverse T3 R is the alanine residue (-CH2-CHNH2-COOH).

Circulation, action and metabolism of thyroid hormones T4 is the major form of thyroid hormone secreted by the thyroid; it is converted enzymatically to either T3 or reverse T3 (rT3) (Fig. 17.21), of which only the former is biologically active. Regulation of the activity of the relevant deiodinases, in the liver and in other tissues, is important both in health and disease. Generally, the production of T3 and rT3 is reciprocally controlled. T4 is a prohormone which becomes active by virtue of its conversion to T3. T4 is, however, the principal circulating form of thyroid hormone and is more than 99% bound to three plasma proteins thyroxine-binding globulin (TBG), thyroxine-binding prealbumin (TBPA; also called transthyretin) and albumin. TBG has the highest affinity but the lowest capacity, followed by TBPA. T3 binds with lower affinity to TBG, also binds to albumin, but not at all to TBPA. In their bound states T3 and T4 are not biologically active. It is the concentration of unbound (or free) hormone that determines the level to which the target cells are exposed. The liver is the major site of degradation. Although much of the iodine is recycled some is inevitably wasted, being excreted principally in the urine as iodinated degradation products and inorganic iodide. Under normal circumstances this is matched by a daily intake of iodide of about 150-300 ug in non-iodine deficient areas. The normal circulating level of T4 (50-150 nmol/L) exceeds that of T3 (1-3 nmol/L) by 50-fold; the difference in their unbound concentrations is about 10 times less than this, because of both the lower binding of T3 to TBG and its failure to bind to TBPA.

Physiological changes in thyroid hormones Physiological changes in thyroid hormones are shown in Figure 17.22. The fetus derives a small yet potentially important amount of T4 but no TSH from the mother. During the second trimester the fetal pituitary-thyroid axis becomes active. Up to the time of birth fetal hypothy-

FIG. 17.22 Relative changes of serum total T4, T3, reverse T3 and thyroxine-binding globulin (TBG) in different physiological states and in illness.

roidism does not greatly impair fetal development unless the mother is also hypothyroid, in which case severe neurological damage results. This typically occurs in areas of iodine deficiency, causing endemic cretinism. Although T4 is secreted in the fetus, very little T3 is produced before birth, rT3 being the major product. After birth there is an immediate switch of hepatic metabolism of T4 from rT3 to T3, so that levels of T3 rise briskly during the first week of life. This change may well be important in adaptation to postfetal life. TBG levels in the fetus are low, and rise after birth. T3 levels throughout infancy and childhood tend to be higher than those of the adult. In pregnancy total T4 and T3 levels rise passively in response to an oestrogen-induced rise in TBG. A similar change is produced by oestrogen therapy. In a wide variety of conditions of stress and ill-health, hepatic production of T3 declines and that of rT3 increases. Exposure to extreme cold is associated with reduced production of rT3 and increased production of T3. 2

PATHOLOGY OF THYROID DISEASE The thyroid gland may be the subject of defective development, inborn biosynthetic errors, toxic or nutritional damage, autoimmune disease, and benign and malignant tumours. Many of these abnormalities give rise to enlargement of the gland and to the appearance of a goitre (thyroid enlargement). This may be focal (as in adenoma, cyst or carcinoma) or generalized, in which case it may be either multinodular or diffuse and smooth. Clinical assessment alone will often allow an accurate assessment of the cause of the goitre.

Autoimmune disease and the thyroid (see p. 889) The thyroid is the endocrine gland most commonly affected by autoimmune disease. Immune mechanisms are involved

913

in the pathogenesis of Graves' disease, 1 Hashimoto's thyroiditis and primary myxoedema. For reasons that are as yet unclear, these disorders are about five times more common in women than in men. Similar mechanisms may underlie these conditions, which frequently occur together in some families. Patients occasionally progress from Graves' disease to autoimmune hypothyroidism or vice versa. A combination of genetic, environmental and hormonal factors is involved in pathogenesis, some of which may also account for the frequent associations with other autoimmune disorders. In Graves' disease the TSH receptor is the critical antigen to which antibodies are produced. These antibodies can bind to the receptor, causing stimulation. Antibodies may also be produced to TG and TPO. The factors which determine what antibodies are produced in an individual case are uncertain. Measurement of circulating levels of relevant antibodies may be of diagnostic value. In general, however, the presence of antibody is rather non-specific. The most commonly measured antibodies (which are of the IgG class) are those against TG and TPO, which are present in about 10% of apparently healthy individuals, usually women. Antibody-positive subjects have a higher than normal chance of future thyroid autoimmune disease. TPO antibodies are present in 85% or more of patients with autoimmune thyroid disease. Very high litres indicate extensive lymphocytic infiltration and Hashimoto's thyroiditis. Antibodies to the TSH receptor Antibodies to the TSH receptor are of importance in the pathogenesis of Graves' disease and hypothyroidism. Thyrotoxicosis in Graves' disease is due to the presence of antibodies which bind to and activate the TSH receptor, thereby simulating the action of TSH. These are now best detected in assays using cells transfected with recombinant TSH receptor and determining adenylate cyclase activation. The assay is generally used only for research purposes but has clinical value in pregnant patients with Graves' disease. These antibodies can cross the placenta and cause thyrotoxicosis in the newborn infant. Measurement of TSH-receptor-stimulating antibodies at the beginning of the third trimester in the mother allows the detection of babies at risk. Other antibodies bind to the TSH receptor and block it, resulting in hypothyroidism. In most cases of myxoedema, however, the thyroid dysfunction is due to the destructive effect of cell-mediated immunity and cytotoxic antibodies.

Nutrition and thyroid disease There is a wide geographical variation in the incidence of goitre, which is only partly understood. Goitre is said to be

1

914

Fig. 17.11

2

MCQ 17.14

endemic when more than 10% of children in a population are affected: approximately 200 million people have endemic goitre worldwide. The major factors in endemic goitre are dietary iodine deficiency and the ingestion of goitrogenic substances that interfere with either the uptake or the organification of iodine. For example, cassava, a root crop eaten extensively in Africa, is rich in goitrogens. Women are particularly susceptible to goitre because of the demands of the fetus for iodine. Endemic cretinism is a neurological disease that afflicts the offspring of goitrous mothers in parts of the world where iodine is deficient, especially in the mountainous regions of New Guinea. It is completely prevented by providing the mother with iodized salt.

INVESTIGATION OF THYROID DISEASE Examination of the thyroid gland The thyroid should be carefully examined by observation from the front, and by palpation from behind while the patient swallows. The consistency of the gland should be assessed. In Graves' disease, the gland is typically diffusely enlarged. The swelling may be firmer with a carcinoma and associated with enlargement of local lymph nodes. Loss of mobility on swallowing is another sign suggesting malignancy. A midline swelling that moves up with protrusion of the tongue is likely to be a thyroglossal cyst or ectopic thyroid. A bruit over a goitre indicates a vascular gland and usually points to Graves' disease.

Laboratory assessment of hypo- ana hyperthyroidism Rodioimmunoassay The laboratory assessment of hypo- and hyperthyroidism is now extremely simple, thanks to specific radioimmunoassays for T3 and T4, to a specific and sensitive immunoradiometric assay for TSH, and (where relevant) to methods of assessing thyroid hormone binding. In most cases clinical assessment combined with a single measurement of serum T4 and TSH levels is sufficient. Serum TSH is a useful screening test, as a normal level rules out primary thyroid disease, but an abnormal level does not necessarily indicate clinical thyroid disease (Table 17.8). Estimation of the T4 and, sometimes, T3 level is therefore required to confirm the presence of thyrotoxicosis or hypothyroidism. In primary hypothyroidism T4 levels are low and TSH rises markedly. T3 remains normal for a long time in some patients owing to peripheral conversion of T4, and need not be measured. In the early stages, or in patients who have had a subtotal thyroidectomy, T4 and T3 levels are normal and TSH modestly raised. It is a matter of clinical judgement whether or not to treat such patients; at the least, regular follow-up is required. In most cases of symptomatic hypothyroidism TSH will be more than four times the upper limit of normal.

TABLE 17.9 Features of the sick euthyroid syndrome

TABLE 17.8 Causes of an abnormal TSH level Cause Elevated TSH

Suppressed TSH

Free thyroid hormone level

Clinical hypothyroidism Subclinical hypothyroidism (risk of future hypothyroidism) Transient: non-thyroidal illness (recovery phase) Pituitary tumour secreting TSH Thyroid hormone resistance syndrome

Low

Clinical thyrotoxicosis Recently treated thyrotoxicosis Multinodular goitre (risk of future thyrotoxicosis) Thyroid-associated ophthalmopathy (risk of future thyrotoxicosis) Pituitary failure Severe non-thyroidal illness First trimester of pregnancy Excessive thyroid hormone replacement

Elevated Normal

Normal Normal Elevated Elevated

Normal Normal Low Normal or low Normal Normal or raised

In secondary (pituitary) hypothyroidism one would expect TSH to be undetectable, but this is frequently not the case; T4 levels are low. In thyrotoxicosis serum TSH is low and T3 and T4 levels are elevated. In mild cases serum T3 may be raised before T4 (T3 toxicosis), and so is a more reliable test for early detection of thyrotoxicosis. Protein-binding tests Tests of protein binding have been widely used until recently. Oestrogens promote the hepatic synthesis of TBG along with other binding proteins; thus, the situation most commonly encountered, where the T4 is raised and protein binding has to be taken into account, is in a woman during pregnancy or taking the contraceptive pill. The test long used to elucidate this problem is an in vitro thyroid hormone uptake test, in which TBG in dilute serum and sephadex compete to bind radiolabelled T3. In most centres, however, this has been superseded by direct assay of the free T4 and free T3 levels. Although this is generally reliable, anomalies can rarely occur. Any thyroid function test may be misleading in the 'sick euthyroid' syndrome (Table 17.9). 2 Radioactive uptake tests The main isotope to be used diagnostically is 99mTc (as radiolabelled pertechnetate), a short-lived gamma-emitter with a half-life of 6 hours. Scans are of value in assessing the distribution of uptake within the gland, particularly when palpation is difficult. Uptake may be confined to one area in a functioning autonomous adenoma, as uptake by

17

• Occurs in many non-thyroidal illnesses, including trauma, infection, myocardial infarction, malignancy, inflammation, starvation, psychiatric illness. • The severity of the thyroid hormone changes correlate with mortality. • Patients are believed to be euthyroid, despite abnormal hormone levels, which are due mainly to the effects of cytokines. • The most common form of the sick euthyroid syndrome is a fall in serum T3; T4 levels fall with more severe illness. • TSH levels are low or undetectable in some patients with the most severe illness; TSH levels may rise in others, or during recovery. • Treatment with thyroid hormones has no proven benefit in the sick euthyroid syndrome, • Routine thyroid function testing in all sick patients is far more likely to uncover the sick euthyroid syndrome than unsuspected thyroid disease; thyroid function testing should be reserved for those in whom thyroid disease is suspected clinically. • Repeated thyroid testing after recovery demonstrates normal thyroid function and is useful if there is doubt about the diagnosis.

the rest of the gland is suppressed, or may reveal multiple areas of high uptake in toxic multinodular goitre. Uptake is usually absent in non-functioning benign or malignant nodules, and failure of uptake cannot reliably distinguish between these. Ultrasound Ultrasound is useful for demonstrating the size and nodularity of a goitre. Although it can also reveal cysts, ultrasound cannot exclude malignancy in the solid tissue around a cyst. Fine needle aspiration Fine needle aspiration of the thyroid is a simple and increasingly used first-line test in the diagnosis of thyroid masses, particularly to exclude malignancy. Interpretation requires expertise on the part of the histopathologist. However, once the technique is mastered it greatly reduces the need for thyroid scans or ultrasound, and for surgery on benign nodules.

PRIMARY HYPOTHYROIDISM: CONGENITAL AND JUVENILE Hypothyroidism may be present before birth or may develop at any time thereafter.

Congenital hypothyroidism The causes of congenital hypothyroidism are listed in Table 17.10. Endemic cretinism is a severe developmental disorder occurring in areas of endemic goitre in children born to goitrous and iodine-deficient mothers. Severe neurological defects (deaf-mutism, spasticity and motor impairment) are associated with variable hypothyroidism and

915

TABLE 17.10 Causes of congenital hypothyroidism Non-goitrous Complete failure of thyroid development Ectopic thyroid Endemic cretinism

Selected recommendations for hypothyroidism

Pituitary or hypothalamic insufficiency Transplacental transfer of TSH receptor blocking antibodies (transient) Goitrous Maternal antithyroid drugs Endemic cretinism Iodine treatment of mother Defects in thyroid hormone biosynthesis Iodide transport, organification defects lodotyrosine coupling defect lodotyrosine dehalogenase (iodine recycling) defect Abnormal iodoprotein secretion Thyroid hormone resistance syndrome

goitre. In non-goitrous areas congenital hypothyroidism occurs with a frequency of around 1 in 5000 births, and is due mainly to failure of normal development of the thyroid. It is usually not clinically apparent at birth, but may be diagnosed reliably within 1-2 weeks by TSH assay on a blood spot dried on filter paper; routine screening programmes are now widely established. Suggestive clinical features include postmaturity, hypothermia, enlargement of the posterior fontanelle, neonatal jaundice and delayed passage of meconium. Failure to diagnose and treat early leads to permanent retardation of physical growth and mental development, which by late infancy becomes gross. The baby may have feeding difficulty, failure to thrive, a protuberant abdomen, dry skin and hair, and delayed eruption of teeth, and is slow to reach the usual milestones. Screening and early treatment have virtually eliminated these features.

Juvenile hypothyroidism Hypothyroidism starting in childhood results in mental slowness but not permanent mental retardation. Features seen are less obvious than in the myxoedematous adult, but growth and sexual development are retarded, often markedly. Delay in body development principally affects the long bones and the bones of the face, which therefore appears immature. Epiphyseal dysgenesis with 'stippling' of the epiphyses on X-ray is common, and epiphyseal closure is delayed. Puberty is usually delayed, although, 1

916

Fig. 17.12

2

SUMMARY 7 Summary of the Royal College of Physicians of London Working Party guidelines for management of hypothyroidism and hyperthyroidism

MCQ 17.15

• Patients with hypothyroidism only require specialist referral if aged less than 16 years, pregnant or postpartum, hypopituitary, or presenting a management problem. • The correct dose of thyroxine is that which achieves euthyroidism and restores symptoms; this is usually achieved with normalization of the TSH level. • Patients with subclinical hypothyroidism and positive thyroid antibodies should receive thyroxine; those without antibodies require follow-up. • If there is any doubt about the need for thyroxine therapy, this should be stopped and serum TSH and T4 measured 6 weeks later. • Once thyroxine dosage is established, TSH and T4 levels should be measured annually. Selected recommendations for hyperthyroidism • All patients require specialist referral. • Prolonged medical treatment with antithyroid drugs is generally only indicated in Graves' disease. • Radioiodine is indicated in most types of hyperthyroidism, but caution is needed in the presence of active Graves' ophthalmopathy. • Thyroid surgery, performed by an experienced surgeon, is an alternative method of treatment, especially in patients with a large goitre or because of a patient preference for this treatment. • All patients require an annual check of thyroid function after treatment.

paradoxically, precocious puberty may also occur. Replacement therapy results in some catch-up growth and in spontaneous puberty (if this was delayed).

PRIMARY HYPOTHYROIDISM DEVELOPING IN ADULT LIFE Hypothyroidism with onset in adult life is almost always due to autoimmunity or previous treatment of thyrotoxicosis. It occurs in 1-2% of women, with a 5-10-fold lower prevalence in men.

Clinical features The onset is usually gradual and the symptoms are often attributed incorrectly to increasing age. Increasing lethargy, slowness, memory loss and depression are often coupled with increasing weight (without eating more) and cold intolerance. The skin becomes dry and the hair 'lifeless'. Hair loss is a major concern in some women. Puffiness below the eyes, especially in the morning, is frequently noted. 1 Other early signs include slowness in answering questions, a malar flush and delayed relaxation of the tendon reflexes (e.g. ankle jerks, if present, or biceps jerks). The voice may be husky or 'gruff. There is usually a relative or absolute bradycardia. Symptoms are referred to

131

SUMMARY 8 Clinical features of myxoedema • Symptoms include lethargy, increased weight, cold intolerance, mental slowing. • Signs include dry skin, puffy face, gruff voice. • Complications include serous effusion, psychosis, coma.

TABLE 17.11 Autoimmune polyglandular syndromes (APS)

Type 1 APS

Type 2 APS

I therapy) and a family history of thyroid disease should be sought. Although the complete picture is easy to recognize, the aim should be to make the diagnosis before overt thyroid failure develops, and serum T4 and TSH estimation allows this to be done. The commonest conditions with which hypothyroidism is confused are obesity and depression. With good clinical practice, gross myxoedema should be a thing of the past. In doubtful cases, thyroid function tests (TSH will suffice for screening) should be measured on a serum sample. 2

Main features*

Less frequent features

Autosomal recessive; rare

Alopecia and vitiligo

Investigation

Chronic mucocutaneous candidiasis

Ovarian failure

Hypoparathyroidism

Hypothyroidism

Addison's disease

Enamel hypoplasia Nail dystrophy

Autosomal dominant with incomplete penetrance; common

Alopecia and vitiligo

Autoimmune hypothyrodism

Coeliac disease

Graves' disease

Myasthenia gravis

TSH is raised and total and free T4 are low. T3 may remain normal for a long time. Investigations to establish the aetiology depend on the clinical setting. In a child, a thyroid scan may demonstrate ectopic tissue and, if a goitre is present, a perchlorate discharge test may rarely uncover an organification defect (Pendred's syndrome, associated with sensory nerve deafness). This test depends on the failure of iodide to be used in thyroxine synthesis; preadministered radioiodine can therefore be discharged from the gland when perchlorate is given. TG and TPO antibodies point to an autoimmune aetiology. It is important to carry out investigations to determine the extent of damage that has resulted from the hypothyroidism in a child, such as X-rays of bones and endocrine tests to assess pubertal stage. Juvenile hypothyroidism is sometimes confused with panhypopituitarism, but TSH will be low in the latter and high in the former. An ECG is important in the elderly patient with possible ischaemic heart disease; T4 replacement should be started cautiously, possibly under B-adrenergic blockade. Measurement of serum electrolyte levels may reveal hyponatraemia due to inappropriate ADH secretion.

Type 1 diabetes mellitus

Chronic active hepatitis

Addison's disease

Sjogren's syndrome Dermatitis herpetiformis Autoimmune pituitary disease

* Patients have two or more of these conditions.

many systems; constipation is common and may be associated with abdominal distension; in the nervous system, symptoms of carpal tunnel syndrome and a complaint of dizziness are common. Deafness is a frequent symptom owing to fluid in the middle ear. Rarely, the patient develops cerebellar ataxia or myotonia. There may be a firm, often irregular goitre (Hashimoto's thyroiditis) or the thyroid may be atrophic and impalpable (primary myxoedema). Hypercholesterolaemia predisposes to vascular atheroma, and angina in the untreated patient is particularly important, as treatment may precipitate myocardial infarction. Effusions may occur into any serous cavity - pericardial, pleural, peritoneal or joint spaces. Mild anaemia of macrocytic type is quite common; megaloblastic anaemia suggests coincidental pernicious anaemia. Other autoimmune disorders are often associated (Table 17.11). Menorrhagia and galactorrhoea (with elevated PRL) are often seen, with amenorrhoea in severe cases. In severe myxoedema patients are grossly retarded and may become psychotic when treated; rarely, however, they can also paradoxically appear mentally hyperactive (myxoedema madness). Hypothyroidism predisposes to hypothermia; it may also cause hyponatraemia. If hypothyroidism is suspected, a past history of thyrotoxicosis (with surgery or

17

Management In treating marked hypothyroidism it is important to be certain before commencing thyroid hormone that the patient is not hypopituitary or hypoadrenal, as decompensation of the adrenal failure may occur. If in doubt, hydrocortisone replacement should also be given until ACTH and/or cortisol deficiency has been excluded. In mild cases the patient can be commenced forthwith on 50-100 ug T4 a day, the higher dose being used for patients with high TSH levels (4-5 times or more above normal). Doses should be increased or decreased by 25ug every 2-3 months based on TSH levels, until these are within normal limits. More frequent sampling of TSH can be misleading. Once the TSH level is normal, the need for lifelong treatment should be stressed and an annual estimate of TSH levels should be performed to ensure the correct level of T4 is maintained. A computerized recall scheme is usually used for this. In severe cases, especially if there is ischaemic heart disease, replacement should be more cautious, starting with 25 ug of T4 daily or on alternate days.

917

TABLE 17.12 Features of myxoedema coma • Myxoedema coma is caused by decompensation of undiagnosed or poorly treated hypothyroidism, almost always in the elderly. • Mortality is around 50% even with treatment. • The three key features are hypothermia (as low as 24°C), altered mental state (confusion, psychosis or coma) and a precipitating cause - infection, trauma, congestive heart failure, myocardial infarction, stroke, or drugs that depress respiration. • Prompt treatment is essential and should not await biochemical confirmation. • Non-specific laboratory test abnormalities occur: hyponatraemia, hypoglycaemia, hypoxia, hypercapnia, elevated creatinine phosphokinase. • Ventilatory support is often needed. • Other supportive measures include i.v. fluids (avoid hypotonic fluids with low serum Na+), glucose for hypoglycaemia and gradual warming with space blankets (external warming may produce cardiovascular collapse). • Treat any underlying cause; empirical broad-spectrum antibiotics and hydrocortisone are usually given at the outset. • Replace thyroxine either as a single 500ug dose by nasogastric tube or as an intravenous bolus if a suitable preparation is available. An alternative is triiodothyronine, 10ug every 4-6 hours, given intravenously or orally.

Myxoedema coma (Table 17.12) with hypothermia carries a poor prognosis. There is no clearly established optimum form of thyroid hormone replacement: a large oral dose of T4 may be given by nasogastric tube, or parenteral T3 or T4 can be administered.

Prevention Thyroid failure cannot be prevented, but much can be done to ensure its early detection and treatment. Thyroid hormone replacement is the simplest form of therapy so, for practical purposes, prevention is by early detection. Patients previously treated for thyrotoxicosis should be offered annual follow-up testing of thyroid function to detect early hypothyroidism. This also applies to euthyroid subjects found to have thyroglobulin or thyroid peroxidase antibodies. Borderline thyroid function tests may present a problem; in such cases the tests can be repeated after a 3-month month interval and, if thyroid failure is developing, TSH will rise progressively. 1

THYROTOXICOSIS Thyrotoxicosis is common, with a prevalence of 1-2% in women in non-iodine deficient countries; men have a 10-fold lower prevalence. Most patients have autoimmune thyroid disease (Graves' disease). Thyrotoxicosis due to multinodular goitre, or to a toxic adenoma, is less common.

1

918

MCQ 17.16

Figs 17.13-17.17

There is a wide clinical spectrum of severity of thyroid hormone excess, of the systemic manifestations of the disease, the size and vascularity of the thyroid gland, and the extent to which other systems, notably the extraocular muscles and the skin, are involved in associated autoimmune disorders. Toxic multinodular goitre usually develops after the age of 50 against a background of many years of non-toxic goitre. It is therefore more common in areas of past or present iodine deficiency. For reasons that are not clear, focal autonomous nodule formation develops in the enlarged thyroid (detected as areas of increased radioiodine uptake), with associated thyrotoxicosis which is generally relatively mild in degree and associated with suppressed TSH. In toxic adenoma, an autonomous nodule (follicular adenoma) develops and progressively enlarges. It secretes an increasing amount of thyroid hormone until TSH is completely suppressed and (generally mild) thyrotoxicosis develops. This usually happens only when the nodule reaches 2.5 cm in diameter or more. Neither this disorder nor toxic multinodular goitre is associated with ophthalmopathy. In Graves' disease, antibodies to the TSH receptor stimulate the thyroid gland in ways analogous to the normal action of TSH. The disease presents most commonly in the third and fourth decades of life. Hyperthyroidism may be induced in incipient cases of Graves' disease or toxic multinodular goitre by high doses of iodine or iodinecontaining drugs (the Jod-Basedow phenomenon). Hashimoto's disease, which is also autoimmune in aetiology, may rarely be associated with mild thyrotoxicosis in the early stages, as may subacute (viral) thyroiditis. Occasionally, thyrotoxicosis is due to exogenous hormone consumption.

Clinical features of thyrotoxicosis History Typically, the patient develops weight loss, increased appetite, heat intolerance and sweating, and becomes anxious, nervous, restless, depressed, irritable and difficult to live with. A fine tremor is common, and proximal myopathy may cause muscle weakness, especially in the elderly. In women, menstrual disturbances, especially oligomenorrhoea, are common. Increased gut motility may lead to diarrhoea. Palpitations are common and symptoms of heart failure are especially likely in the elderly, often precipitated by atrial fibrillation. A goitre may be noted and complained of, especially if the gland is large and vascular. It rarely obstructs swallowing or breathing unless it is retrosternal. To assess the patient quickly when thyrotoxicosis is suspected, a clinical checklist, such as that given in Table 17.13, is useful. Examination Most of the features of thyroid hormone excess are due to increased sensitivity to circulating catecholamines. Part of

TABLE 17.13 Symptoms and signs of thyrotoxicosis Mechanisms Primary autoimmune process (specific for Graves' disease)

Thyroid hormone excess (found in all forms of thyrotoxicosis)

Symptoms

Signs

Exophthalmos, discomfort Diplopia, blurred vision Lesions on legs

Exophthalmos, congestion Ocular palsies, papilloedema, Pretibial myxoedema (rare)

Neck swelling, dysphagia Palpitations* Breathlessness

Goitre, bruit

Heat intolerance* Sweating* Increased appetite Weight loss Diarrhoea* Weakness Nervousness and irritability* 'Staring' eyes*

Resting tachycardia* atrial fibrillation, other arrhythmias* Heart failure Warm and sweaty* Increased bowel sounds Malabsorption Proximal myopathy Inability to sit still Emotional lability* Tremor* Lid retraction and lag (also due to autoimmune process)

* Signs and associated symptoms (partially) blocked by B-blockers.

the levator palpebrae superioris is sympathetically innervated, and overactivity causes a staring appearance. This may be associated with signs of anxiety, nervousness, tremor, sweating, tachycardia and supraventricular arrhythmias (notably atrial fibrillation). Other common features include increased reflexes and a proximal myopathy. The signs of thyroid-associated ophthalmopathy and pretibial myxoedema only occur in thyrotoxicosis caused by Graves' disease, and are described later. Thyroid hormones also stimulate the metabolic rate, probably both directly by inducing a number of metabolic enzymes, and indirectly through catecholamine sensitivity. Heat intolerance and weight loss (in spite of increased appetite) result. Features that may occur less frequently include myopathy, cardiac failure, hypercalcaemia, raised liver enzymes and ferritin, gynaecomastia and infertility. Gynaecomastia and infertility probably result from metabolic alterations of androgen and oestrogen metabolism, and increased sexhormone-binding globulin (SHBG) levels, which lead to an increased ratio of free oestradiol to testosterone. Thyrotoxicosis in the elderly may have few clinical features and can even masquerade as depression (apathetic thyrotoxicosis). For reasons at present unclear, a common presentation of thyrotoxicosis (often otherwise mild) in oriental men is with acute hypokalaemic flaccid paralysis.

Natural history of thyrotoxicosis Graves' disease Although it is difficult to be certain, it seems likely that, before treatment became available, up to a quarter of patients with Graves' disease had a spontaneous remission of their thyrotoxicosis. One reason for remission is that about 15% of patients will spontaneously become hypothyroid if followed long enough. The reverse pattern - hypothyroidism progressing to hyperthyroidism - is a rare but recognized presentation of Graves' disease.

17

Toxic adenoma and multinodular goitre Usually, these conditions gradually progress until the patient becomes overtly toxic. Remission does not occur.

Thyroid-associated ophthalmopathy (Fig. 17.23) This condition has been given various names (dysthyroid eye disease, infiltrative or Graves' ophthalmopathy) and remains one of the most puzzling clinical features of endocrinology.

Aetiology Ophthalmopathy occurs in Graves' disease but not in other causes of thyrotoxicosis, and is also infrequently found with autoimmune hypothyroidism. Some patients present with eye signs months or years before thyroid disease is apparent. An autoimmune process involves the extraocular muscles and appears to be indirectly linked with that causing the thyroid disease. The eye muscles become infiltrated with lymphocytes, swollen and ultimately fibrosed. Later, there is involvement of the retro-bulbar fat in the inflammatory process. Patients with Graves' disease who smoke are most likely to develop eye signs, for unknown reasons. Clinical features 2 The symptoms of dysthyroid eye disease include discomfort, grittiness, swelling, redness and pain in the eyes which is often most marked on wakening (because of orbital congestion on recumbency and exposure of the cornea during sleep). The patient may have noticed the development of 'bulging eyes' or diplopia. Occasionally presentation may be with visual failure owing to optic nerve compression by the enlarged extraocular muscles. The signs of a dangerous rise in ocular and orbital pressure are obvious conjunctival or periorbital oedema, painful eyes, exposure keratitis, retinal venous congestion and papilloedema. These should lead to urgent specialist referral. Management In most cases the symptoms and signs are mild and settle spontaneously. Local use of artificial tears is helpful to relieve grittiness, together with simple eye ointment at night, when the eyelids may need to be taped shut if there is corneal exposure. Periorbital and conjunctival oedema

919

FIG. 17.23 Eye signs in thyrotoxicosis and thyroid-associated ophthalmopathy A Eye signs in thyrotoxicosis: (1) lid retraction; (2) lid lag; (3) exophthalmos with conjunctival oedema; (4) strabismus with diplopia due to paralysis/tethering. Arrows indicate direction of gaze. B CT scan through orbits in a case of exophthalmos. Note oedematous and enlarged extraocular muscles (*). C Patient with typical features of ophthalmopathy, including periorbital oedema, lid retraction and exophthalmos. As is sometimes the case, signs are asymmetrical.

are helped by sleeping propped up and by diuretics. Severe congestive ophthalmopathy, particularly associated with visual impairment, is uncommon but requires urgent treatment, usually with systemic steroids in high dosage (e.g. prednisolone, 60-80mg daily). Radiotherapy to the orbits is used as an alternative in some centres. In severe cases, if improvement does not occur, or if deterioration follows gradual reduction in steroid dosage, orbital decompression may save vision. This usually involves removal of the medial wall of the orbit. Later (but only after the condition has been stable for 6 months or more), surgery may be indicated for diplopia that cannot be managed by refractive means (e.g. use of prismatic spectacles) or to improve appearance. Treatment-induced hypothyroidism can lead to deterioration in the eyes because of myxoedematous congestion. There is some evidence that radio-iodine therapy (see below) may also worsen ophthalmopathy. Careful follow-up or even prophylactic steroids (prednisolone in reducing doses for 3 months) 1

920

Fig. 17.18

2 2MCQ17.17 MCQ 17.17

should be offered to patients with eye signs receiving radio-iodine.

Pretibial myxoedema and thyroid acropachy Pretibial myxoedema occurs in 1-5% of patients with Graves' disease, 1 almost always in association with eye signs. There is purplish infiltration of the dermis of the skin over the anterolateral aspect of the lower legs and sometimes elsewhere (Fig. 17.24), and the surface resembles the skin of an orange. Pathology shows deposition of mucin between collagen bundles. Pretibial myxoedema may present incidentally, for cosmetic reasons, or because the lesion is itchy. Thyroid acropachy is even rarer, occurring in around 10% of patients with pretibial myxoedema. There is finger clubbing and periosteal reaction in the bones of the forearm. 2

Management of thyrotoxicosis Only Graves' disease is cured by antithyroid drugs, and even then only 40-50% of patients have a permanent

TABLE 17.14 Features of thyrotoxic crisis

17

A rare complication of poorly treated or untreated thyrotoxicosis. Mortality is 100% when untreated and 20% with treatment. A precipitating cause of the acceleration in thyrotoxicosis can usually be identified - thyroid damage due to surgery or radio-iodine, or severe illness such as trauma, non-thyroid surgery, infection, stroke, pulmonary embolus, diabetic ketoacidosis or parturition. Features include severe hyperthermia, marked tachycardia, congestive heart failure, nausea and vomiting, and progression from agitation to coma. Initial treatment is propylthiouracil 300 mg every 4 hours by nasogastric tube or per rectum. One hour after the first dose of propylthiouracil, stable inorganic iodine is given, which acutely retards thyroid hormone release, e.g. Lugol's iodine 8 drops every 6 hours, or ipodate 0.5 g every 12 hours. Dexamethasone 2 mg every 6 hours inhibits thyroid hormone release and T4-T3 conversion. Propranolol 40-80 mg every 6 hours is used to control the tachycardia, but should be used with caution in patients with heart failure. Supportive measures include external cooling, intravenous fluids, and treatment of the precipitating cause.

FIG. 17.24 Marked pretibial myxoedema extending over the feet.

remission. Remission occurs because the drugs not only reduce thyroid hormone synthesis (by inhibiting TPO) but also have an immunomodulatory effect. Traditionally, younger patients with Graves' disease have been treated with antithyroid drugs (carbimazole or propylthiouracil). If the condition relapses they have been offered subtotal thyroidectomy, whereas patients over about 40 years of age have been given radioactive iodine therapy. Different strategies are used in different centres, many physicians now prescribing 131I as definitive treatment for young patients, and others giving carbimazole long term. As a general principle, surgery should not be undertaken, nor 131 I given, while the patient is grossly thyrotoxic. The patient should first be made euthyroid with antithyroid drugs because there is a risk of precipitating a thyrotoxic crisis or 'storm' with 131I or surgery when the patient is still thyrotoxic (Table 17.14). Drug therapy The major antithyroid drugs used are thiourea compounds, of which carbimazole is usually used in the UK and methimazole and propylthiouracil in the USA. They are generally well tolerated, with reversible skin rashes developing in only about 1 % of cases, usually during the first 6 weeks of therapy. Arthralgia and hepatitis are less common. A rare but serious complication is agranulocytosis, which cannot be predicted. Patients should be warned in writing to stop the drug if they develop a sore throat, mouth ulcers or a fever, and to have a white cell count urgently, treatment being resumed if the neutrophil count is normal. Carbimazole is given at first in a dose of 40-60 mg daily, in three or four divided doses because of its short half-life.

This is tapered off after 3-4 weeks to a maintenance dose of 5-30mg daily (the titration regimen). An alternative policy is to continue carbimazole at 40 mg daily and maintain euthyroidism by adding in T4, 100-150 ug daily (the block-replace regimen). The latter achieves maximum remission rates after 6 months, whereas 18-24 months are usually needed for equivalent rates with the titration regimen. Graves' disease most commonly relapses within a year of stopping treatment but can recur up to 15 years later, and so annual follow-up should be arranged. Relapse is almost inevitable with a large goitre or if a second course of antithyroid drugs is given for a recurrence, and in some centres such patients are offered long-term antithyroid drugs. B-Adrenergic blocking drugs are useful for symptomatic treatment during the first 4-8 weeks until euthyroidism is achieved. Usually, small doses (e.g. 10-40 mg of propranolol three times daily) are sufficient. In mild cases, Bblockers alone may be used while the effect of 131I therapy is awaited. Potassium perchlorate is now no longer used because of a high incidence of toxic reactions. Thyroid surgery The indications for thyroid surgery are not clear-cut and depend largely on patient preference. Patients should be rendered euthyroid medically for 3 months before surgery is undertaken (see above). Surgery is particularly useful in thyrotoxic patients with large goitres, retrosternal goitres or toxic adenomas (although the latter also do well with radio-iodine). It is the treatment of choice for toxic multinodular goitre where there are obstructive symptoms (p. 924).

921

SUMMARY 9 Management of hyperthyroidism • Carbimazole is usually the first treatment for patients of all ages. • B-Blockers are useful symptomatically for the first 1-2 months. • Surgery is considered for failed medical treatment, especially in children and adolescents and large goitres. • 131I is considered for older patients and for toxic adenomas. • All patients should have regular follow-up.

Pre- and postoperative management Potassium iodide is generally administered by mouth in a dose of at least 5mg daily (many surgeons use much more) for 7-10 days before operation. This blocks the release of thyroid hormones but also, more importantly, leads to a marked accumulation of colloid in the thyroid follicles and, partly as a consequence of this, reduces the vascularity of the gland. Laryngoscopic examination of the cords is often undertaken routinely before surgery. Most surgeons remove about seven-eighths of the gland, taking care to identify and preserve the parathyroid glands and, of course, the recurrent laryngeal nerves. Serum calcium levels should be checked 12 and 24 hours after surgery. Hypoparathyroidism is uncommon, often transient, and can be treated with alfacalcidol (initially lug daily). Other early complications include bleeding and laryngeal oedema; later, hypothyroidism or a recurrence of thyrotoxicosis can occur. Radio-iodine treatment This is the treatment of choice for toxic multinodular goitres and is being used increasingly for Graves' disease and toxic adenoma. There are many dosage regimens, based on the size of the gland, the age and health of the patient and, in some cases, on preliminary isotope uptake studies. None has proved superior. One practical, simple scheme is to give 200 MBq radio-iodine with a small goitre, 400-600MBq with a moderate-large goitre and 800 MBq in severe thyrotoxicosis, in elderly patients with heart problems or atrial fibrillation and with a nodular goitre. Antithyroid drugs should be stopped for a few days beforehand and can then be restarted 7-10 days afterwards, but are not required routinely. There is an acute effect in the first few weeks or months and a more prolonged effect thereafter. Around 10% of patients become hypothyroid within a year and 3-5% do so each year thereafter. Patients should be told of this complication and reassured that it is easily treated. There does not appear to be any increased risk of cancer after 131I therapy. The treatment is obviously contraindicated in pregnancy and breast-feeding, and pregnancy should be avoided for 4-6 months after treatment.

1

922

Figs 17.19,17.25

Radio-iodine may worsen existing thyroid-associated ophthalmopathy. As with other treatments, all patients should be offered regular follow-up, usually by means of a computerized recall scheme. Management of thyrotoxicosis in the elderly In the elderly the same principles of management of thyrotoxicosis apply, except that surgery is rarely indicated. When atrial fibrillation is present it is advisable to anticoagulate the patient or give aspirin (if there is no contraindication). Cardioversion may be undertaken if the patient is off digoxin and does not revert to sinus rhythm spontaneously. Treatment of toxic adenoma is either surgical excision or, particularly in the elderly or frail, 131I. Thyrotoxicosis and pregnancy A number of autoimmune conditions commonly improve during pregnancy, probably due in part to hormonal changes. Graves' disease often enters remission during pregnancy and relapses in the puerperium. It is important to avoid overtreatment with antithyroid drugs during pregnancy. In the fetus these can cause hypothyroidism, increased TSH secretion and thyroid enlargement (which may cause respiratory obstruction in the newborn). It is therefore the rule to avoid the block-replacement regimen and keep the mother on the minimum dose needed to maintain euthyroidism, and usually to discontinue the drug in the last trimester. If the level of thyroid-stimulating antibodies (which can cross the placenta) is high this may lead to significant thyroid stimulation in the baby and neonatal thyrotoxicosis. As the level of antibodies declines, so does the thyrotoxicosis. The infant is treated with B-blockers and carbimazole until the condition subsides spontaneously. Neonatal thyrotoxicosis can be predicted by measuring maternal levels of thyroid-stimulating antibodies in the last trimester and by fetal tachycardia and failure to thrive. All babies born to mothers with Graves' disease should have urgent thyroid function tests at birth (and 4-7 days later if the mother is on antithyroid drugs, as these may have crossed the placenta and masked neonatal thyrotoxicosis).

OTHER THYROID DISORDERS Simple diffuse goitre Simple diffuse goitre 1 with euthyroidism is common, especially in women, and declines with age. The aetiology is poorly understood. Most such goitres are TSH dependent and generally diminish on T4 in replacement doses, but treatment is not generally needed. Pseudogoitre is also common. Here, a normal gland is readily visible because it is high-lying, usually in a girl with a long neck.

17

CASE STUDY 17.2 WEIGHT LOSS AND TREMOR A 33-year-old nurse developed tremor and heat intolerance over 2 months. At the same time she noticed a 3 kg weight loss, but was attempting to diet anyway. Although she was continuing to work, she felt more tired than usual and her partner had commented on her irritability. She had recently returned to work, having had a baby 6 months previously. Her periods were regular and she had not experienced palpitations, but was occasionally aware of her heart 'racing'. There was no significant past medical history. Her aunt had been treated for hypothyroidism. She smoked 10 cigarettes a day, drank no alcohol and was taking no medication. On examination, she was slim and rather anxious. There was a fine tremor of her outstretched hands, the palms were warm and sweaty and the pulse was 92 in sinus rhythm. Examination of the neck revealed a small, firm, diffuse goitre which was mobile, non-tender and not associated with lymphadenopathy. There were no eye signs. The reflexes were symmetrically brisk and the remainder of the examination was normal.

Discussion The symptoms and signs are very suggestive of thyrotoxicosis. The main differential diagnosis is anxiety, but the heat intolerance, warm extremities and goitre all make this less likely. Phaeochromocytoma and other causes of weight loss are much less likely without other symptoms and signs. Once a diagnosis of thyrotoxicosis is suspected, the next step is to confirm the diagnosis of excess thyroid hormone production. This is achieved by measuring the serum TSH and free T4; measurement of free T3 is indicated if the TSH is suppressed and the free T4 is normal, to rule out T3 toxicosis. If thyrotoxicosis is confirmed, the cause must be established. The commonest cause is Graves' disease in a woman of this age, and she has a family history of thyroid autoimmunity, which would be compatible with Graves' disease, as would the character of the goitre. There is no thyroid-associated ophthalmopathy, which would clinch the diagnosis, but these signs are absent in 50% of cases. She has recently had a baby, and this suggests the alternative possibility of

postpartum thyroiditis. Against this diagnosis is the duration of symptoms: thyrotoxic symptoms rarely last more than 4 weeks and are not severe in postpartum thyroiditis. However, the postpartum period is also a time of increased risk for the development of Graves' disease, and there is therefore a need to distinguish between these two conditions. A 99mTc scan will reveal increased isotope uptake and confirm the diffuse goitre in Graves' disease; in postpartum thyroiditis the isotope is not taken up by the damaged gland. Thyroid peroxidase antibodies are found in both conditions and will not help to distinguish them, but the presence of these antibodies tends to rule out other causes of thyrotoxicosis. In this patient the TSH was <0.03 mU/L, FT4 35nmol/L and FT310.9 pmol/L. Thyroid peroxidase antibodies were positive and there was a diffuse goitre with elevated 99m Tc uptake after radionuclide administration. This confirmed the diagnosis of Graves' disease and treatment with carbimazole was started.

Questions 1, What are the most likely causes of this woman's symptoms? 2, What investigations are indicated?

Diffuse multinodular goitre

Management of multinodular goitre

Diffuse multinodular goitre is common, but poorly understood. In western countries, where dietary iodine is sufficient, new cases are now seldom due to iodine deficiency. The high prevalence in many underdeveloped parts of the world is due in part to the ingestion of goitrogens, e.g. cassava. Eventually in multinodular goitre, the nodules (which are initially TSH dependent) may become autonomous and thyrotoxicosis develops.

In the early stages, non-toxic goitre will often respond either to suppression of TSH with exogenous thyroid hormones, or to removal of the cause (goitrogens or iodine deficiency). Once autonomous nodules have developed, TSH suppression with T4 is much less effective, and iodine administration carries the risk of precipitating hyperthyroidism. The chances of a successful response are highest if TSH levels are not suppressed before giving T4.

923

RECENT ADVANCES SIMPLE GOITRE

THE TREATMENT OF

Diffuse non-toxic goitre is common and affects 5-10% of the population. Those affected are euthyroid. Women are three to four times more likely to develop a goitre, with a peak incidence between the ages of 20 and 30. In up to half of these subjects the goitre may disappear spontaneously, but in the remainder it may progress and cause cosmetic problems or even compressive symptoms. The aetiology is unknown, but a proportion of such goitres become multinodular, which may in turn cause hyperthyroidism. The traditional method of treatment is surgery or thyroxine treatment. Recurrence occurs in around 20% of goitres treated surgically. Thyroxine, in doses sufficient to suppress TSH levels, decreases goitre size modestly in two-thirds of patients but there is a risk of atrial fibrillation developing. Thyroxine is ineffective in patients whose TSH is already low, but is the treatment of choice in a goitre associated with a high TSH: in such patients the diagnosis is autoimmune hypothyroidism rather than simple goitre. Radio-iodine has recently been used to treat simple goitre. The mean reduction in goitre size is around 50% and hypothyroidism, in the short term, occurs in 50%. This treatment seems equally effective in non-toxic nodular goitre. Best results are obtained when the goitre is small. It therefore appears that radio-iodine is an effective treatment for simple goitre, in the absence of hyperthyroidism. Such treatment spares the patient the risks of surgery or suppressive thyroxine treatment, at the expense of hypothyroidism requiring replacement doses of thyroxine. Toxic multinodular goitre is usually treated with 131I therapy, though it is often relatively resistant and repeated doses may be needed. Patients should generally be rendered euthyroid with carbimazole first. Where the goitre is large or causing severe obstructive symptoms (especially pressure on the trachea), surgery is the treatment of choice, especially as 131I therapy may temporarily cause further thyroid enlargement. Tracheal compression is best confirmed by analysing flow-volume loops and by CT scanning. 1 Familial goitre There are a number of rare conditions where defects in thyroid hormone synthesis lead to goitre. Most are autosomal recessive. The commonest is Pendred's syndrome, caused by an iodine organification defect (as a result of mutation in the pendrin gene) and nerve deafness.

1

924

Fig. 17.20

2

Case 17.2

3

MCQ 17.18

Riedel's thyroiditis Riedel's thyroiditis is a rare condition of unknown aetiology associated with woody hard fibrous infiltration of the thyroid gland and surrounding tissues. Some patients have more widespread disease, including mediastinal and retroperitoneal fibrosis and sclerosing cholangitis.

Subacute (de Quervain's) thyroiditis 2 This commonly presents with acute painful generalized swelling of the thyroid gland, fever, and mild thyrotoxicosis followed by transient hypothyroidism. The ESR is raised and radionuclide uptake by the thyroid is low, in contrast to non-destructive types of thyrotoxicosis. Destruction by one of many different viruses leads to the release of stored hormone and later impaired thyroid function, accounting for this clinical picture. Treatment is with aspirin or, for moderate or severe disease, a short course of prednisolone, beginning at 40 mg daily and tapering over 6-8 weeks.

Localized thyroid swellings Apparently solitary thyroid nodules are common, affecting around 5% of the population, but only 1-2% of these are malignant. Over two-thirds of carcinomas are papillary (more in areas of high iodine intake), which have a very good prognosis, metastasizing via lymph nodes. Follicular carcinomas account for 10-15% of cases and also have a good prognosis, particularly in younger patients; bloodborne metastases occur especially in lung, liver and bone. Anaplastic carcinomas account for 5-10% of tumours and are commonest in the elderly. They are highly malignant and have a poor prognosis. All these tumours are derived from the thyroid epithelial cell. About 5% of thyroid carcinomas are of different origin, being derived from the calcitonin-secreting parafollicular C cells. These are the medullary carcinomas, associated with high circulating levels of calcitonin. They are sometimes familial and may be associated with multiple endocrine adenomatosis type 2 (see p. 939). Lymphoma of the thyroid is a rare but important complication of Hashimoto's thyroiditis. Clinical features are unhelpful in most patients, but malignancy should be suspected more strongly with a hard, fixed mass; rapid enlargement; lymphadenopathy; or a change in voice. Both a cystic lesion and a dominant nodule in a multinodular goitre (Fig. 17.25) have a reduced but still appreciable risk of malignancy compared to a solid lesion. Most centres now use fine needle aspiration biopsy in the initial investigation of isolated thyroid swellings, sometimes supplemented by thyroid scans or ultrasound. A toxic adenoma should be excluded by measuring TSH and free T4. If doubt remains, the nodule is removed surgically.

Management of thyroid cancer Where possible, thyroid carcinomas are treated surgically. 131 I therapy is of value if the carcinoma takes up the

RECENT ADVANCES IN POSTPARTUM THYROIDITIS (PPT)

17

PPT is defined as the appearance of thyroid dysfunction, usually based on biochemical evidence, within the year after delivery. It is due to transient autoimmune destruction. PPT is common, affecting 5% of women postpartum, and may be predicted by the presence of thyroid peroxidase (TPO) antibodies in the first trimester. At present it is unclear whether such screening is worthwhile, as the condition is usually mild and self-limiting and does not require treatment. However, recent evidence suggests that 10-20% of such women will develop permanent autoimmune hypothyroidism within 10 years, and that postpartum depression is more severe in women with TPO antibodies. Future studies will determine the value of screening all pregnant women. At present, patients with type 1 diabetes are known to have a threefold increased risk of PPT and should be screened: a high index of suspicion should be maintained for the emergence of thyroid dysfunction in other women during the postpartum period. FIG. 17.25 Dominant left-sided nodule in a multinodular goitre

Bodenner D L, Lash R W 1998 Thyroid disease mediated by molecular defects in cell surface and nuclear receptors. Am J Med 105: 524-538. Hermus A R, Huysmans D A 1998 Treatment of benign nodular thyroid disease. N Engl J Med 338:1438-1447. Le Moli R, Wesche M F T, Tiel-van Buul M M C, Wiersinga W M 1999 Determinants of longterm outcome of radioiodine therapy of sporadic non-toxic goitre. Clin Endocrinol 50: 783-789. Lindsay R S,Toft A D 1997 Hypothyroidism. Lancet 349: 413-417. Schlumberger M J 1998 Papillary and follicular thyroid carcinoma. N Engl J Med 338: 297-306. Vanderpump M P J, Ahlquist J A O, Franklyn J A, Clayton R N 1996 Development of guidelines for good practice and audit measures in the management of hypothyroidism and hyperthyroidism. Br Med J 313: 539-543. Weetman A P 1997 Hypothyroidism: screening and subclinical disease. Br Med J 314:1175-1178. Weetman A P 2000 Graves' disease New Engl J Med 343:1236-1248. The use of radioiodine in the management of hyperthyroidism 1995. Royal College of Physicians, London.

isotope, and is required for patients with residual and metastatic disease. Uptake only occurs in papillary and follicular tumours. If the patient has had a subtotal or near total thyroidectomy, the remaining normal thyroid is ablated with a high dose of 131I. Thyroid hormone replacement therapy is then started and continued for 6 weeks. It is then withdrawn to allow TSH levels to rise, following which an ablative dose of 131I is given to treat metastatic or persistent local disease. T3 is often used initially instead of T4, as less time is needed off treatment to achieve high TSH levels, but patients may none the less experience hypothyroidism. This can now be avoided by giving recombinant TSH to achieve the necessary levels for 131I uptake, but recombinant TSH is expensive. Treatment with 131I can be repeated as long as isotope scans show persistent metastatic or local disease. Measurement of circulating TG levels is useful in monitoring for relapse of differentiated carcinoma. Anaplastic cancers, and those that do not take up 131 I, are treated with radiotherapy to the local tumour. Results of treatment of metastatic disease with chemotherapy are disappointing. Lymphoma is treated with radiotherapy or chemotherapy. 3

ANATOMY AND PHYSIOLOGY

FURTHER READING ON THE THYROID GLAND

The gross anatomy of the adrenal glands is shown in Figure 17.26. The adult adrenal glands weigh 4-5 g each and are situated at the upper poles of the kidneys.

Amino N,Tada H, Hidaka Y, Crapo L M, Stagnaro-Green A 1999 Screening for postpartum thyroiditis. J Clin Endocrinol Metab 84:1813-1821. Bartalena L, Pinchera A, Marcocci C 2000 Management of Graves' ophthalmopathy: reality and perspectives. Endocrine Revs 21:168-199.

THE ADRENAL GLANDS

The adrenal cortex The cortex, whose role is to secrete steroid hormones, consists of a capsule, a subcapsular layer (the zona glomerulosa), the middle zone (zona fasciculata) and a zona

925

FIG. 17.27 Two-dimensional structures of three adrenal steroids aldosterone, cortisol and dehydroepiandrosterone sulphate

ences between the three zones are due to the presence or absence of a few critical enzymes.

Physiological control of the adrenal cortex

FIG. 17.26 Coronal section through the adrenal gland

926

reticularis. The latter envelops the adrenal medulla. The blood supply runs from the cortex to the medulla. All the steroid-secreting cells appear to be of common origin. Cell division occurs between the zona glomerulosa and zona fasciculata, and cell cords loop up through the zona glomerulosa and then down through the zona fasciculata to the zona reticularis, cells changing function as they go. In utero the glands are much larger and the cortex is dominated by a large innermost (fetal) zone, which degenerates shortly after birth. The zona glomerulosa synthesizes and secretes mineralocorticoids, of which the most important is aldosterone. The zona fasciculata synthesizes cortisol, and the reticularis and fetal zone the weak adrenal androgens, notably dehydroepiandrosterone (DHEA) and its sulphate. The fetal zone appears to develop as a consequence of the high levels of ACTH and oestrogen in utero. Oestrogens impair cortisol synthesis by inhibiting the action of the enzyme 3B-hydroxysteroid dehydrogenase. The two-dimensional structures of aldosterone, cortisol and dehydroepiandrosterone sulphate are shown in Figure 17.27. The basic pathways of steroidogenesis in the adrenal cortex are set out in Figure 17.28. The functional differ-

M ineralocorticoids Aldosterone is the most potent adrenal steroid; about 100 ug are secreted per day, compared to 10 mg for cortisol. Aldosterone acts on intranuclear mineralocorticoid receptors in the distal renal tubule to promote sodium uptake and potassium secretion. The receptor does not distinguish between mineralocorticoids and glucocorticoids. Specificity is conferred by rapid local conversion of cortisol to cortisone (by the enzyme llB-hydroxysteroid dehydrogenase), which ensures that aldosterone alone is available to the receptor. Normally, cortisone produced in the kidney is then 'shuttled' back to cortisol by the reciprocal enzyme in the liver. Liquorice can inhibit the renal enzyme and hence lead to excess mineralocorticoid action. Rare genetic defects in llB-hydroxysteroid dehydrogenase also produce hypertension and hypokalaemia. The synthesis and secretion of aldosterone are relatively independent of ACTH secretion. The major control factors are angiotensin II and the extracellular sodium and potassium concentration (see Fig. 20.25, p. 1073). Angiotensin II acts through cell surface receptors by activation of phospholipase C (p. 885), and principally stimulates enzymes late in the biosynthetic pathway. Renin is an enzyme secreted by the juxtaglomerular apparatus, which acts on angiotensinogen (of hepatic origin) to release the decapeptide angiotensin I. This is then converted to the active octapeptide angiotensin II in peripheral tissues, including lung, by angiotensin-converting enzyme (ACE). Glucocorticoids Glucocorticoids, of which the main one in humans is cortisol (hydrocortisone), have a wide range of actions on many

17

FIG. 17.28 The major steroid pathways in adrenal zona glomerulosa, fasciculata and reticularis, plus ovary or testis The balance of steroids secreted is determined by the enzymes present (represented by arrows).

different cells and tissues. They have a major influence on glucose homeostasis, with catabolic actions on muscle and enhanced hepatic gluconeogenesis. In their absence, the response to stress of any kind is greatly impaired. Glucocorticoids inhibit the immune system, repair processes and collagen synthesis in bone and soft tissues. They also modify the effect of many other hormones on a range of cell types. Glucocorticoids are secreted episodically, with a pronounced diurnal variation. It is probable that they facilitate the immediate requirements of daytime activity while inhibiting other processes, such as tissue repair, which then occur at night. Adrenal androgens Apart from their role as oestrogen precursors in utero, adrenal androgens appear to be of relatively minor importance. Increased synthesis before puberty (the adrenarche) contributes to the development of secondary sexual hair.

The adrenal medulla The adrenal medulla receives a sympathetic nerve supply, which is responsible for initiating adrenaline (epinephrine) secretion in response to stress. Its blood supply comes largely from the adrenal cortex and is therefore very rich in cortisol. The pathway of adrenaline synthesis from tyrosine, derived from the diet or synthesized by the liver from phenylalanine, is indicated in Figure 17.29. The last stage is cortisol dependent; cortisol appears to control the synthesis of the enzyme phenylalanine N-methyl transferase.

FIG. 17.29 Pathway for the biosynthesis of adrenal catecholamines

927

Adrenal glands and the response to changes in trophic hormones Chronic ACTH excess leads to hypertrophy of the adrenal cortex, particularly the zona fasciculata-reticularis, and to alterations in the blood vessels. In some individuals the chronic consequence appears to be multinodular hyperplasia, in which the nodules of hyperplastic tissue tend to become semi-autonomous; rarely, malignant tumours may develop in these nodules. Chronic ACTH insufficiency leads to adrenal atrophy. The zona glomerulosa continues to function under the trophic stimulation of angiotensin II, so that the picture is predominantly one of glucocorticoid (and adrenal androgen) deficiency, rather than of mineralocorticoid lack.

INVESTIGATION OF ADRENAL DISORDERS Methods once considered essential have been supplanted by direct, simple, cheap and reliable techniques. For example, chemical methods for the analysis of steroid metabolites have now been largely replaced by radioimmunoassay of the levels of the steroids themselves. The clinical investigator should only use such tests as are necessary to answer a question beyond reasonable doubt. Electrolyte disturbances may provide an important clue to the presence of an adrenal disorder.

Mineralocorticoid assays Aldosterone can be measured by radioimmunoassay in serum or urine; plasma renin measurements are also reasonably reliable. Provided the patient is not on diuretics and the electrolyte levels are known, renin and aldosterone measurements in lying and standing positions are helpful in the investigation of suspected mineralocorticoid disorders, notably in the diagnosis of primary hyperaldosteronism.

Adrenal androgen assays A virilizing tumour of the adrenal in women can be excluded by measuring serum testosterone levels (p. 951). Plasma androstenedione and DHEAS assays are available but contribute little extra to the practical management of hirsutism or adrenal disorders.

Use of dynamic function tests Adrenal function is intrinsically dynamic, and much can be learnt from examining spontaneous changes in activity, e.g. diurnal changes in serum cortisol. Repeated 9 a.m. and midnight samples will be helpful in diagnosing Cushing's syndrome, where there is loss of the diurnal rhythm and levels remain high in the evening. In contrast, there is little point in taking late evening samples in a patient with suspected cortisol deficiency, as cortisol levels are normally low at that time.

Cortisol assay

928

The urinary free cortisol excretion in 24 hours correlates well with both the cortisol secretion rate (a tedious radioisotopic measurement) and the mean 24-hour free plasma cortisol level. It is more widely used to diagnose hyper- than hypocortisolism. Random serum cortisol assays have the disadvantage that the hormone is normally secreted episodically, so that repeated measurements are usually needed for reliable interpretation. However, any morning cortisol value below l00nmol/L supports a diagnosis of hypoadrenalism, and one above 400 nmol/L virtually excludes this diagnosis. The midnight cortisol assay, taken while the patient is asleep (often used in the investigation of Cushing's syndrome), has the disadvantage that stress can cause its elevation, even in normal individuals. Also, the patient has to be admitted to hospital, which makes it expensive. Assays for precursors of cortisol are of use in diagnosing enzyme deficiencies or assessing drug-induced enzyme blocks; they include 17-hydroxyprogesterone (or its urinary metabolite pregnanetriol) and 11-desoxycortisol. The two-site immunoradiometric assay has made the plasma ACTH assay more sensitive and reliable. An elevated ACTH level is found in Addison's disease and ACTH-dependent forms of Cushing's syndrome, but ACTH is undetectable when Cushing's syndrome is due to an adrenal tumour.

Suppression tests (Fig. 17.30) The most useful suppression test involves the administration of dexamethasone (which does not cross-react in the cortisol assay), with subsequent measurement of serum cortisol. The overnight dexamethasone suppression test involves the patient taking 1 mg of dexamethasone orally at midnight, with a serum sample being taken for cortisol assay 8 hours later. In a normal individual dexamethasone suppresses the morning secretion of ACTH, so that the morning serum cortisol value is low. This can be done without admitting the patient and is a useful screening test for Cushing's syndrome. Failure to suppress indicates Cushing's syndrome or excess metabolism of dexamethasone due to enzyme induction by alcohol, or drugs such as phenytoin and carbamazepine. Depression, obesity and pregnancy may also cause false positive tests. Confirmation of Cushing's syndrome involves the use of low-dose (0.5 mg 6-hourly for 48 hours) suppression with dexamethasone; serum and sometimes urine cortisol levels are assayed. The high-dose dexamethasone suppression test is used to distinguish between the causes of Cushing's syndrome. After 2mg dexamethasone 6-hourly for 48 hours there is at least 50% suppression of serum cortisol in pituitary-dependent (but not other forms of) Cushing's syndrome. Unfortunately, false positive and false negative results occur with this test.

and has been used to differentiate between pituitarydependent Cushing's syndrome, in which ACTH and cortisol precursors rise, and ectopic ACTH syndrome, in which there is no response. It is no longer widely used. Petrosal sinus sampling plus CRH testing is now the most reliable way to distinguish between these two types of Cushing's syndrome. Catheters are inserted via the femoral veins into the inferior petrosal sinuses, where ACTH levels can be measured directly after release from the pituitary. A gradient between petrosal and peripheral vein ACTH levels indicates Cushing's syndrome due to a pituitary adenoma, whereas with an ectopic source of ACTH the two levels will be similar. Administration of CRH and demonstration of a rise in petrosal sinus ACTH levels further strengthens the diagnosis of a pituitary corticotroph adenoma. The CRH test with peripheral blood sampling alone can be used prior to petrosal sinus sampling, as a 50% or greater rise in ACTH or cortisol is strongly suggestive of a pituitary rather than an ectopic source of ACTH in Cushing's syndrome.

17

Investigation of the adrenal medulla

FIG. 17.30 Dexamethasone suppression test A Normal feedback mechanism for cortisol production. B Normally, dexamethasone suppresses ACTH and hence cortisol production, [c] In Cushing's syndrome due to pituitary or ectopic ACTH excess, or adrenal adenoma, dexamethasone has no effect, and cortisol is not suppressed.

Stimulation tests The short synacthen test is generally used to identify adrenal insufficiency. Synacthen is synthetic ACTH; i.m. injection of 250 ug should produce a rise in cortisol to 550 nmol/L 30-60 minutes later. This does not occur in primary or secondary adrenal failure, but the test is unreliable if secondary failure is recent (e.g. after pituitary surgery). In this situation the insulin tolerance test (Fig. 17.13, p. 899) should be used. In some centres a low-dose (lug) synacthen test is used, as this is believed to be more physiological and sensitive. Primary and secondary adrenal failure are usually distinguished by measuring serum ACTH levels, which are high in the former and low in the latter. An alternative is to give synacthen repeatedly for 4-5 days (the long synacthen test). A cortisol response will be found only in secondary adrenal failure. Corticotrophin-releasing hormone (CRH) can be of value in distinguishing pituitary-dependent ACTH secretion (where ACTH responds to CRH) from ectopic ACTH secretion (where there is no response). Metyrapone blocks the last stage in cortisol biosynthesis

The diagnosis of phaeochromocytoma is confirmed by one or more of a number of methods for measuring catecholamines or their metabolites. The simplest is the urinary excretion of vanillylmandelic acid (VMA), but urinary metanephrines are more reliable. Serum assays for noradrenaline (norepinephrine) and adrenaline (epinephrine) taken lying and standing and their metabolites, normetanephrine and metanephrine, are also of considerable diagnostic value. The clonidine suppression test is useful in difficult cases: prompt suppression of plasma catecholamines normally occurs after giving 0.3mg of this a2-agonist.

Radiological methods CT scanning is particularly valuable in localizing adrenal tumours of the cortex and medulla (Fig. 17.31). Radioisotope scans using a labelled catecholamine precursor (131I-MIBG) are useful to localize phaeochromocytomas.

DISEASES OF THE ADRENAL CORTEX ADRENOCORTICAL INSUFFICIENCY Aetiology The major causes of adrenal insufficiency are shown in Table 17.15. Acute adrenal destruction occurs in septicaemia, particularly that due to meningococcus (the so-called Waterhouse-Friedrichsen syndrome). Meningococcal septicaemia is associated with purpura and haem-

929

TABLE 17.15 Causes of adrenocortical insufficiency Primary (High ACTH) Congenital/familial Adrenal enzyme defects (usually 21-hydroxylase deficiency) Congenital adrenal hypoplasia Adrenal leukodystrophy Acquired Infections - tuberculosis1 - histoplasmosis - HIV infection Autoimmune1 Vascular - meningococcal septicaemia and haemorrhagic destruction - thrombosis (adrenal vein catheterisation) Neoplastic - secondary carcinoma Degenerative - amyloid Drugs - ketoconazole, etomidate, metyrapone Secondary (Low ACTH) Hypopituitarism Isolated ACTH deficiency Glucocorticoid therapy Enhanced glucocorticoid metabolism Drug-induced, e.g. rifampicin*, carbamazepine 1 These commonly present as Addison's disease (chronic adrenal insufficiency with pigmentation). * Important - may precipitate adrenal failure if already compromised, e.g. in tuberculosis. Accelerated hepatic steroid metabolism enhances steroid requirements.

FIG. 17.31 CT scan of the adrenal glands In A the scan shows an adrenal adenoma (arrowed) in a patient with Cushing's syndrome; in B there is bilateral adrenal enlargement (arrowed) due to ectopic ACTH secretion in a patient with pancreatic carcinoma and Cushing's syndrome.

orrhage, including massive haemorrhagic destruction of the adrenals. Untreated, it leads rapidly to circulatory collapse and death. Chronic destruction of the adrenal glands is most commonly due to tuberculosis or autoimmune disease, leading to the clinical features of Addison's disease (see below). Adrenal insufficiency should always be considered in any patient with circulatory collapse; the picture will vary somewhat depending on whether it is primary or secondary (p. 897). However, the presentation of adrenal 1

930

Fig. 17.21

0 Fig. 17.22

insufficiency tends to be much more insidious. Tuberculosis leads to enlargement of the adrenals and, eventually, total destruction of the normal tissue and caseation. Possibly the high cortisol levels within the adrenal make it difficult to mount the necessary cellular immune response to eradicate the bacteria. In western countries the decline in tuberculosis makes autoimmune destruction of the adrenals the commonest cause of chronic adrenal insufficiency, with a prevalence of 30-60 per million. Autoimmune adrenal insufficiency is often associated with a high incidence of other organ-specific autoimmune disease, especially thyroid failure (Table 17.11, p. 917). There is an association with HLA-DR3. Secondary adrenal insufficiency due to hypopituitarism and isolated ACTH deficiency is discussed on page 897.

ADDISON'S DISEASE Clinical features This was first described by Thomas Addison in 1855. The speed of onset varies from insidious to sudden. When the

SUMMARY 10 Clinical features of Addison's disease • Increased pigmentation in mouth, skin creases, pressure areas. • Crises of hypotension, salt and water loss, hyperkalaemia, hypercalcaemia, hypoglycaemia. • Associated vitiligo, thyroid disease or pernicious anaemia.

onset is insidious the early symptoms are lassitude, somnolence and depression. Dizziness due to postural hypotension, or fasting hypoglycaemia and nocturia owing to failure to excrete a water load in the daytime, may occur. Frequently, the patient notices an increased pigmentation or simply retention of the summer suntan; as a result they may appear 'healthy' to relatives. In its most dramatic form the disease presents with Addisonian crisis, which in the absence of effective treatment is also the terminal picture. The patient is ill and anorexic, with vomiting and abdominal pain, often with rigidity simulating an abdominal emergency. There is usually a fever, tachycardia, hypotension (with a marked postural drop in blood pressure), cramps and postural dizziness, and signs of sodium and water depletion. There is marked loss of skin turgor, with loose skin .like damp leather and sunken eyes. The blood urea rises because of poor renal perfusion, the serum potassium concentration is raised and serum sodium is low. Hypercalcaemia sometimes occurs. Hypoglycaemia may be present, particularly in children. The features of the associated ACTH excess are variable, and consist of increased pigmentation 1 of the skin (especially exposed areas, pressure points, creases and recent scars), buccal mucous membrane and sometimes sclerae, and streaked pigmentation of the nail beds. These changes are due to stimulation of melanocytes by ACTH and related peptides, and are absent in secondary hypoadrenalism. The clinical features due to mineralocorticoid deficiency - principally sodium and water depletion, hyperkalaemia and hypercalcaemia - are unique to primary adrenal failure, and are absent or much less marked in secondary adrenal failure. The serum sodium concentration is usually low in both conditions, in part owing to excessive secretion of ADH. Features due to glucocorticoid deficiency include hypoglycaemia (more marked in pituitary insufficiency, as GH is also usually lacking), inability to respond to stress and infection, and also, in part at least, hypotension and pyrexia. Anaemia is common, with a lymphocytosis and an elevated eosinophil count. Other suggestive clinical features are the presence of other forms of organ-specific autoimmune disease such as vitiligo 2 (Table 17.11, p. 917) or a history, or signs of, tuberculosis.

Investigation A high index of suspicion is important in the diagnosis of Addison's disease. Once suspected in a collapsed or ill

patient, the appropriate blood samples should be taken for electrolyte and cortisol assay, and treatment instituted forthwith. Hyperkalaemia (with or without a low serum sodium concentration) in a patient with Addisonian-type pigmentation, hypotension and fluid depletion is very suggestive of primary adrenal insufficiency. A low or undetectable cortisol in a stressed, sick patient confirms adrenal failure. Once the patient is well, further investigation can be conducted to establish the aetiology. Ideally the diagnosis should be made before severe insufficiency presents as an emergency. The short synacthen test (p. 929) is the investigation of choice to confirm adrenal insufficiency. ACTH levels or the long synacthen test are used to show that this is primary adrenal failure. The aetiology can be identified by tests for adrenal (and other) autoantibodies, plain X-ray of the abdomen (to show adrenal calcification due to tuberculosis) and CT scanning of the adrenals (to exclude tuberculosis and adrenal infiltration).

17

Management Immediate treatment of the acutely ill patient is with fluids, electrolytes and hydrocortisone (Table 17.16). Mineralocorticoid therapy is only required when the dose of hydrocortisone is being reduced to two to three times normal replacement doses (i.e. about 60mg/day), and then only in primary adrenal insufficiency. Long-term management of adrenal failure is generally simple if the patient adheres to medication. It is not possible to simulate precisely the normal episodic pattern of hormone secretion. Cortisol itself can be conveniently given as hydrocortisone in a split dose, which for an adult is often given as 20 mg on rising and 10 mg at about 5 pm, although patients often feel better if the morning dose is split to 10 mg on rising and 5-10 mg at lunchtime, and then 5-10 mg at teatime. Some tailoring of the dose is possible by repeated cortisol measurements through a day. Mineralocorticoid replacement is with the synthetic mineralocorticoid fludrocortisone (0.05-0.2mg daily). The dose is adjusted according to blood pressure, serum potassium levels and, more precisely, by maintaining renin levels TABLE 17.16 Treatment of acute adrenal insufficiency (Addisonian crisis) • Treat as soon as the diagnosis is suspected: do not await biochemical confirmation. • Take a blood sample for retrospective cortisol and ACTH analysis prior to the administration of steroids. • Establish i.v. access and administer 1L of normal saline over 2-4 hours; monitor for signs of overload using CVP if possible; continue normal saline thereafter at a slower rate. • Monitor glucose levels; give 5% dextrose in normal saline as appropriate. • Hydrocortisone is given as a 100mg bolus intravenously immediately and then every 6 hours; alternatively, give as an intravenous infusion of 4mg/h. • Treat any underlying precipitant, such as infection.

931

within normal limits. Elevated renin levels indicate the need to increase fludrocortisone. A minority of Addisonian patients do not need mineralocorticoid therapy, because zona glomerulosa function may be spared and replacement glucocorticoids have a weak mineralocorticoid effect. Recent work indicates that dehydroepiandrosterone improves libido and general wellbeing in women with Addison's disease. Patients on long-term steroid replacement should be given a special card clearly stating the steroids that they are on, and should keep it with them at all times. It is important that additional glucocorticoid is given in emergencies and for incidental infections. If the patient has a gastrointestinal upset he or she should be given an increased dose parenterally. Some drugs increase steroid requirements because they accelerate steroid metabolism; this occurs particularly with anticonvulsants and the antituberculous drug rifampicin, which is given as part of triple chemotherapy if tuberculosis is the cause of Addison's disease. 1

with azoospermia. Ultimate height is reduced and, if treatment with replacement glucocorticoids is delayed until the epiphyseal bone age has advanced to 11 years or more, affected males often undergo true precocious puberty. There is a high prevalence (up to 1 in 100) of a mild enzyme deficiency resulting in the non-classic form of the disorder, which is the commonest autosomal recessive condition in humans. There is no obvious abnormality at birth, but girls may present with premature pubic hair development, short stature and, later, menstrual irregularities and hirsutism. Most such individuals, however, do not come to medical attention. Once suspected, the diagnosis is easily made by demonstrating elevated levels of plasma, salivary or urinary levels of 17-hydroxyprogesterone or its urinary metabolite pregnanetriol, with or without ACTH stimulation. The importance of early diagnosis in female infants is obvious, and corrective genital surgery is generally required in infancy and again later with the onset of sexual activity. Replacement therapy is with glucocorticoids and mineralocorticoids.

ADRENAL ENZYME DEFECTS

11 B-hydroxylase deficiency

The pathways of steroidogenesis are discussed on pages 925-928. The principal defects in these pathways involve the cytochrome P450 enzymes 21-hydroxylase and 11Bhydroxylase deficiency (Fig. 17.32), although other defects occur rarely.

21 -hydroxylase deficiency 21-hydroxylase deficiency is inherited as a recessive condition. It is due to a mutation or loss of the gene for 21hydroxylase, which is located on chromosome 6, between the HLA-B and HLA-D loci. The incidence, based on neonatal 17-hydroxyprogesterone screening, is around 1 in 15 000 births, but this misses many with a milder defect. The severe 'classic' form can be recognized in newborn girls by ambiguous external genitalia caused by high levels of adrenal androgens. Around 75% of affected newborn babies develop severe mineralocorticoid deficiency and salt loss. Male infants appear normally virilized. Salt-losing males present with acute adrenal insufficiency and sodium and water depletion, usually within the first few weeks of life. The mortality is therefore much higher with male infants and the cause of death may not be recognized unless previous affected siblings have drawn attention to its likelihood. Non-salt-losing males will present with accelerated growth and sexual development but small testes (sexual precocity and pseudopuberty) and, later,

1MCQ17.19

932

3

2

Fig. 17.23

Figs 17.24, 17.25

Deficiency of llB-hydroxylase (Fig. 17.32) is 20 times less common than 21-hydroxylase deficiency. It also causes virilization, but the androgen excess is generally less marked and the picture is dominated instead by the presence of hypertension. This is due to ACTH-induced excessive secretion of 11-desoxycorticosterone (DOC). The treatment is similar to that of 21-hydroxylase deficiency, although transient salt wasting may occur when treatment is started.

CUSHING'S SYNDROME Harvey Cushing first identified this syndrome in 1912 and later suggested that pituitary dysfunction was a cause. The diagnosis and the treatment of the disorders that lead to adrenal glucocorticoid excess (Cushing's syndrome) are often difficult. latrogenic Cushing's syndrome is common but readily identified. The annual incidence of the other types combined is around 25 per million.

Aetiology Cushing's syndrome is either ACTH dependent or independent (Table 17.17). Around 80% of non-iatrogenic cases are due to pituitary adenomas, although ectopic secretion of ACTH is actually more common owing to the prevalence of lung carcinoma. Most such cases are not recognized because of the high mortality rate. ACTHindependent Cushing's syndrome is caused by a steroidsecreting tumour (carcinoma or adenoma) of the adrenal cortex. Adrenal tumours generally predominate in children and young adults; adenoma is the easiest form of the syndrome to treat.

Glomerulosa

Reticularis

Fasciculata

17

Cholesterol (20Q-OH Cholesterol)

DHEA Sulphate

Pregnenolone

17a-OH preonenolone

DHEA

Progesterone

17a-OH progesterone

Androstenedione

11-Desoxycorticosterone

11-Desoxycortisol

Corticosterone

Cortisol

A 5 Androstenediol

21-Hydroxylase

11B-Hydroxylase

18-OH Corticosterone

Aldosterone FIG. 17.32 Normal adrenal steroid pathways with sites of action of 21-hydroxylase and 11p-hydroxylase, the two enzymes most commonly defective in congenital adrenal hyperplasia

ACTH-dependent causes are pituitary or non-pituitary in origin. Pituitary Cushing's syndrome is almost always due to a pituitary adenoma (Cushing's disease) and is three times more common in women. Some rare cases are due to secretion of CRH by a hypothalamic or extrahypothalamic tumour. Ectopic ACTH syndrome is caused by ACTH secretion from a tumour outside the pituitary gland. The commonest types are carcinoid tumours of the bronchus, thymus and pancreas, small cell lung cancer and thymoma. Pancreatic carcinoma, medullary thyroid carcinoma and phaeochromocytoma are other rare causes (see Table 6.4, p. 154). Alcohol can sometimes induce clinical features of Cushing's syndrome, but these and abnormal cortisol levels become normal with abstinence. Depression and obesity can cause diagnostic problems, as abnormalities in cortisol levels can mimic Cushing's syndrome (p. 928).

TABLE 17.17 Causes of Cushing's syndrome Cause ACTH driven Pituitary dependent

Ectopic ACTH secretion

Pituitary adenoma* Pituitary carcinoma** CRH-secreting tumour** Hypothalamic Ectopic Tumours Small cell carcinoma Branchial carcinoid Medullary thyroid carcinoma Pancreatic carcinoma Phaeochromocytoma

Non-ACTH driven

Adrenal adenoma Adrenal carcinoma Micronodular or macronodular hyperplasia Glucocorticoid therapy

Mixed aetiology

Multinodular hyperplasia Alcohol-induced 'pseudo' Cushing's Cyclical Cushing's syndrome

Clinical features Common features Features common to all forms of Cushing's syndrome are those due to glucocorticoid excess. They include centripetal obesity, proximal muscle weakness and wasting, thin skin, striae, 0 poor wound healing, osteoporosis often presenting as vertebral fractures, hypertension, venous thrombosis, diabetes mellitus, depression and psychosis (Fig. 17.33). Mild hirsutism and acne are common, 3

Example

* 'Cushing's disease' ** Very rare

933

TABLE 17.18 Investigation of suspected Cushing's syndrome 1. Clinical assessment 2. Confirmation:

24-hour urine free cortisol Overnight dexamethasone suppression

3. If one or both abnormal:

Plasma electrolytes Further 24-hour urine free cortisol Circadian blood cortisol 9 am and midnight ACTH levels Low-dose dexamethasone suppression test (2 mg/day)

4. If abnormal, find cause:

Alcohol? Ectopic ACTH? Adrenal tumour? Pituitary dependent?

- Withdraw and reassess - CXR and mediastinal CT - Adrenal CT ACTH levels - High-dose dexamethasone suppression test Pituitary CT or MRI scan CRH test with petrosal sinus sampling

5. Treat

SUMMARY 11 Clinical features of Cushing's syndrome

FIG. 17.33 Classic features of severe Cushing's syndrome

owing to adrenal androgen as well as glucocorticoid excess. In men, LH and testosterone secretion may be inhibited and libido consequently reduced. Oligomenorrhoea or amenorrhoea are frequent in women.

• • • • • • •

Central obesity, moon face, buffalo hump Proximal myopathy Osteoporosis Striae and poor wound healing Depression and psychosis Hypertension Diabetes mellitus

may well be absent, and a rapid onset also suggests a malignant ectopic source.

Investigation

934

Features unique to specific forms Cushing's syndrome that disappears or diminishes under hospital scrutiny is commonly alcohol induced. Some other forms may be cyclical. Addisonian-type pigmentation suggests marked ACTH excess, which is found most often with the ectopic ACTH syndrome (unless the adrenals have been removed, when it suggests Nelson's syndrome; see below and p. 906). The presence of profound hypokalaemia and oedema suggests the ectopic ACTH syndrome. The ectopic ACTH syndrome due to a small cell carcinoma of lung may not be attended by centripetal obesity, because of cachexia. Other typical clinical features

The investigation of suspected Cushing's syndrome is summarised in Table 17.18. The difficulties have been discussed under dynamic testing (see p. 928). The 24-hour urinary free cortisol excretion and overnight low-dose dexamethasone test (p. 928) are the mainstays of initial diagnosis. ACTH-independent causes can be diagnosed by adrenal CT scan (Fig. 17.31, p. 930) and the suppressed plasma ACTH level. Pituitary CT or MR scanning is often surprisingly unhelpful (and occasionally misleading) in the investigation of ACTH-dependent causes of Cushing's syndrome. Pituitary tumours are usually small, and rarely cause any distortion of the pituitary fossa. CT scan of the

17

CASE STUDY 17.3 DEPRESSION AND WEIGHT GAIN A 43-year-old woman complained of weight gain of approximately 2kg over several months, associated with swelling of the ankles. She noticed that, in particular, her abdomen had become distended. At the same time she had noticed increasing tiredness, which had become progressive, and her skin had become dark. Friends had commented that she appeared to have a suntan, although there was no recent exposure to the sun. In the previous month her skin had started to bruise easily. Her periods were less regular than previously, occurring every 40 days or so. Her sleep was disturbed and she was waking up to 2 hours earlier than normal. She was also short of breath. She smoked 20 cigarettes a day but drank no alcohol. She had had a laminectomy for a prolapsed intervertebral disc, and took ibuprofen and dihydrocodeine to relieve back pain. She was also taking furosemide (frusemide), recently prescribed by her GP for ankle oedema. On examination she was tearful and acknowledged that she had been feeling depressed over the preceding 2-3 weeks. Her skin thickness was slightly reduced and the abdomen was distended, with striae. There was filling-in of the supraclavicular fossae with fat, a moon-shaped face and a small pad of fat over the nape of the neck. Several bruises were noted over the shins in particular. Her blood pressure was normal at 130/80, and the remainder of the examination was normal except for mild weakness of the proximal muscles. Questions 1. What is the most likely cause of this woman's symptoms? 2. What investigations are indicated? Discussion The symptoms and signs point to a diagnosis of Cushing's syndrome. The features are very typical, although

missing from the presentation in this case are other fairly frequent findings, such as hypertension (85% of cases), hirsutism (around 70% of women) and acne (around 50% of women). Other less frequent features are polydipsia (30%) and poor wound healing (around 35%). Back pain and bone pain occur in around 50% of patients with Cushing's syndrome, but in this woman's case the problems were ascribed to pre-existing musculoskeletal problems which had been treated surgically. Indeed, her back problem had not deteriorated during recent months. It was noted that the patient was depressed, and biochemical changes of Cushing's syndrome can occur with severe depression. On the other hand, mental changes occur in 50% of patients with true Cushing's syndrome, and therefore her depression would fit with this diagnosis. Another common problem that can cause confusion is alcoholism, which can produce a clinical and biochemical picture of Cushing's syndrome (so-called pseudo-Cushing's syndrome) but this patient denied alcohol ingestion and a random blood alcohol level test was negative. She was taking no drugs that would affect cortisol metabolism and interfere with tests for Cushing's syndrome (in particular, anticonvulsants can interfere with diagnosis). Investigation consists of a twostage process. First, the diagnosis of Cushing's syndrome must be confirmed biochemically, and once this is done the exact cause needs to be determined. The first of these goals can be accomplished as an outpatient procedure using urine free cortisol measurement or administering 1 mg of dexamethasone at midnight, followed by estimation of the cortisol level at 8.00-9.00 am, values of plasma cortisol greater than 50nmol/L being highly suggestive of Cushing's syndrome. There are problems with ensuring a complete

collection of urine as an outpatient, and therefore if there is doubt about the diagnosis arrangements should be made for admission to hospital, where the diurnal rhythm of cortisol can be assessed. This is best done by taking the cortisol at midnight when the patient is asleep, a value above 50 nmol/L being very suggestive of Cushing's syndrome. This woman's cortisol at midnight was 254 nmol/L and the diagnosis of Cushing's syndrome was further confirmed by performing a long, low-dose dexamethasone suppression test (dexamethasone 0.5 mg, four times a day for 2 days). The cortisol value at the end of this was 210 nmol/L, indicating Cushing's syndrome (normal <50 nmol/L). The second stage is to confirm the cause of the syndrome. A number of different tests can be used, but essentially the first distinction to make is whether the ACTH level is detectable or not: a detectable ACTH implies Cushing's disease (i.e. pituitary-driven Cushing's syndrome) or an ectopic ACTH syndrome, whereas an undetectable ACTH is compatible with an adrenal adenoma or carcinoma. Rare patients have bilateral micronodular adrenal hyperplasia, again with an undetectable ACTH. When an adrenal problem is suggested by low ACTH levels, an adrenal CT or MRI scan will reveal the source of the Cushing's syndrome. ACTH levels tend to be very high with the ectopic ACTH syndrome and the source is often obvious, most typically small cell carcinoma of the lung, visible on a chest X-ray. In pituitary-dependent Cushing's disease there is often suppression of the cortisol after highdose dexamethasone (2mg dexamethasone four times a day for 2 days, with suppression being judged as a 50% or more decrease in the basal cortisol level taken at 9.00 am). It can be difficult to distinguish between an occult, ectopic source of ACTH and a pituitary tumour.

935

CASE STUDY 17.3 CONTINUED Unfortunately, the mere presence of a pituitary tumour on MRI or CT does not confirm that this is the source of ACTH, as pituitary tumours are found in up to 5% of the healthy population. Moreover, not all pituitary tumours secreting ACTH are visible by current scanning methods. The use of a corticotrophinreleasing hormone (CRH) test, often combined with petrosal sinus sampling, is currently the best way to differentiate between these two possibilities. Patients with Cushing's disease have a rise in plasma cortisol

and plasma ACTH levels 15 minutes after administration of CRH, but this does not occur in ectopic ACTH syndrome. This test is also useful in differentiating pseudo-Cushing states from Cushing's syndrome, as there is no response to CRH with pseudoCushing's syndrome due to alcoholism, depression, obesity or drugs. This test can be performed using peripheral blood samples, but if the diagnosis remains unclear the test can be extended by carrying out cannulation and sampling of the petrosal sinuses draining the pituitary,

thorax may reveal a small bronchial or thymic carcinoid, which is the source of ectopic ACTH. 1

Management Drug therapy Drugs have a place in the preoperative, and occasionally also the long-term, management of Cushing's syndrome. Steroid production can readily be inhibited in patients with adrenal tumours by the llp-hydroxylase inhibitor metyrapone. Indeed, there is real risk of inducing adrenal insufficiency. The dose (250-750 mg every 4 hours, omitting 4am dosage) should be titrated against the serum and urinary cortisol response. Ketoconazole (200-400 mg twice a day) is an alternative that does not exacerbate hirsutism in women, unlike metyrapone. It causes nausea and regular checks of liver function are required. Before surgery, it is wise for the patient to be treated with either drug for several weeks to improve wound healing. Palliative drug therapy in ectopic ACTH syndrome generally requires the simultaneous use of metyrapone and ketoconazole to block biosynthesis of cortisol; replacement doses of steroids can be given once endogenous production is blocked. Aminoglutethimide (0.25-1 g daily in divided doses) can also be used. The DDT analogue mitotane (10-20 g daily) also has an established place in the palliation of adrenal carcinoma; it induces adrenal cell necrosis but causes major gastrointestinal and neurological side-effects; lower doses may be preferable. Cytotoxic chemotherapy is the treatment of choice for small cell lung cancer, but the prognosis of the disease is especially poor when accompanied by ectopic ACTH secretion.

1

936

MCQ 17.20

2

Figs 17.26-17.28

and showing a direct release of ACTH in response to CRH, which confirms a pituitary-dependent cause. In this case the ACTH level was 50ng/L (reference range <20) and the cortisol was suppressed to >50% after a high-dose dexamethasone suppression test. The plasma cortisol level rose by more than 100% in response to CRH, and a pituitary MRI scan revealed a 3 mm microadenoma on the left side of the pituitary. Removal of the adenoma by the transsphenoidal route led to cure of the Cushing's disease.

Adrenal surgery Surgery is clearly indicated for an adrenal adenoma or carcinoma, where operable. Surgery is also the treatment of choice for carcinoid tumours causing ectopic ACTH secretion. Bilateral adrenalectomy was formerly the preferred treatment for Cushing's disease. This has the advantage of eliminating the possibility of cortisol excess (provided no remnant is left behind). The disadvantages are that it involves major surgery, with a significant mortality, and also a significant (10-20%) risk of developing Nelson's syndrome (the condition of hyperpigmentation, increasing ACTH excess and an expanding pituitary tumour). The syndrome probably only develops where there is already a semi-autonomous ACTH-secreting pituitary tumour, previously kept under partial restraint by the high circulating cortisol levels. If bilateral adrenalectomy is performed in Cushing's disease, for instance when pituitary surgery fails, external pituitary irradiation minimizes the risk of this unpleasant complication. 2 Bilateral adrenalectomy is the treatment of choice in multinodular adrenal hyperplasia. Pituitary surgery The treatment of choice for Cushing's disease is surgery by the transsphenoidal route (p. 901). If a tumour is present it is removed; if not, then as much as possible of the pituitary gland is resected. Hypopituitarism (with replacement therapy) is generally preferable to continuing active Cushing's syndrome. Pituitary radiotherapy Pituitary radiotherapy can be an effective treatment for Cushing's disease. Radiotherapy may be particularly useful in the elderly or frail patient with moderately active disease who has shown some hormonal response to drug therapy.

CLINICALLY SILENT ADRENAL TUMOURS Adrenal tumours are increasingly being detected in the absence of symptoms owing to the widespread use of CT and MRI carried out to diagnose non-adrenal disease. Between 0.4 and 4.0% of patients scanned have a previously unsuspected adrenal mass. Adrenal carcinoma is very rarely silent, whereas metastases from other cancers are common, although in most such patients there will be evidence of metastases elsewhere or of the primary tumour. The majority of adrenal masses are benign adenomas, of which only around 15% exhibit hormonal hypersecretion. There is currently no consensus on the best way of eliciting the nature of these clinically silent tumours. At present, a careful history and examination remain the key to determining further investigations, but all patients should probably also have screening for phaeochromocytoma (usually urinary catecholamine metabolites), Conn's syndrome (serum potassium) and Cushing's syndrome (urine or serum cortisol). Dehydroepiandrosterone sulphate may be elevated in adrenal carcinoma. Any mass greater than 5 cm has an increased risk of malignancy and should be removed; repeat scans at 3-6 months are indicated to verify that smaller lesions have not increased in size. CTguided fine needle biopsy may improve future diagnostic accuracy.

DISORDERS OF ALDOSTERONE SECRETION The basic components (renin, angiotensin and aldosterone) of the mineralocorticoid system are discussed on page 926. In addition, the heart is also involved in salt and water homeostasis, through atrial natriuretic peptide (ANP). This peptide is produced by endocrine cells in the right atrium, and acts on the renal tubule to cause a sodium diuresis (Ch. 20, p. 1035).

Hypoaldosteronism Hypoaldosteronism may be a result of general adrenal insufficiency owing to destruction of the adrenal glands (p. 929), or it may occur in the presence of normal secretion of other adrenal steroids. The disorder may rarely be primary, owing to a selective adrenal enzyme deficiency late in the aldosterone pathway. Hypoaldosteronism secondary to renin deficiency occurs quite commonly in long-standing diabetes mellitus, and may be a consequence of autonomic neuropathy. Aldosterone deficiency is manifested by a tendency to sodium depletion, hypotension and hyperkalaemia. Treatment is with fludrocortisone. Pseudohypoaldosteronism is rare and mimics aldosterone deficiency, but is due to mutations in the epithelial sodium channel or in the mineralocorticoid receptor. In both types there is resistance to aldosterone, which rises to high levels.

Hyperaldosteronism Primary hyperaldosteronism (Conn's syndrome) is a rare cause of a generally mild hypertension and hypokalaemia, occurring in around 1% of the hypertensive population. Not all patients have hypokalaemia, and in some patients the diagnosis is suggested by the provocation of significant hypokalaemia with diuretics. In the Chinese especially primary hypoaldosteronism may present with hypokalaemic paralysis. It is caused by an adrenal adenoma. Idiopathic zona glomerulosa hyperplasia results in a similar clinical picture, but there is recent evidence that this is not a primary adrenal disorder; it may be due to enhanced sensitivity of the zona glomerulosa to angiotensin II. The diagnosis of primary hyperaldosteronism is confirmed by the finding of elevated serum aldosterone and low renin levels, and by demonstrating an adenoma on CT or MRI scan. Treatment is with spironolactone or (in the case of definite adenoma) adrenal surgery. Hypokalaemia and hypertension may rarely occur with llp-hydroxysteroid dehydrogenase deficiency (p. 926) and this responds to suppression of endogenous cortisol levels by dexamethasone. Secondary hyperaldosteronism is a common physiological adaptation seen in such disorders as congestive cardiac failure, accelerated hypertension, cirrhosis of the liver and nephrotic syndrome. Reduced renal perfusion is the major stimulus to increased renin secretion, angiotensin II production and aldosterone secretion, and thus to sodium retention and potassium excretion. Rarely, hyperaldosteronism is caused by a renin-secreting tumour of the kidney. Liddle's syndrome is rare and mimics hyperaldosteronism, with hypertension and hypokalaemia, but is due to mutations in the epithelial sodium channel in the collecting tubule, leading to sodium retention and potassium wasting. Aldosterone and renin levels are low. In Bartter's syndrome there is hypokalaemia but no hypertension, despite high aldosterone and renin levels. Several gene mutations lead to this syndrome, all producing defective sodium chloride transport in the kidney. Gitelman's syndrome is a further variant associated with hypomagnesaemia and hypocalciuria. Salt wasting in these syndromes leads to hyperaldosteronism.

17

FURTHER READING ON DISEASES OF THE ADRENAL CORTEX Fraser R, Murray G D, Connell J M C 1998 Conn's syndrome: no longer a needle in a haystack? Clin Endocrinol 49: 709-710. Hughes I A 1998 Congenital adrenal hyperplasia - a continuum of disorders. Lancet 352: 752-754. Latronico A C, Chrousos G P 1997 Adrenocortical tumors. J Clin Endocrinol Metab 82: 1317-1324. Mantero F, Arnaldi G 1999 Investigation protocol: adrenal enlargement. Clin Endocrinol 50:141-146. Newell-Price J, Grossman A 1999 Diagnosis and management of Cushing's syndrome. Lancet 353: 2087-2088. Oelkers W 1996 Adrenal insufficiency. N Engl J Med 335:1206-1212.

937

Peppercorn P D, Grossman A B, Reznek R H 1998 Imaging of incidentally discovered adrenal masses. Clin Endocrinol 48: 379-388. Ross R J M, Trainer P J 1998 Endocrine investigation: Cushing's syndrome. Clin Endocrinol 49:153-155. Scheinman S J, Guay-Woodford L M, Thakker R V, Warnock D G 1999 Genetic disorders of renal electrolyte transport. N Engl J Med 340:1177-1187. Speiser P W, White P C 1998 Congenital adrenal hyperplasia due to steroid 21-hydroxylase deficiency. Clin Endocrinol 49: 411417.

DISEASES OF THE ADRENAL MEDULLA Adrenaline (epinephrine) is the major product of the normal adrenal medulla and adrenaline in the circulation arises from this gland. Under basal conditions, plasma noradrenaline (norepinephrine) is derived mainly from the sympathetic nervous system, but large amounts can be released rapidly with adrenaline by the adrenal medulla on stimulation. These hormones are of particular importance in the acute response to stress, causing increased alertness, tachycardia, nervousness, glucose and fat mobilization, and inhibition of insulin secretion.

ADRENAL MEDULLARY INSUFFICIENCY There are no demonstrable effects of adrenal medullary insufficiency, e.g. after bilateral adrenalectomy. The sympathetic nervous system alone suffices to maintain normal catecholamine-dependent functions.

ADRENAL MEDULLARY EXCESS Phaeochromocytoma Phaeochromocytomas are tumours of the adrenal medulla and are, fortunately, rare. The usual figure quoted is 1 per 1000 cases of hypertension; however, in a large autopsy series from the Mayo Clinic, three-quarters of the patients found to have had phaeochromocytoma were undiagnosed during life, so the true incidence may be considerably higher. Death in cases previously undiagnosed is commonly from cardiovascular disease or cerebral haemorrhage. The tumours are usually located in the adrenal medulla, 10% being bilateral; 10% are located in extraadrenal chromaffin tissue, usually in the sympathetic chain in the abdomen. Most predominantly secrete

1

938

MCQ 17.21

noradrenaline. Ten per cent of all phaeochromocytomas are malignant. They can occur at any age but present most commonly between 40 and 60 years, and are slightly more common in women. Phaeochromocytomas are associated with neurofibromatosis, von Hippel-Lindau syndrome (p. 1341) and multiple endocrine neoplasia type 2 (see below). Clinical features Presentation may be with sustained, difficult-to-treat hypertension, hypertensive crisis or 'attacks' of headache, sweating, palpitations, flushing and nervousness. The blood pressure is generally moderately raised and becomes grossly so during 'attacks', which may be precipitated by exercise or abdominal palpation. Because the hypertension is accompanied by secondary volume depletion, there is commonly a significant drop in blood pressure on standing; this is unusual in untreated hypertension from other causes. Investigation The diagnosis of phaeochromocytoma is made by measurement of plasma levels of adrenaline and noradrenaline, or their urinary precursors or metabolites - vanillylmandelic acid, metanephrines and normetanephrines. The paroxysmal nature of the secretion may lead to normal levels being obtained at times, and repeated blood and urine measurements may therefore be necessary. The tumour can be localized by adrenal CT scan, but MRI is superior as the T2-weighted image is unusually intense. Radioisotope scanning with 131I-MIBG is helpful to demonstrate an ectopic site. Management The principles of treatment are to start with pharmacological measures, namely adrenergic blockade (initially with the cc-adrenoreceptor blocking agent phenoxybenzamine 10-20mg twice daily or doxazosin 1-2mg twice daily), followed a few days later by B-blockade (e.g. with propranolol 10 mg three times daily, and titrated to control the pulse rate). At least 2 weeks should be allowed for volume repletion before surgery. During surgery, blood pressure is closely monitored and any elevation treated with an intravenous infusion of phentolamine (an oc-adrenoreceptor antagonist). 1

Multiple endocrine neoplasia (MEN) There are at least three different syndromes associated with multiple neoplasia of endocrine glands (Table 17.19); in each case inheritance is autosomal dominant. The involved glands may be the subject of hyperplasia, adenoma or carcinoma development, and different glands may be involved in different individuals in the same family. In MEN type 1, hyperparathyroidism due to parathyroid hyperplasia is present in 95 % of cases at the time of diagnosis. Major morbidity and mortality result from the de-

TABU 17.19 Multiple endocrine neoplasia (MEN)

FURTHER READING ADRENAL MEDULLA

DISEASES OF THE

Major glands/areas involved

Clinical features

Type

1 (Autosomal dominant)

Parathyroid Pancreas

Hypercalcaemia Insulinoma Gastrinoma Glucagonoma

Anterior pituitary

Prolactinoma Acromegaly Cushing's syndrome

Bronchus, bowel

Carcinoid tumours Lipomas

2a (Autosomat dominant)

Thyroid (C cell) Adrenal medulla Parathyroid

Medullary carcinoma Phaeochromocytorna Hypercalcaemia

2b (Autosomal dominant)

Thyroid (C cell) Adrenal medulla

Medullary carcinoma Phaeochromocytorna

ANATOMY AND PHYSIOLOGY

Tongue etc.

Mucosal neuroma Martanoid habitus

The testes have a crucial early role in the male fetus in determining that primary sex differentiation is male. After puberty they serve the dual function of spermatogenesis and androgen (testosterone) production. The initial gonadotrophin stimulation in utero comes principally from the placental hormone human chorionic gonadotrophin (hCG). However, during and after puberty the stimulation comes from the pituitary via its two gonadotrophins, FSH and LH.

velopment of pancreatic tumours, notably insulin-secreting islet cell tumours causing hypoglycaemia, and gastrinomas causing intractable peptic ulceration (Zollinger-Ellison syndrome). Pituitary tumours (prolactinomas, GH or ACTH secreting) are also common, and carcinoid tumours too are recognized as part of the syndrome. The gene responsible is located on chromosome 11 and has been given the name menin, but its function is unknown. Its identification will greatly help family screening. In type 2 MEN the components include medullary cell thyroid carcinoma (calcitonin secreting) and phaeochromocytoma. Hyperparathyroidism is found in 20% of patients with MEN 2a but not MEN 2b. The latter patients uniquely have mucosal neuromas. MEN 2a and 2b are caused by separate point mutations in the RET protooncogene on chromosome 10. Similar mutations to MEN 2a occur in some patients with familial medullary thyroid carcinoma and no other features of MEN 2a, so presumably other genes are involved in MEN. Once MEN 1 or MEN 2 has been identified, first-degree relatives should be offered screening. This is biochemical at present in MEN 1 and is based on annual measurements of calcium, prolactin and gut hormones, but DNA screening may be possible if the menin gene mutation is known. Unfortunately, mutations are diverse, which makes such screening difficult. Genetic screening is now possible in MEN 2a and 2b, allowing thyroidectomy before the emergence of medullary carcinoma of the thyroid. DNA testing for MEN 2 should be done in infancy, as medullary carcinoma can be fatal once established (preceded by years of C-cell hyperplasia).

17

Brandi M L, Gagel R F, Agneli A, Bilezikian V P, Beck-Peccoz P, Bordi C et al 2001 Consensus Guidelines for diagnosis and management of MEN type 1. J Clin Endocrinol Metab 86:5658-5671 Bravo E L 1994 Evolving concepts in the pathophysiology, diagnosis, and treatment of pheochromocytoma. Endocrine Reviews 15: 356-368 Eisenhofer G, Lenders J W M, Linehan W M, Walther M M, Goldstein D S, Keiser H R 1999 Plasma normetanephrine and metanephrine for detecting pheochromocytoma in von Hippel-Lindau disease and MEN2. N Engl J Med 340:1872-1879. Thakkar R V 1998 Multiple endocrine neoplasia - syndromes of the twentieth century. J Clin Endocrinol Metab 83: 2617-2620.

THE TESTES

The testes in utero: sex differentiation After the primitive germ cells have been joined by steroidproducing cells, the primitive 'indifferent' gonads develop in fetuses of both sexes. Much of the current knowledge on the role of the testes was obtained more than 50 years ago, in pioneering experiments by the French biologist Alfred lost on rabbit fetuses. The Y chromosome contains a testisdetermining gene (SRY) that controls differentiation at 6 weeks into a testis. Shortly thereafter, two pairs of duct systems, the Miillerian and the Wolffian, extend from the region of the gonads to the genital groove. The basic pattern of sex differentiation, which requires no hormonal function for its persistence, is female. In the absence of testes the Miillerian ducts persist and become the uterus and fallopian tubes, and the Wolffian ducts disappear. The Sertoli cells of the developing testes produce a peptide hormone, the anti-Miillerian hormone, which causes the Miillerian ducts to disappear permanently. The major pathway of steroidogenesis that leads to testosterone production occurs in the Leydig cells and is shown in Figure 17.28 (p. 927). At around 12 weeks testosterone is secreted, under stimulation from hCG, and acts on the Wolffian ducts to ensure they persist to form the male internal genital ducts. If testosterone is absent at this time, the duct cells die.

939

The development of the external genitalia is also dictated by androgen secretion from the Leydig cells between weeks 10 and 20. Full virilization involves fusion of the labioscrotal folds to form a scrotum, enlargement of the genital tubercle to become a penis, and incorporation into the latter of a penile urethra. Both these effects and the development of the prostate require conversion of testosterone to dihydrotestosterone; this occurs by local action of the enzyme 5a-reductase in target tissues. Later in utero, at around 20 weeks, pituitary secretion of LH and FSH commences. hCG secretion reaches a peak at around 12 weeks and subsequently declines.

The testes in infancy and childhood Late in utero or shortly after birth the testes descend to the scrotum. During the first 6 months of extrauterine life, there is a transient rise in plasma levels of LH and secretion of testosterone. Because levels of the plasma-binding protein sex-hormone-binding globulin (SHBG) also rise during the first week of life, plasma testosterone levels in early infancy are in the low adult male range (although free testosterone levels are lower). Gradually, the activity of the hypothalamopituitary-testis axis subsides and remains low throughout childhood.

form tight junctions which ensure that the germ cells and contents of the tubule are kept secluded by a very effective 'blood-testis barrier'. During puberty the systemic secretion of testosterone initially occurs only at night, but later, as LH levels rise during the day, throughout the 24 hours. Testosterone and, in some tissues, dihydrotestosterone lead to the progressive development of secondary sexual characteristics. In humans these include the enlargement and development of external and internal genitalia; male pattern hair growth; muscular development; growth of long bones followed by closure of the epiphyses; and male sex drive and sexual behaviour, including nocturnal emissions and masturbation. There is no evidence that homosexuality is in any way related to abnormal sex hormone levels. Oestradiol is also produced by conversion from testosterone, in amounts otherwise sufficient to cause breast development,

The testes in puberty The first endocrine event of puberty is the onset of pulsatile secretion of LH during sleep; this initiates pulsatile nocturnal secretion of testosterone (Table 17.20). FSH levels also rise, which, together with testosterone from the Leydig cells, stimulates spermatogenesis. During the course of puberty the volume of the testes increases some 20-fold, from about 1mL to 15-25mL each; 95% of the volume consists of the tubules and their contents. The intersitital (Leydig) cells are the sole source of de novo steroidogenesis, which occurs under stimulation of LH (Fig. 17.34). The action of LH is via specific cell surface receptors, mediated principally by cyclic AMP. FSH also acts via surface receptors and cyclic AMP, but on the Sertoli cells. The Sertoli cells are now known to be crucial in the control of spermatogenesis. Beyond the stage of type A spermatogonia, the germ cells are enclosed in the cytoplasm of Sertoli cells (Fig. 17.35). Adjacent Sertoli cells

FIG. 17.34 Regulation of testicular function

TABLE 17.20 Actions of LH and FSH Effect on ovary

940

Hormone

Effect on testis

Follicular phase

Midcycle

Luteal

LH

Stimulates Leydig cells to produce testosterone, which has local and general effects

Thecal cell androgen secretion

Midcycle surge induces ovulation and luteinization

Sustains corpus iuteum until hCG from blastocyst takes over

FSH

Stimulates Sertoli cells. With local testosterone, regulates spermatogenesis

Stimulates next cohort of follicles, and growth of dominant follicle

Small midcycle surge possibly helps ovum maturation

FSH prepares next crop of follicles if no pregnancy

TABLE 17.21 Features of precocious puberty

17

• Pubertal development before the age of 8 years in girls and 9 years in boys. • The commonest cause is idiopathic central precocious puberty, which affects girls and is due to abnormal sensitivity to GnRH. • Central precocious puberty may also be secondary to cerebral tumours, hydrocephalus, trauma and radiotherapy. • Other causes include hypothyroidism, congenital adrenal hyperplasia, adrenal and gonadal tumours and the McCune-Albright syndrome (due to mutation of the Gsa protein, causing pigmented skin patches, polyostotic fibrous dysplasia and hormone hypersecretion). • Premature thelarche is isolated premature breast development in girls without other signs of puberty; the cause is unknown. • Precocious puberty causes major social and psychological disturbance and retards final adult height. • Treatment is with steroid antagonists (cyproterone acetate, ketoconazole, testoiactone) or GnRH analogues, which prevent pulsatile endogenous GnRH release. FIG. 17.35 Relationship between testicular tubules (cut in crosssection) and interstitial (Leydig) cells Only type A spermatogonia lie outside the Sertoli cells (blue) and so outside the blood-testis barrier.

but testosterone normally inhibits this. Plasma SHBG levels decline by about 50% during male puberty.

The testes in adult life In men there is no evidence of an abrupt decline in gonadal function in old age equivalent to that of the female menopause. This is because there is a constant supply of new germ cells from mitotic division of type A spermatogonia; in women the supply of oocytes declines progressively even from before birth, and eventually none remain.

CLINICAL ASSESSMENT OF TESTICULAR FUNCTION The clinical assessment of testicular function depends upon the age of the subject. A detailed history of sexual activity is important. For example, if a patient has normal nocturnal erections and ejaculation (wet dreams or masturbation), a complaint of lack of potency (erection or ejaculation) during intercourse is unlikely to be due to androgen deficiency. Assessment of testicular size can best be made by reference to an orchidometer - a series of ovoids of known volume. Small testes are usually a sign of damage to the seminiferous tubules. Examination of external genitalia will also determine if there has been any failure of intrauterine virilization. Failure of normal pubertal development may be associated with a degree of growth failure or, conversely, with excessive growth of the long bones before the epiphyses close, so that span exceeds height. Female fat distribution, gynaecomastia and failure

of development of secondary sexual hair are also features of failure of pubertal development, but the latter is most difficult to interpret in view of the great inherited variability in potential for body hair. Precocious puberty is commoner in girls than boys; clinical features are summarized in Table 17.21. Where the patient is normally virilized and the clinical problem is subfertility, examination for proper intrascrotal descent or previous damage (leading to a small testis or testes), phimosis and varicocele (which may depress the sperm count) is important.

DEFECTIVE TESTICULAR FUNCTION IN UTERO Pseudohermaphroditism is the appearance of ambiguous genital ducts and/or external genitalia in the presence of a gonad that is exclusively ovary or testis. True hermaphroditism is much rarer and is due to sex chromosome abnormalities that give rise to both ovarian and testicular tissue in the gonads. Defective testicular function in utero leads to male pseudohermaphroditism. The major causes are listed in Table 17.22. The defects in testosterone and dihydrotestosterone biosynthesis are inherited as autosomal recessive disorders and are extremely rare. They are discussed below, under disorders of sex differentiation (p. 949).

DEFECTIVE TESTICULAR FUNCTION IN ADULT LIFE Male infertility without androgen deficiency In at least one-third of couples presenting with infertility the condition is attributable largely or entirely to abnor-

941

TABLE 17.22 Causes of male pseudohermaphroditism

TABLE 17.23 Causes of male Infertility

XY gonadal dysgenesis

Level of dysfunction

Cause

Testicular enzyme deficiencies With adrenal enzyme deficiency Side-chain cleavage enzyme 3p-hydroxysteroid dehydrogenase 17a-hydroxylase Without adrenal deficiency 17, 20-lyase 17B-hydroxysteroid oxidoreductase

Pituitary/ hypothalamic

Gonadotrophin deficiency Primary*1 Secondary, e.g. tumour**, radiation etc. Gonadotrophin suppression Exogenous androgens e.g. body builders* Endogenous, e.g. 21-hydroxylase deficiency*

Testicular

Chromosomal defect, e.g. Klinefelter's syndrome1 Germ cell damage, e.g. Sertoli cell only syndrome Undescended testes Radiotherapy1 Drugs: cytotoxic, sulphasalazine* Temperature-induced damage: varicocele*, pyrexia* Castration1 Anorchia1

Post-testicular

Absent or obstructed vasa Vasectomy* Epididymal damage Prostatitis, orchitis*

Ejaculatory failure

Erectile failure: psychogenic, vascular, neurogenic Retrograde ejaculation Spinal cord or cauda equina damage

Spermatozoal dysfunction

Infection* Antisperm antibodies Failure to penetrate mucus/poor viability Miscellaneous, e.g. defects in sperm structure

Target organ abnormalities 5a-reductase deficiency Androgen receptor defect - testicular feminization, complete or partial

malities in the male, leading either to azoospermia, oligospermia or abnormal sperm. A general classification of causes is given in Table 17.23. Management is in most cases unsatisfactory and is beyond the scope of this book. In general, if the serum level of FSH is elevated in a patient with azoospermia or severe oligospermia, severe irreversible testicular damage will be present and no treatment will be effective. With subnormal semen, in vitro fertilization may be useful in increasing the chances of conception, possibly because problems of poor sperm motility are overcome.

Androgen deficiency (male hypogonadism) The presentation of testicular dysfunction with androgen deficiency may be at the time of puberty, or much later. A few men presenting with infertility have a mild degree of androgen deficiency and associated oligo- or azoospermia; most such men, and all who are severely androgen deficient, are sexually inactive. The causes of androgen deficiency, classified according to whether the gonadotrophin levels are low (hypogonadotrophic) or high (hypergonadotrophic), are given in Table 17.24. In both groups the severity of the androgen deficiency can vary considerably. It is usually least severe in Klinefelter's syndrome 1 (which in most cases is associated with an XXY chromosomal constitution), where the Leydig cells continue to function (although subnormally) despite severely damaged tubules. The incidence is about 1 per 1000 men. The testes are usually firm and pea-sized. Gynaecomastia may occur, as in any form of hypogonadism, owing in part to increased oestrogen but mainly to reduced testosterone

1

942

Fig 17.29

* Partially or totally treatable or reversible Associated with androgen deficiency and therefore also with infrequent or absent intercourse.

action on the breasts. Typically the limbs continue to grow, so that the final height is increased, with long extremities ('eunuchoidal proportions', Fig. 17.36). Isolated gonadotrophin deficiency (owing usually to GnRH defficiency) may be sporadic or familial (usually X-linked). When associated with anosmia, it is termed Kallmann's syndrome. The anosmia is due to hypoplasia of the olfactory lobes. When hypogonadism develops after puberty it is highly likely that a pituitary tumour or craniopharyngioma is responsible, a suspicion that can be confirmed by appropriate radiology and hormone assays. A common clinical problem is the teenage boy with delayed puberty. The differential diagnosis between simple delayed puberty and permanent isolated gonadotrophin deficiency may be impossible until age 18, when failure of gonadotrophins and testosterone to rise confirms the latter. For psychological wellbeing boys over 14 with delayed puberty often benefit from 6-month courses of testosterone (e.g. Sustanon 50-100 mg i.m. every 4 weeks,

TABLE 17.24 Causes of androgen deficiency in phenotypically normal men

These tests are generally no more informative than basal hormone measurements.

Gonadotrophin levels

Cause of androgen deficiency

Management

Low LH and FSH

Simple delayed puberty Illness and malnutrition Anorexia nervosa Isolated gonadotrophin/GnRH deficiency Normal sense of smell Anosmia (Kallmann's syndrome) Panhypopituitarism Drugs - cyproterone acetate*

High LH and FSH

Developmental Klinefelter's syndrome (XXY and its variants)

Male infertility Semen analysis should be the first investigation in an infertile couple when the woman has regular periods. The principal treatable causes of male infertility (Table 17.23) are oligospermia due to varicocele, or exposure of the testes to high temperature (the scrotum keeps the normal testes at about 32°C and prolonged high temperature impairs spermatogenesis). Sulphasalazine-induced infertility in ulcerative colitis can be treated by transfer to 5-aminosalicylic acid. Gonadotrophin deficiency from any cause may lead to some permanent testicular damage. In most such cases adequate spermatogenesis can be stimulated by administering hCG, with the later addition of FSH (as human menopausal gonadotrophins). An adequate trial of therapy is 3-6 months, the response being followed by semen analysis (count and motility). An alternative is to administer GnRH by subcutaneous pump in a pulsatile manner (15 ug every 90 minutes) but this is not clearly better than hCG plus FSH. Where gonadotrophin deficiency is caused by extratesticular androgens (as in congenital adrenal hyperplasia) the appropriate treatment is with replacement glucocorticoid therapy.

Acquired Trauma Anorchia (testicular degeneration) Viral orchitis (mumps) Radiation Drugs, e.g. cyclophosphamide Dystrophia myotonica Variable LH and FSH

Prader-Willi syndrome Hyperprolactinaemia Noonan's syndrome Cirrhosis of liver Drugs, e.g. spironolactone, alcohol Cushing's syndrome

* Gonadotrophin suppression combined with androgen resistance.

or oral testosterone undecanoate 40mg 2-4 times daily). Evidence of spontaneous puberty can be assessed during a 4-6-month period off treatment. Usually only one or two courses are required, and final height is not significantly affected.

Investigation of endocrine function of the testes A great deal can be learned about the endocrine function of the testes from a single measurment of the serum testosterone (especially when combined with measurement of sex-hormone-binding globulin, to allow determination of the free androgen index) plus the gonadotrophins LH and FSH. A grossly raised prolactin points to a prolactinoma, which, although rare in men, is important to diagnose as early as possible as such tumours are often rapidly growing. If gynaecomastia is present, serum oestradiol should be included in the assessment. Stimulation tests can be carried out at the level of the pituitary (with GnRH 100 ug i.v.) or the testes (with hCG 1500-5000IU i.m.), or oestrogen negative feedback can be assessed (with clomiphene citrate).

17

Androgen deficiency Testosterone esters (e.g. Sustanon 250 mg i.m. every 2-4 weeks) are the most practical form of replacement, although testosterone patches are now available and provide more physiological levels. To induce puberty, smaller doses (e.g. 50 mg) should be used and gradually increased. An alternative is to implant testosterone pellets, under local anaesthetic, subcutaneously in the lower abdominal wall, but this is now little used. Oral testosterone undecanoate is less effective than i.m. testosterone esters.

Impotence Around 2-5% of men aged 40 are impotent; this increases to around 15% by age 60 and continues to rise thereafter. Changing attitudes to sexuality have enhanced both general awareness of this problem and the patient's expectations for successful treatment, now fortunately matched by advances in therapy. The main causes of impotence are . detailed in Table 17.25; psychogenic impotence is considered in Chapter 8. A detailed sexual history and examination generally reveal a likely cause and initial investigation should be limited to full blood count, plasma glucose, serum creatinine and electrolytes, liver function tests and testosterone, unless the history and examination point to a rare cause of impotence such as thyrotoxicosis. In specialized centres, direct measuring of penile erection, particularly sleep-related circumference increase, is employed to differentiate between psychogenic impotence (in which

943

FIG. 17.36 Testicular dysfunction with androgen deficiency Grossly eunuchoidal 31-year-old man with Kallmann's syndrome [A] before and [£] after 2 years of androgen replacement. C A 21-year-old man with Klinefelter's syndrome. Note partial virilizaton and minimal gynaecomastia.

TABLE 17.25 Main causes of impotence Psychogenic

Depression, stress

Vascular Neurological Genitourinary

Aortic or iliac arteriosclerosis, hypertension Autonomic neuropathy, spinal cord damage Micropenis or hypospadias, Peyronie's disease, phimosis, prostatic, retroperitoneal or rectal surgery, pelvic fracture

Drugs

Antihypertensive agents, antidepressants, sedatives, alcohol, antiandrogens, cocaine, amphetamines, liquorice Hypogonadism (primary or secondary), hyperprolactinaemia, oestrogen-secreting tumours, thyrotoxicosis, hypothyroidism, Cushing's syndrome Cardiac, hepatic, renal, lung Due usually to a combination of the above factors

Endocrine

Organ failure Age

spontaneous erections occur during sleep) and organic causes (in which all erections are lost). Unfortunately, in around a quarter of patients it is difficult to differentiate between these two broad categories. Apart from diabetes mellitus, endocrine causes of im1

944

MCQ 17.22

potence are rare and can usually be excluded by finding a normal testosterone level. Low testosterone warrants further investigation to determine the cause, as detailed in Table 17.24. A prolactin-secreting tumour in particular must be excluded. When testosterone deficiency is the cause of impotence, replacement usually restores normal function. There is no place for testosterone treatment of impotent men with normal levels of testosterone, although some patients may have a placebo response, and testosterone is contraindicated if a patient has prostatic or breast cancer. When impotence is psychogenic, or related to vascular or neurological disorders, intracavernosal injections of prostaglandin E1, alone or with papaverine and phentolamine, are effective in around two-thirds of patients. Age does not seem to alter efficacy. The main side-effects are pain, plaque formation and priapism. A refinement of this is to administer the prostaglandin as a gel inserted into the urethra. However, the most important development is the oral drug sildenafil, which relaxes the smooth muscle of the corpus cavernosum by enhancing nitric oxide levels. Caution is required in patients with cardiovascular disease. In selected cases psychotherapy, external vacuum devices or implanted inflatable penile prostheses have been successful. With all of these therapies, specialist counselling (often undertaken by urologists) is important to achieve optimum results. 1

FURTHER READING ON THE TESTES

17

Albanese A, Stanhope R 1995 Investigation of delayed puberty. Clin Endocrinol 43:105-110. de Kretser D M 1997 Male infertility. Lancet 349: 787-790. Hawton K 1998 Integration of treatments for male sexual dysfunction. Lancet 351: 7-8. Elders M J, Scott C R, Frindik J P, Kemp S F 1997 Workup for precocious puberty. Lancet 350: 457-458. Fabbri A, Aversa A, Isidori A 1999 Sildenafil and erectile dysfunction. J Endocrinol Invest 22: 486-492. National Institute on Aging Advisory Panel 2001 Special report. Report of National Institute on Aging Advisory Panel on testosterone replacement in men. J Chin Endocrinol Metab 86:4611-4614.

THE OVARIES

ANATOMY AND PHYSIOLOGY The ovaries consist of germ cells surrounded by granulosa cells (the homologue of the Sertoli cell); these are surrounded in turn by stromal cells of the theca interna. The ovaries have no role in female genital development, which simply requires the absence of testes. They secrete little steroid before birth, and so FSH and LH are higher in female fetuses. After birth and a transient surge in LH levels, the hypothalamopituitary-ovarian axis becomes quiescent until puberty.

Changes during puberty On average, puberty starts about 2 years earlier in girls than in boys. The full complement of ovarian follicles (about half a million) is complete at birth, and their number declines progressively thereafter, until exhaustion at the time of the menopause. Crops of immature follicles advance to a certain stage and become atretic even before puberty. With the rise in FSH levels, these crops of follicles produce increasing amounts of oestrogen; at about 7 days before ovulation, one follicle becomes sensitive to FSH and LH (via increased numbers of receptors) and makes more oestradiol (Fig. 17.37). This suppresses FSH and leads to atresia of the less advanced follicles in that cohort. There is good evidence that the first cycles in girls are anovulatory, and they may be associated with erratic 'breakthrough' menstrual bleeding, owing to oestrogenic stimulation of the endometrium. Eventually, however, conditions are right for the first ovulatory cycle.

Ovulation and the normal menstrual cycle The biological purpose of ovulation is pregnancy. Menstruation may be thought of as biological failure and the starting point of the next cycle, whose purpose is once again to achieve the release of a single fertilizable ovum

FIG. 17.37 The normal menstrual cycle [A] The sequence of events of the normal menstrual cycle. (1) FSH leads to follicle maturation and secretion of oestradiol, which leads to proliferative endometrium and, with other factors, (2) feedback on pituitary and hypothalamus. The midcycle positive feedback triggers LH surge (3), leading to follicle rupture and luteinization. Corpus luteum (4) secretes progesterone and oestradiol, leading to a secretory endometrium. If conception occurs, the implanted blastocyst secretes hCG, which maintains the corpus luteum and its secretions. B Overall pattern of circulating hormone levels during a normal menstrual cycle.

and to prepare the reproductive tract for pregnancy. The menstrual cycle depends on inherent cyclicity within the ovary, the hypothalamus and the pituitary gland. Defects in this complex system are common. Figure 17.37 shows the stylized profile of hormone levels obtained if blood samples are taken at daily intervals throughout a normal cycle. Oestradiol exerts predominantly negative and intermittently positive feedback. When follicular oestradiol secretion rises, this appears to

945

switch the pituitary gonadotrophs from an FSH- to a mainly LH-secreting mode, and to reduce the pulse amplitude of GnRH secretion. Meanwhile, the gonadotrophs become more sensitive to GnRH. The signal for the preovulatory surge in LH secretion is a decline in oestradiol levels, together with the onset of progesterone secretion with an attendant rise in GnRH pulse amplitude. The midcycle LH surge, characterized by 24-48 hours of high peaks of LH secretion, is one of the turning points of the cycle. It arrives at the ovary when one follicle has reached a diameter of 2cm, and its granulosa cells have proliferated and developed LH receptors sufficient to be induced to ovulate. The LH surge completes ovulation, and induces the granulosa cells to become the corpus luteum. Postovulatory steroid secretion, which is still dependent on low concentrations of LH, comes principally from the corpus luteum, whose cells have developed the capacity for de novo steroid synthesis and secrete both progesterone and oestradiol. It should be noted that plasma oestradiol levels fluctuate 10-fold and progesterone more than 20fold during the normal menstrual cycle. In the uterus the changes are first, proliferative-phase development, due to the preovulatory oestradiol surge; and second, postovulatory progesterone and oestradiol conversion of this to the receptive secretory-phase endometrium. Mucus secretion from the cervix is only conducive to sperm penetration during the late proliferative phase and midcycle; thereafter, progesterone rapidly inhibits it. If fertilization occurs, the blastocyst penetrates the endometrium and secretes its own LH-like hormone, hCG, which is absorbed into the circulation and maintains the corpus luteum. In the absence of hCG the corpus luteum atrophies, progesterone secretion declines, the spiral arteries of the endometrium collapse and the surface layers of endometrium die and are shed as menstrual blood.

Secondary sexual changes during and after female puberty Breast development commences at around 9-11 years. Primary changes are caused by the action of oestradiol on the duct epithelium; this in turn probably stimulates stromal fat development (most of the bulk of the breast). Progesterone also acts on the oestrogen-primed breast epithelium, which is important in pregnancy to prepare for lactation. Progesterone is responsible for the increase in breast volume (about 20%) during the luteal phase. Oestradiol, and ovarian and adrenal androgens are responsible for the stimulation of female pubic and axillary hair growth. Changes in uterine size (myometrial and endometrial) are caused by the joint effects of oestrogen and progesterone.

Oestradiol is principally responsible for the pubertal growth spurt by a direct effect on epiphyseal cartilage, and an indirect effect on GH and gonadotrophin secretion. The earlier time of puberty and the potent effect of oestrogens on epiphyseal plates leads to the shorter final stature of women compared to men. Oestradiol has a major effect in increasing endosteal new bone formation and skeletal mass, but is less potent in this respect than testosterone. The increased muscle mass in males in probably also a factor in increasing the bone density (cortical and trabecular) in men (see also p. 954).

Age-related changes in ovarian function and the menopause The menopause is preceded by hormone changes over 10 years or more, which stem from the progressive depletion of follicles, causing deeper troughs in oestradiol levels and higher peaks of FSH. The role of the ovarian peptide hormone inhibin is still not entirely clear. High FSH levels stimulate premature development of the next crop of follicles. Luteinization without ovulation is commonplace; bleeding becomes erratic and women are subfertile long before the last menstrual period, which usually occurs at about 50 years of age. The menopause is commonly attended by frequent disabling and embarrassing hot flushes, owing to central sympathetic nervous discharge. The flushes coincide with discharge of LH and FSH, but are not caused by them. Vaginal dryness, lack of libido, breast atrophy, urethritis and bone loss are all common features of postmenopausal oestrogen deficiency.

OVARIAN DYSFUNCTION: AMENORRHOEA AND OLIGOMENORRHOEA Ovarian dysfunction is usually associated with erratic and infrequent cycles (oligomenorrhoea) or absent periods (amenorrhoea). The presenting complaint may be of the menstrual disorder, infertility, or accompanying problems such as breast discomfort, galactorrhoea or hirsutism. Amenorrhoea or oligomenorrhoea may occur from the time of puberty (primary) or develop later (secondary). It is obviously always important to exclude pregnancy by a pregnancy test or ultrasound scan, even in women who ovulate only occasionally. It is especially important in patients with primary amenorrhoea to try to determine whether the external and internal genitalia are those of a normal female. Pituitary/hypothalamic causes

1 946

MCQ 17.23

Gonadotrophin deficiency Deficiency of GnRH is most commonly secondary to weight loss. Many teenage girls diet excessively and, as a

consequence, develop amenorrhoea at a weight or lean body mass that varies between different individuals. In anorexia nervosa (Ch. 8, p. 239) circulating LH and FSH levels are low, as are those of oestradiol and progesterone. As a consequence, in addition to amenorrhoea there is a degree of regression of secondary sexual characteristics towards a prepubertal state, with a reduction in oestrogenization of the breasts and genital tract. A variable degree of osteoporosis will occur in the long term. A similar state can result from weight loss or starvation, or in athletes (especially long-distance runners) and ballet dancers: the periods usually, but not always, return if and when weight is regained. Isolated gonadotrophin deficiency without weight loss (due almost certainly to GnRH deficiency) may occur as a selective hypothalamic defect in the female, as in the male. If associated with anosmia, it is called Kallmann's syndrome (p. 942). It may also be seen in patients with craniopharyngioma, or after pituitary irradiation, in which case it is likely to be associated with other pituitary hormone deficiencies. Proloctin excess PRL excess from any cause is frequently associated with disordered ovarian function, with amenorrhoea occurring when the PRL level rises to more than two or three times above normal. PRL excess is believed to exert a 'short loop feedback' effect at hypothalamic level, reducing the pulse frequency of GnRH secretion. A classification of the causes is given in Table 17.7 and the condition discussed on pages 907-909.

Ovarian causes The ovarian causes of amenorrhoea and oligomenorrhoea are summarized in Table 17.26. In such cases the circulating levels of LH and FSH are elevated. Premature menopause Menopause is considered premature below the age of 40, and this occurs in 1 % of women. Occasionally the immune system may be implicated, with the production of ovarian antibodies (p. 889). An identical syndrome often results from irradiation of the ovaries or the administration of chemotherapeutic agents, usually directed against Hodgkin's lymphoma or other malignancies (p. 161). Gonadal dysgenesis An important group comes under the general heading of gonadal dysgenesis. Here, the ovaries may either fail to develop or degenerate before puberty, often because of an abnormal X chromosome complement, which leads to follicular death. Turner's syndrome Turner's syndrome (see Fig. 3.15, p. 68) is the commonest form, with a prevalence of around 1 in 5000 female births. Primary amenorrhoea is associated with short stature and a variable number of skeletal and somatic abnormalities

TABLE 17.26 Ovarian causes of amenorrhoea or oligomenorrhoea

17

Failure of normal ovarian development Pure XX gonadal dysgenesis Familial Sporadic Turner's syndrome (XO) Ovarian enzyme deficiencies (e.g. 17, 20 lyase) Premature ovarian failure (follicular degeneration) Idiopathic (low initial follicle number) Autoimmune (antiovarian antibodies) Infection (e.g. mumps oophoritis) Radiotherapy Cytotoxic drugs (e.g. for Hodgkin's disease, leukaemia) Surgical/traumatic Galactosaemia Ovarian dysfunction Polycystic ovary syndrome (possibly ovarian in origin) Resistant ovary syndrome Mutations in gonadotrophin receptors

(Table 17.27; see also Ch. 3, p. 69), but patients vary considerably in their expression of these. The condition is most commonly caused by a 45,X (termed XO) chromosome constitution, which results from the loss of an X or Y chromosome before the first mitotic division of the fertilized ovum. The features of this syndrome result in part from the existence of only one X chromosome (without a Y), and in part from the resulting sex hormone deficiency from the time of puberty onwards. Other forms of gonadal dysgenesis Some patients with a classic Turner's phenotype (Fig. 3.14) have a normal chromosome complement but have part of an X chromosome missing. With XO/XY mosaicism a range of clinical features, from Turner's syndrome to male genitalia, may be present. These patients frequently develop gonadal tumours, and prophylactic removal of dysgenetic gonads is indicated. Rare individuals have an apparently normal male genotype (46, XY) but normal female or ambiguous genitalia (pure XY gonadal dysgenesis; see p. 949). These patients also have an increased risk of gonadal tumours. 1

SUMMARY 12 Polycystic ovary syndrome • • • • • • •

Erratic cycles or amenorrhoea Subfertility Variable obesity and hirsutism Polycystic ovaries on ultrasound Inappropriately high LH Modest increase in androgens Mild hyperprolactinaemia

947

TABLE 17.27 Clinical features of 45, XO Turner's syndrome Type

Features

Sexual

Streak ovaries Sexual infantilism Primary amenorrhoea High gonadotrophins Normal female genitalia

Skeletal

Short stature (<150 cm) Characteristic facies - micrognathia, epicanthic folds, low-set ears, fish-like mouth, high arched palate Shield-like chest Wide carrying angle Short neck, low hairline Short fourth metacarpals Scoliosis Localised areas of rarefaction Osteoporosis

Soft tissues

Lymphoedema Webbing of neck

Cardiovascular

Coarctation of aorta Bicuspid aortic valve Aortic stenosis

Renal

Abnormal position of kidney or horseshoe shape Duplex ureters Hydronephrosis

Cutaneous

Small nipples Pigmented naevi Keloid formation

The polycystic ovary syndrome This was originally described by Stein and Leventhal in 1935 as hirsutism, obesity and oligomenorrhoea associated with enlarged ovaries with a thickened and whitish-looking 'capsule' and multiple follicular cysts. There is persistent elevation of serum LH. Hirsutism occurs in around 70% of affected women, amenorrhoea in 50% and obesity in 40%. It is likely that the cause varies, from hypothalamic, pituitary, ovarian and possibly even adrenal defects. About 20% of normal women have polycystic ovaries on ultrasound, although many of these have no specific signs or symptoms. The high LH levels are generally associated with increased ovarian secretion of androstenedione, which is converted in peripheral tissues to testosterone.

1 948

MCQ 17.24

Plasma oestradiol is relatively constant at levels seen in the midfollicular phase of the cycle; levels of SHBG are reduced, and so the unbound testosterone level is modestly raised. The extent of development of hirsutism and more marked virilization in this syndrome is very variable, and depends both on the hormone abnormality (which can be measured) and on the capacity of target tissues to respond. Likewise, the ovulatory disturbance varies from slightly erratic cycles through to long-standing amenorrhoea (primary or secondary).

Management of amenorrhoea The treatment of patients with erratic cycles depends on the underlying cause, the age of the patient and their requirements (e.g. for contraception, treatment of hirsutism or infertility). With PRL excess, restoring the PRL level to normal is usually sufficient to restore normal ovulation and fertility. If the latter is not desired, then the oral contraceptive pill can be given safely with bromocriptine or cabergoline. With gonadal dysgenesis or ovarian failure, attempts to restore ovarian function and fertility are generally unsuccessful. These women are suitable candidates for ovarian hormone replacement and embryo implantation (donor eggs fertilized by husband's sperm). Sex hormone replacement is indicated in most cases, at least until the time of the menopause. This can be provided by low-dose oral contraceptive pills, most of which provide 20-30 ug daily of ethinyl oestradiol, and a progestogen (21 days out of 28). When oral oestrogens are contraindicated (e.g. a history of venous thrombosis) percutaneous oestrogen can be given. Adhesive plastic patches containing oestradiol dissolved in ethanol (Estraderm), which are applied to the skin, or a gel preparation (Oestrogel) applied to a wide surface area are alternatives. A progestogen (oral or, in the case of patches, percutaneous) is needed in addition for 10-14 days every 1-2 months to prevent endometrial hyperplasia and induce periods. After the menopause, for most women, opinion increasingly favours hormone replacement therapy (HRT). There is little doubt that it protects the bones against osteoporosis and the genital tract against atrophy, and that many women feel much better on it. Cyclical progestogens prevent an increased risk of endometrial cancer. It is not clear by how much HRT increases the risk of breast cancer but the benefits usually outweigh any risk, at least up to 5 years of treatment, and probably longer. 1

Management of female infertility With hyperprolactinaemia, returning the PRL to normal usually restores both periods and fertility. Gonadotrophin deficiency may be treated either by gonadotrophin administration, or by pulsatile administration of GnRH. Both require careful monitoring. Gonadotrophin therapy consists usually of 3-7 days of

injections of a preparation of human FSH. The dose is titrated by measuring blood or urine oestrogens and by ovarian ultrasound scanning. When the right amount of follicular development has occurred, ovulation is induced by an injection of hCG. This treatment is only given in specialized units where there is also an egg retrieval and fertilization programme. Polycystic ovaries. Treatment of the infertility is difficult and often unsatisfactory. If the patient is obese, weight reduction may restore fertility. Clomiphene citrate (an antioestrogen) is often given, 50 mg daily for 5 days, commencing on the fifth day after a period starts, and induces ovulation in about half the cases. Side-effects include multiple pregnancy. Wedge resection of the ovaries, and more recently laser surgery, may be effective in cases resistant to clomiphene. Gonadotrophin therapy can also be used.

ate sex of rearing. Parents are under great pressure to declare the sex of the infant and to stick to it. They require very sensitive handling. Most cases of abnormal sex differentiation are cases of pseudo- (rather than true) hermaphroditism.

17

MALE PSEUDOHERMAPHRODITISM The male pseudohermaphrodite is a genetically male infant who does not have completely normal male sexual characteristics. In its mildest form the infant has failure of development of the penile urethra (hypospadias); in its most severe form the external genitalia appear to be those of a completely normal female. The causes are listed in Table 17.22 (p. 942).

Pure XY gonadal dysgenesis PREMENSTRUAL TENSION Premenstrual tension is a common but rather poorly defined condition in which a wide range of physical and emotional symptoms and signs are worst during the week or two preceding a menstrual period, and generally improve with its onset. It is not primarily an endocrine disorder, but is probably best thought of as a cyclical disorder 'locked in' on the normal fluctuation in hormone levels that accompanies the menstrual cycle. In many women, however, it can be extremely disabling. Treatment may be directed at smoothing out or eliminating the normal hormone changes, by using an oral contraceptive preparation, or at treating the symptoms.

FURTHER READING ON THE OVARIES Adashi E Y, Hennebold J D 1999 Single gene mutations resulting in reproductive dysfunction in women. N Engl J Med 340:709-718. Con way G S 1996 Polycystic ovary syndrome: clinical aspects. Bailliere's Clin Endocrinol Metab 10: 263-279. Greendale G A, Lee N P, Arriola E R 1999 The menopause. Lancet 353:571-580. Kyei-Mensah A A, Jacobs H S 1995 The investigation of female infertility. Clin Endocrinol 43: 251-255. Saenger P, Albertsson Wikland K, Conway G S, Davenport M, Gravholt C H, Hintz P et al 2001 Recommendations for the diagnosis and management of Turner syndrome. J Clin Endocrinol Metab 86: 3061-3069.

DISORDERS OF SEX DIFFERENTIATION The underlying mechanisms of primary sex differentiation are described on page 939. Where sex differentiation is evidently abnormal at birth, the problem is made particularly difficult by the urgency of determining the appropri-

Here, the individual is XY in sex chromosome constitution but testes never imposed maleness on the 'background' female phenotype. Apart from the absence of gonads, the patient is a completely normal female. The gonadal remnants (streak gonads) are at risk of developing tumours (dysgerminomas) at or after the age of puberty, and should therefore be removed. From puberty onwards, female sex hormone replacement should be given.

Testicular feminization Testicular feminization occurs with a frequency of about 1 in 20000 male births and is caused by a heterogeneous group of molecular defects in the androgen receptor, inherited in an X-linked fashion. XY individuals develop testes as normal but are unable to respond to the androgen they produce. The heterogeneity of genetic defects is responsible for the complete and partial forms of the syndrome. In the complete form the external genitalia are those of a normal female, but with a very short blind vagina. Miillerian structures are absent. There is no development either at or after puberty of even the normal female amount of pubic and axillary hair. The testes are in the inguinal canal. Presentation is often with inguinal herniae. The correct sex of rearing for individuals with either the partial or the complete form of the syndrome is undoubtedly female. It is generally felt to be advisable to leave the testes in situ until breast development has taken place at the normal age of puberty. Feminization of the breasts is usually excellent, under the influence of normal levels of testicular oestradiol (unopposed by androgen action). Circulating levels of LH and FSH are usually somewhat elevated, and testosterone levels in the high adult male range (after puberty). Oestradiol levels are also elevated owing to increased testicular production and peripheral conversion from testosterone. The gonads should be removed in early adult life, as there is considerable risk of development of tumours (dysgerminomas) within them.

949

5a-reductqse deficiency 5a-reductase deficiency is an autosomal recessive disorder, and is therefore most common where there has been intermarriage. The primary defect lies in the enzyme 5areductase, which is absent or reduced in activity. The production of dihydrotestosterone (DHT) in androgen target tissues is reduced (Fig. 17.28, p. 927) and those effects that depend on DHT are therefore compromised. Females with the condition appear normal, but males have impairment of virilization of the external genitalia. 1The phallus is small, and there is complete perineoscrotal hypospadias and incomplete fusion of the scrotum, with a blind vaginal opening. The testes, epididymes, vasa and seminal vesicles are normal, but the prostate gland is, and remains, atrophic. Regardless of the sex of rearing, affected individuals generally insist that they are male, both at the time of puberty and beyond, and generally demonstrate male patterns of behaviour. After puberty, the voice deepens and normal male muscular development occurs, but facial hair is absent or scanty.

FEMALE PSEUDOHERMAPHRODITISM In female pseudohermaphroditism a genetic female is partially or completely virilized. The most likely cause is excess androgen secretion in utero, generally as a result of the action of increased ACTH on the (abnormal) adrenal glands. The commonest cause is 21-hydroxylase deficiency (p. 932).

RARE DISORDERS OF STEROID BIOSYNTHESIS Rare deficiencies of almost all the enzymes have been described, and the clinical picture can usually be understood by reference to these pathways and the normal physiology of the glands (Fig. 17.28, p. ••). Conditions that affect the testosterone pathway lead to male pseudohermaphroditism; if they lie early on the pathway (which is shared with the adrenals) then adrenal insufficiency also occurs. In addition to 21-hydroxylase deficiency (which is about nine times more common than the others put together), there are a number of other disorders of steroid biosynthesis that also lead to pseudohermaphroditism in affected females. These include 3B-hydroxysteroid dehydrogenase deficiency, in which the weak A5-androgens from the adrenals cause some virilization in females (but males are inadequately virilized). Likewise, llB-hydroxylase defi-

1

950

Fig. 17.30

2

MCQ 17.25

TABLE 17.28 Causes of gynaecomastia Physiological Neonatal Pubertal Old age 'Refeeding' (after starvation) Pharmacological Spironolactone Digitalis glycosides Cannabis Cyproterone acetate Oestrogen ingestion Alcohol

Pathological Hypogonadism Steroid disorders 17B-hydroxysteroid oxidoreductase deficiency Testicular feminization Cirrhosis Thyrotoxicosis Tumours hCG-secreting Leydig cell Adrenal

ciency is associated with a build-up of adrenal androgens, leading to virilization (p. 932).

DISORDERS OF MALE SECONDARY SEX DIFFERENTIATION The physiological mechanisms that lead to male secondary sex differentiation are discussed on page 939.

Gynaecomastia Gynaecomastia (breast development in the male) is so common in puberty and old age as to be regarded as physiological. It also features in a wide range of disorders (Table 17.28). These are generally associated with an increase in oestrogen production and/or a decline in testosterone production. Some drugs cause gynaecomastia, the most notable being the aldosterone antagonist Spironolactone. This compound also binds to androgen receptors and is an antiandrogen. It also inhibits the secretion of androgen. Digitoxin and the main metabolite of cannabis cause gynaecomastia because of intrinsic oestrogenic activity. Treatment may be directed at the underlying cause, but that is often not apparent. Hypogonadism should be treated with testosterone replacement. Drugs such as danazol and tamoxifen may help if used early, but the best treatment for severe gynaecomastia is surgery. 2

Prostatic hyperplasia Prostatic hyperplasia begins in the fifth decade and results from increased androgen effect (via dihydrotestosterone and its hydroxylated metabolites). It commonly causes urethral obstruction and is usually treated by transurethral resection, although the 5a-reductase inhibitor finasteride (5mg daily) may be helpful in mild to moderate cases, sometimes combined with a-adrenergic blockade.

Prostatic cancer Prostatic cancer is the most common cancer in elderly males and is generally androgen dependent. The management of early-stage disease is with surgery or radiotherapy, depending on stage. For more advanced cases and in metastatic disease, a mainstay of treatment is suppression of androgen secretion and/or action. GnRH analogues such as buserelin, goserelin or leuprorelin are as effective. These agents can cause a transient worsening of disease owing to initial LH stimulation before LH becomes downregulated. Such 'flares' may be prevented by the antiandrogens cyproterone acetate (100 mg two or three times daily) or flutamide (250mg three times daily). These agents may also be used in palliation.

DISORDERS OF FEMALE SECONDARY SEX DIFFERENTIATION Hirsutism in women is the obverse of gynaecomastia in men and is almost equally common. The term virilization is reserved for the much less common androgenization associated with very high levels of testosterone, in or approaching the adult male range. This may be due to ovarian or adrenal tumours, or to the so-called late-onset form of congenital adrenal hyperplasia. Virilization usually declares itself by a rapid development of hirsutism, with other signs such as male pattern muscle development, frontal balding, clitoral enlargement and deepening of the voice. Measurement of serum testosterone is sufficient to exclude a virilizing tumour of the adrenal or ovary. The cut-off point for further investigations is about 5 nmol/L.

Hirsutism Most cases are due to the polycystic ovary syndrome or are idiopathic. In the latter there is probably heightened sensitivity of hair follicles to normal androgen levels. It is evident from the wide variation in the extent of facial and body hair in men that the same genetic variability must also exist in women. Thus, some women with chronic anovulation, erratic cycles, polycystic ovaries and

high levels of testosterone and androstenedione coming from the ovaries under the stimulation of LH show little or no hirsutism, whereas others, with a lesser endocrine defect, are severely hirsute. Another problem is to determine how much of the androgen excess is arising from the ovaries and how much from the adrenals. With the exception of 21-hydroxylase deficiency, the ovaries are usually the major source.

17

Management of hirsutism There is no ideal treatment and a careful explanation of the likely outcome is important. Psychological support is of great importance and must include reassurance that the woman is not losing her femininity, or 'turning into a man'. Sometimes minimal hirsutism is a focus for an anxiety state; conversely, some women react by refusing to accept that there is a problem, and avoid even elementary cosmetic measures. Cosmetic treatment involves the use of creams, waxes (which root out the hairs), shaving or electrolysis. The latter is usually reserved for strong terminal hairs on the face and breasts, or other psychologically important areas, and should only be carried out by a trained electrologist. Contrary to popular belief, shaving does not make unwanted hair grow faster or thicker. Hormonal treatment involves the use of adrenal and/or ovarian suppression and antiandrogens. Adrenal androgens can generally be suppressed with doses of glucocorticoid that are less than those required to suppress cortisol, but this tends to be unhelpful in practice. Ovarian suppression is achieved by using a low-dose oral contraceptive. The antiandrogens include cyproterone acetate, which in low dosage is a component of a contraceptive pill (Dianette); much higher doses (50-100 mg daily) are needed for maximum antiandrogen effect. Spironolactone (50-200 mg daily) is sometimes remarkably effective, but can lead in some cases to menstrual disturbance. Finasteride has recently been reported to be of benefit. Drug treatment does not cure the problem and can take 6-9 months to have an effect. Pregnancy must be avoided while taking these drugs, as male fetal development would be affected. An oestrogen is usually therefore given as a contraceptive and to ensure regular periods. The duration of drug therapy for hirsutism can be openended and determined by the severity of the problem and the response; the patient will simply have the same problem (no worse) when treatment is stopped.

SUMMARY 13 Management of hirsutism • Exclude a virilizing tumour by history and serum testosterone level • Psychological support has considerable importance • Cosmetic treatments are often effective and do not increase hair growth • Antiandrogens such as spironolactone, cyproterone and finasteride are effective in severe cases but require careful monitoring for side-effects

SEX HORMONES AND BEHAVIOUR Androgens, oestrogens and progestogens all affect the individual's behaviour, either directly or indirectly. However, homosexuality, transsexualism and sexual deviation are not directly linked to androgen or oestrogen levels. It has been suggested that XYY men are more violent than normal men, in view of the finding of an increased number

951

in penal institutions; however, the significance of this is uncertain. Androgen-deficient men may develop protective mechanisms and become aggressive to compensate for the lack of physical development. If they are then abruptly put on to androgens at high doses, the aggression may become more overt. The transition through an endocrine adolescence very late requires medical guidance. Generally, hypogonadal men adapt very well to their new endocrine status and many go on to develop satisfactory heterosexual relationships.

952

FURTHER READING ON DISORDERS OF SEX DIFFERENTIATION Azziz R, Dewailly D, Owerbach D 1994 Non-classic adrenal hyperplasia: current concepts. J Clin Endocrinol Metab 78: 810-815. Azziz R, Carmina E, Sawaya M E 2000 Idiophic hirsutism. Endocrine Revs 21: 347-362. Braunstein G D 1993 Gynecomastia. N Engl J Med 328: 490-495. Rey R, Picard J-Y 1998 Embryology and endocrinology of genital development. Bailliere's Clin Endocrinol Metab 12:17-33. Rittmaster R S 1997 Hirsutism. Lancet 349:191-195.

18 Metabolic Bone

Disease

and

Mineral

Metabolism John Cunningham and John P Monson

lar fluid, the free intracellular calcium ion concentration is approximately 10-7-10-6 molar, which is one thousandth of the free calcium ion concentration in the ECF (10-4 molar, or ImM). Most of the intracellular calcium is sequestered in, or on, specific cell organelles. Calcium in blood plasma at 2.4 mM is present in three main forms: free ionized, protein bound and anion bound (Table 18.1). The biological functions of ECF calcium are thought to be effected by the ionized fraction only. Inorganic phosphate in plasma is only slightly protein bound (approximately 12%), the rest being present as HPO2/H2PO4 (in the ratio 4:1 at physiological pH), as well as in the form of sodium, calcium and magnesium phosphates. Most of the intracellular phosphate is covalently bound to organic compounds or in the form of phospholipids. The free inorganic phosphate concentration in cells is probably less than 1 mM.

Bone composition Anatomy and physiology 953

Rickets and osteomalacia 962

Hormonal control of bone, calcium and phosphorus metabolism 954

Hypercalcaemia 965

Magnesium homeostasis 956 Investigation of bone disease 957

Hypocalcaemia 972 Paget's disease of bone 974 Miscellaneous rare bone diseases 978

Osteoporosis 959

ANATOMY AND PHYSIOLOGY The skeleton: • Supports the soft tissues • Encases and protects certain organs such as the central nervous system (CNS) • Acts as an enormous reservoir for minerals, in particular calcium, phosphorus, magnesium and sodium. Although in health these functions are not in conflict, there are circumstances - for example calcium deprivation when the skeleton's contribution to chemical homeostasis may lead to progressive demineralization and compromise its mechanical functions.

Distribution of mineral The adult skeleton contains some 99% of total body calcium (Table 18.1), ranging from 1 to 2kg (25-50mol) depending on size, 88% of total body phosphorus, 80% of body carbonate, 50% of body magnesium and 35% of body sodium. Although calcium is distributed approximately equally between extracellular fluid (ECF) and intracellu-

Bone consists of a calcified (mineralized) organic matrix. The organic matrix, osteoid, makes up about 30% of the total skeletal mass, and consists mainly of type I collagen and the vitamin K-dependent protein osteocalcin, with small amounts of other proteins. The mineral component, which is deposited throughout the organic matrix, makes up the remaining 70% of skeletal mass and consists mainly of a complex crystalline salt of calcium and phosphate called hydroxyapatite. There are also smaller amounts of amorphous acid phosphate, carbonate, sodium and magnesium. Bone undergoes continual modelling and remodelling by bone cells, which allows the skeleton to grow and adapt itself to prevailing mechanical needs. The result is an organ that, for its weight, is extremely rigid and strong without being brittle. The skeletal adaptation provides for appropriate strength only where it is needed, thereby avoiding excess weight. At least three distinct cell types contribute to these processes (Fig. 18.1). The osteoblast The osteoblast is derived from mesenchymal precursor cells and is the main bone-forming cell. It synthesizes organic matrix (osteoid) and probably also controls the subsequent mineralization of the matrix. It has plentiful mitochondria and an extensive Golgi apparatus associated with rapid protein synthesis. It is rich in alkaline phosphatase. Cells of the osteoblast line possess specific receptors for parathyroid hormone (PTH) and 1,25dihydroxyvitamin D (1,25(OH)2D, calcitriol). The osteocyte Osteocytes are derived from osteoblasts that no longer have a major biosynthetic role and have become encased in bone, communicating with each other by extensions running through canaliculae in the bone. They are likely to be involved in the movement of mineral in the vicinity of the osteocyte lacunae and in detecting stresses in bone.

953

FIG. 18.2 Relationship between bone mass and age

FIG. 18.1 Diagram of normal trabecular bone undergoing osteoclastic resorption followed by new bone formation by osteoblasts

citonin, thyroxine, glucocorticoids and the sex steroids. The functions of osteoclasts and osteoblasts are closely linked, osteoclastic bone resorption being followed by an increase in osteoblastic bone formation. This coupling is effected by locally produced regulators.

Bone qrowth, modelling and remodelling TAlLE 18.1 Distribution of body calcium Compartment

Quantity/concentration

Whole body Skeleton Rapidly exchangeable/bone surface Extracellular fluid (ECF) Intracellular fluid (ICF)

1kg(25mol) 8g(200mmol) 1g(25mmol) 1g(25mmol)

In plasma Free ionised Ca2+ Protein-bound Albumin Globulin Ultrafiltrable bound (by bicarbonate citrate and phosphate) Total in plasma

954

1.1mmol/L 0.75mmol/L 0.25mmol/L 0.3mmol/L 2.4mmol/L

The osteoclast Osteoclasts are multinucleated cells derived from cells related to macrophages. Their function is bone resorption. When actively resorbing they have a deeply folded plasma membrane, creating a 'ruffled border' in close proximity to the bone surface. Although they are influenced by the calcium-regulating hormones, PTH and 1,25(OH)2D, they do not apparently possess specific receptors for these hormones; it is therefore likely that hormonal control is exerted by way of communication with osteoblasts, or by effects of hormones on osteoclast precursors. Osteoclasts may also be influenced by other hormones, including cal-

Longitudinal bone growth in childhood depends on the proliferation of cartilage at the hypertrophic zone, which becomes mineralized and then formed into bone. This process is mediated by growth hormone, through the action of locally produced insulin-like growth factor 1 (IGF-1), and thyroid hormone. Diameter is increased by periosteal bone formation. Once growth has ceased the coupled activity of osteoblasts and osteoclasts continues, although at a much slower rate, a process which allows alteration of bone architecture (without growth) in response to factors such as mechanical loading. As shown in Figure 18.2, skeletal mass peaks at about the age of 25, decreasing thereafter at a constant rate in males, and with a transient acceleration after the menopause in females. This progressive decrease indicates that, throughout most of adult life, net bone resorption just exceeds net formation. In some individuals, especially postmenopausal women, bone mass may decrease to the point where clinically important mechanical weakening arises (osteoporosis), leading to fractures that may be spontaneous or follow minimal trauma.

HORMONAL CONTROL OF BONE, CALCIUM AND PHOSPHORUS METABOLISM The interrelationships between calcium and inorganic phosphorus in the ECF, the skeletal reservoirs of these ions, and their main sites of entry and exit in the body (the intestine and the kidney, respectively), are controlled by a hormonal system that is dominated by PTH and 1,25(OH)2D (Fig. 18.3). The role of the third potential calcium-regulating hormone, calcitonin, is more enigmatic.

• To decrease the renal tubular reabsorption of inorganic phosphate, thereby promoting phosphaturia; • To stimulate the production of 1,25(OH)2D, thereby promoting intestinal calcium absorption.

18

1,25-dihydroxyvitamin D (1,25(OH)2D, calcitriol)

FIG. 18.3 Vitamin D metabolism and the hormonal regulation of calcium metabolism Processes raising blood calcium are shown in blue and calcium-lowering ones in black.

Parathyroid hormone PTH is an 84 amino acid peptide hormone that is manufactured, stored and released by the chief cells of the parathyroid glands. Two intracellular precursors have been identified, prepro-PTH being cleaved successively to yield pro-PTH and, finally, PTH 1-84 itself. Its synthesis is regulated by the concentrations of extracellular calcium, extracellular phosphate and 1,25(OH)2D3, each forming a negative feedback loop with the parathyroid gland (Fig. 18.3). Calcium and phosphate regulate PTH production at the post-transcriptional level, whereas 1,25(OH)2D, a steroid hormone, acts via a vitamin D receptor and a vitamin D response element on the PTH gene. Both PTH synthesis and release are highly responsive to the prevailing ionized calcium concentration in the ECF, particularly in and near the range of physiological calcaemia, and are mediated by a calcium sensor located in the membrane of the parathyroid cells. Thus, modest hypocalcaemia triggers a substantial PTH response and modest hypercalcaemia almost completely suppresses the secretion of biologically active PTH. As shown in Figure 18.3, PTH forms part of a feedback loop that is well suited to maintaining calcium homeostasis. The actions of PTH itself are directed towards the elevation of blood and ECF calcium concentration. These actions are: • To promote recruitment and activation of osteoclasts and so accelerate bone resorption (indirectly through osteoblasts); • To increase the renal tubular resorption of calcium;

The potent steroid hormone 1,25-dihydroxyvitamin D is produced almost entirely in the kidney, although small amounts are found in other tissues, such as alveolar macrophages and the placenta. It is the product of a series of chemical transformations that starts with the production of cholecalciferol (vitamin D3) in the skin under the influence of ultraviolet light, or with the ingestion of ergocalciferol (vitamin D2) added to the diet. Normally, the latter is quantitatively less important. 25-Hydroxylation of vitamin D in the liver, and subsequent 1-hydroxylation of 25hydroxyvitamin D in the kidney, result in production of the active hormonal form of vitamin D, 1,25(OH)2D3 (Fig. 18.3). The 1-hydroxylation in the kidney is stimulated by PTH and hypocalcaemia. Other hydroxylated metabolites of vitamin D are also produced by the kidney, but 1,25(OH)2D is the only one with clearly defined biological actions. Like PTH, 1,25(OH)2D is a calcaemic hormone; its known actions complement those of PTH. 1,25(OH)2D has three major target organs, the intestine, bone and the parathyroid glands. In the intestine, 1,25(OH)2D increases the absorption of calcium and, to a lesser extent, phosphate by: • Increasing the permeability of the brush border to calcium; • Promoting the synthesis of a high-affinity calciumbinding protein in the cell; • Accelerating active extrusion of calcium across the basolateral membrane. In bone, the main action of 1,25(OH)2D is to increase resorption. This is probably by stimulating proliferation and differentiation of osteoclast precursors. In addition, osteoblasts (which possess receptors for 1,25(OH)2D, can probably modulate the function of nearby osteoclasts (Fig. 18.1). In the parathyroids, 1,25(OH)2D suppresses PTH gene transcription, thereby completing a negative feedback loop that limits excessive PTH response to hypocalcaemia. These actions of 1,25(OH)2D favour elevation of the concentration of ionized calcium and inorganic phosphorus in blood. The extent to which bone formation and resorption are under the direct control of vitamin D metabolites is unclear, although indirectly the vitamin D system certainly influences bone formation by controlling calcium and phosphate concentrations in the ECF. Renal production of 1,25(OH)2D is under tight control by PTH, hypocalcaemia and hypophosphataemia, all of which stimulate production of la-hydroxylase. The phosphaturic effect of PTH prevents the calcium phosphate product rising when PTH rises and bone resorption increases.

955

Other hormones Although many other hormones have demonstrable effects on bone metabolism under appropriate circumstances, their physiological role is uncertain. Calcitonin Calcitonin is a 32 amino acid peptide hormone secreted by the C cells of the thyroid gland. Its best-defined actions are on bone, where it binds to specific receptors in osteoclasts, leading to decreased resorptive activity. This effect is most striking where bone turnover is high, such as in growing bone or in Paget's disease (p. 974). Even when given in pharmacological doses, the influence of calcitonin on calcium and bone metabolism in the normal adult is minimal. Sex steroids Oestrogens The acceleration of bone loss after the menopause and its amelioration by oestrogen therapy, together with the high prevalance of osteoporosis in women with untreated primary or secondary hypogonadism, has long suggested that oestrogens exert an important influence on bone metabolism. Recent evidence suggests a direct effect, mediated by oestrogen receptors on osteoblasts. Androgens These are responsible for the attainment of higher peak bone mass in males (Fig. 18.2), either directly, or indirectly by increasing muscle bulk and strength. Androgens and oestrogens also act on growth cartilage, in concert with growth hormone, promoting growth and, later, epiphyseal closure. Glucocorticoids In physiological amounts, glucocorticoids are important regulators of bone growth, probably by actions on both bone-forming and bone-resorbing cells. Pharmacological doses of these steroids reduce bone mass, leading to osteoporosis. Thyroxine About one-third of patients with thyrotoxicosis have mild hypercalcaemia, hypercalciuria or both, and neglected thyrotoxicosis may lead to osteoporosis. Conversely, hypothyroidism is associated with low bone turnover, which may be restored to normal by physiological doses of thyroxine. Growth hormone Even after cessation of statural growth, growth hormone (GH), acting through IGF-1, remains an important regulator of bone remodelling. Acquired GH deficiency in adult life is associated with a reduced bone density.

1

956

MCQ18.1

Local modulation of bone, calcium and phosphorus metabolism A variety of local factors contribute to the delicate balance that exists between the actions of the different types of bone cell, although many of these have not yet been fully characterized. Prostaglandins Local production of prostaglandins, such as PGE2, may influence bone metabolism directly, or by promoting the synthesis of other locally acting factors, such as epidermal growth factor and platelet-derived growth factor. Inhibitors of prostaglandin synthesis have little if any effect on bone. Cytokines There are many cytokines with proven or possible paracrine function in bone. Interleukin-1 (IL-1) is a potent bone-resorbing agent derived from activated macrophages. It has many interactions with other cytokines, such as yinterferon, consistent with local regulatory functions in bone. Tumour necrosis factors and transforming growth factor p also have effects, and some or all may be involved in bone cell coupling. IL-4 and IL-11 exert inhibitory effects on osteoblasts. Osteoclast development and/or function is influenced by IL-3, IL-6 and IL-8, as well as by macrophage colony-stimulating factor (MC-CSF) and by granulocyte macrophage CSF. Cytokine-induced bone resorption is an important mechanism in bone metastasis and hypercalcaemia of malignancy (see Fig. 18.8, p. 970). O

MAGNESIUM HOMEOSTASIS Magnesium ions have a similar cellular/extracellular distribution to potassium, and parallel alterations in concentration of these cations are observed in a number of situations. The high content of magnesium in bone results in important associations with calcium metabolism; for example, magnesium is an effector of PTH secretion. Hypomagnesaemia may occur with: • Severe malnutrition - lack of magnesium intake; • Long-term intravenous feeding in the absence of magnesium supplements; • Some long-term renal tubular disorders; • Chronic diuretic therapy causing renal loss; • Uncontrolled diabetes mellitus - renal loss; • Alcoholism - malnutrition; • Following surgical correction of hyperparathyroidism owing to calcium and magnesium deposition in bone; • Cytotoxic treatment with cisplatin (as a result of failure of magnesium reabsorption owing to renal tubular damage). Hypomagnesaemia is rarely severe, but may occasionally give rise to symptoms and signs similar to those of hypocalcaemia, with tetany progressing to convulsions and cardiac

arrhythmias, with or without associated hypocalcaemia. This may necessitate magnesium supplementation by oral magnesium salts (of which magnesium glycerophosphate is least likely to cause diarrhoea) or parenteral magnesium sulphate. Coexisting hypocalcaemia is refractory until the hypomagnesaemia has been corrected; this is owing to impaired secretion of PTH, and also to target organ resistance to circulating PTH during hypomagnesaemia. Hypermagnesaemia may complicate renal impairment. It results in gastrointestinal disturbance and may, if severe, cause neuromuscular dysfunction and CNS depression. Cardiac toxicity, including complete heart block, may occur with plasma magnesium concentrations in excess of 7.5mmol/L (normal is 0.7-0.8mmol/L). However, hypermagnesaemia is rarely of clinical significance.

INVESTIGATION OF BONE DISEASE Investigations of bone disease fall into five groups: • • • • •

Biochemical: blood and urine Radiographic Radionuclide studies Measurement of bone mass Histology.

TABLE 18.2 Biochemical investigation of bone disease Measurements Blood plasma Urea creatinine Na+, K+, CI-, HC03Calcium

Phosphate

Alkaline phosphatase (total or bone-specific) and osteocalcin Parathyroid hormone (PTH) 25-hydroxyvitamin D Urine Calcium Phosphate

Biochemistry (Table 18.2) Blood calcium Total calcium in plasma or serum is the usual measurement taken in clinical practice, although in some instances direct measurement of the physiologically important ionized fraction is undertaken. The latter is likely to be used increasingly as it avoids the difficulties that may arise when abnormalities of plasma protein concentration or acid-base status alter the normal relationship between total and ionic calcium in blood. A reduced plasma protein concentration leads to a reduction in plasma total calcium concentration, even though free ionic calcium is normal. The error may be reduced by 'adjustment' of the observed calcium concentration, using the formula: l

Ca (adjusted) = Ca + 0.02(46 - alb) 1

where Ca is the observed concentration of total calcium in mmol/L and alb is the albumin concentration in g/L. Some laboratories have a reference range for serum calcium based on correction to a serum albumin of 40 g/L. As pH falls, so does calcium binding to plasma proteins. Thus, during severe acidosis, plasma total calcium may be low but with a normal free ionic calcium concentration. Increases in blood calcium usually reflect accelerated entry into the ECF. This occurs as a result of either rapid bone resorption (primary hyperparathyroidism, vitamin D intoxication, osteolytic metastases or PTH-rp in humoral hypercalcaemia of malignancy), or increased intestinal

Hydroxyproline, pyridinoline and deoxypyridinoline

18

Associated disease Renal disease Acid-base disturbance Increased in primary hyperparathyroidism, malignant disease, granulomatous disease, vitamin D intoxication Decreased in renal disease, osteomalacia, hypoparathyroidism Increased in renal disease, hypoparathyroidism and pseudohypoparathyroidism Decreased in primary hyperparathyroidism, vitamin D deficiency, renal tubular disorders Increased in childhood, hyperparathyroidism, vitamin D deficiency, fracture, malignancy, Paget's disease Decreased in hypophosphatasia Increased in hyperparathyroidism and pseudohypoparathyroidism Vitamin D deficiency or intoxication Hypercalciuria Familial hypocalciuric hypercalcaemia Decreased tubular reabsorption in hyperparathyroidism Reflect increased bone resorption, e.g. in hyperparathyroidism, thyrotoxicosis and immobilization

absorption (high dietary calcium intake, primary hyperparathyroidism, vitamin D intoxication or milk alkali syndrome), or both processes acting together. Decreases in blood calcium are due to diminished bone resorption and/or diminished intestinal absorption of calcium. Important causes are hypoparathyroidism, vitamin D deficiency with calcium malabsorption, and severe renal disease. Urine calcium Urine calcium ranges from 2.5 to 7.5mmol/24h in adults and is increased in situations where calcium entry to the ECF is enhanced (leading to potential or actual hypercalcaemia), or where the renal tubular reabsorption of calcium is decreased ('renal leak' type of hypercalciuria), or both. Plasma inorganic phosphate Plasma phosphate is normally determined by the level of renal tubular reabsorption of phosphate, a process that is itself modulated by the dietary intake of phosphate and by PTH. Tubular reabsorption of phosphate is decreased by PTH and increased by low dietary intake of phosphate. A low plasma phosphate is found in association with primary

957

FIG. 18.4 X-ray showing subperiosteal erosions in the hands in a case of hyperparathyroidism

hyperparathyroidism, vitamin D deficiency, certain renal tubular disorders (p. 964) and prolonged dietary phosphate deprivation. Serum phosphate is elevated in patients with substantial impairment of renal function with reduction of glomerular filtration rate (GFR), in hypoparathyroidism, and occasionally in patients with greatly accelerated tissue catabolism (e.g. rhabdomyolysis, tumour necrosis).

vitamin D status. 25(OH)D in plasma shows a seasonal variation, being high in summer and lower in winter because cutaneous production of vitamin D depends on ultraviolet light. Plasma levels are low in vitamin D deficiency, because of lack of sunlight exposure and poor diet normal in patients with renal disease, elevated in patients with vitamin D intoxication.

Alkaline phosphatase The bone-specific isoenzyme of alkaline phosphatase is produced by osteoblasts, and its release into the circulation increases as osteoblast numbers and/or activity increase. This is seen in normal children during the phase of active bone growth, and also in generalized disorders characterized by high bone turnover (e.g. hyperparathyroidism) and in focal lesions (e.g. healing fractures, Paget's disease or bone metastases). Many other tissues also contain alkaline phosphatases, which differ slightly from one another (isoenzymes). To avoid diagnostic confusion, elevation of blood alkaline phosphatase must be interpreted in the clinical context and not in isolation. Specific isoenzymes can be measured to identify the tissue of origin, but even the best assays suffer from cross-reactivity between two or more of the isoenzymes. Serum concentration of osteocalcin provides an additional marker of osteoblastic activity but has little clinical application at present.

Urinary deoxypyridinoline Deoxypyridinoline and pyridinoline are cross-linking compounds in mature collagen. Bone collagen contains both compounds, and their release from bone undergoing resorption constitutes the main source of these compounds excreted in the urine. Urinary collagen cross-links are more sensitive and more specific markers of bone resorption than hydroxyproline and have the advantage of not being susceptible to dietary influences. Deoxypyridinoline has greater specificity for breakdown of collagen from bone than other tissues.

Parathyroid hormone Measurement of PTH by radioimmunoassay is an important adjunct to the diagnosis of some metabolic bone disorders. PTH in plasma is present as intact PTH (84 amino acids) and also as several different peptide fragments. Current assays use antibodies to two sites on the molecule and detect only intact PTH. Elevated PTH is found in primary or secondary hyperparathyroidism, and in all patients with target organ resistance to the action of PTH (pseudohypoparathyroidism, p. 974). PTH is undetectable in patients with hypoparathyroidism. 25-hydroxyvitamin D Although 25-hydroxyvitamin D (25(OH)D) is only the precursor of the active hormonal form of vitamin D (Fig. 18.3), measurements of 25(OH)D are the best guide to

Radiography Plain radiographs are of great value in the diagnosis of localized.disorders of bone, e.g. developmental abnormalities, fractures, tumours and infections. Some diseases affecting the entire skeleton may also be diagnosed in this way. For example, severe osteoporosis may lead to radiologically evident fractures, and hyperparathyroidism may be associated with visible subperiosteal erosions in the hands and elsewhere (Fig. 18.4). In these instances, local abnormalities are acting as markers for generalized disease.

Radionuclide studies Bone-seeking radionuclides such as 99mTc methylene bisphosphonate may be given by intravenous injection and the pattern of uptake assessed by external counting using a gamma camera. Uptake is largely determined by blood flow to bone and by the rate of bone turnover; the technique is therefore extremely good at detecting focal lesions, such as metastases, infections, Paget's disease or subtle fractures, that may not be visible on conventional radiographs.

Measurement of bone mass 1

958

Figs 18.1-18.2

2

Figs 18.3-18.6

The increasing proportion of elderly people in populations worldwide had led to an increase in the incidence of agerelated diseases such as osteoporosis. Accurate measure-

ment of bone mass, simply and safely, is essential for the identification, assessment and monitoring of patients with osteoporosis. Conventional radiography is imprecise and has little to offer apart from the identification of fractures. Dual X-ray absorptiometry (DXA or DEXA) gives an accurate measurement of bone and the mineral content of the spine and femoral neck. 1 In contrast to quantitative computerized tomography (CT), it involves very low radiation exposure. DXA measurements are good predictors of fracture risk. Peripheral sites such as the calcaneum are also used as surrogates for the wider skeleton - X-ray absorptiometry and heel ultrasound (speed of sound transmission) both show promise as screening tests.

Histology The technique of transiliac bone biopsy, whereby a cortexto-cortex core of bone 5-8 mm in diameter is taken from the ilium under local anaesthetic, is sometimes used in the diagnosis and assessment of generalized metabolic bone disease. The specimens may yield useful information on the number and activity of osteoblasts and osteoclasts, and indicate whether organic matrix production is normal and being appropriately mineralized. 0 A useful adjunct is the technique of tetracycline labelling. Here, tetracycline (600-900 mg of demeclocycline) is given twice, with a 10day interval between; the biopsy is performed 4 or more days after the second label is taken. The drug is taken up at the mineralization front in bone, and can be detected in the biopsy specimen by fluorescence under ultraviolet light. Uptake is poor or absent in osteomalacia (pp. 962-965) and enhanced in states of high bone turnover (e.g. hyperparathyroidism and thyrotoxicosis).

OSTEOPOROSIS The term osteoporosis denotes loss of bone mass per unit volume; bone composition is normal, although microscopic trabecular bone architecture is not. The WHO definition of osteoporosis is based on a T-score (standard deviation score compared to the young adult reference range) of less than -2.5 and osteopenia as less than -1.0 at the lumbar spine and/or femoral neck. It is estimated that T-scores of -1.0 to -2.5 are associated with a doubling of fracture risk. Osteoporosis becomes clinically important when bone is so weak that pathological fractures result. Most examples of osteoporosis arise as a result of the progressive decrease in bone mass with age (see Fig. 18.2); as this occurs to a greater or lesser extent in all individuals, the distinction between pathology and the normal ageing process is blurred.

Pathological changes in bone mass Bone mass is governed by the balance between bone for-

mation and resorption. Positive, neutral and negative bone balances are found at different ages. In infancy and adolescence, bone formation clearly exceeds resorption; in early adult life the processes are approximately equal, and later a disparity in favour of resorption results in a slow decline in total bone mass (Fig. 18.2). Whether or not this gives rise to clinical problems (fractures) depends to a large extent on the maximum bone mass achieved as an adult. This is greater in men than women, and is also greater in Afro-Carribeans than in whites. Women are further disadvantaged by a more rapid decrease in bone mass in later adult life, and the net result is a much higher incidence of osteoporotic complications. On average, women also live for about 5 years longer than men. The reasons for the sex differences in rate of bone loss are probably multiple, but postmenopausal oestrogen deficiency is most strongly implicated. This condition may sensitize the skeleton to the bone-resorbing effects of PTH, and possibly also impairs renal synthesis of 1,25(OH)2D. These metabolic changes may lead to mild, but chronic, calcium malabsorption and consequent secondary hyperparathyroidism (pp. 962 and 972). In extreme old age, bone loss in men catches up with that in women. Idiopathic osteoporosis may occasionally occur in young people, especially during the adolescent growth spurt. It is usually self-limiting, but may be complicated by fractures. A particularly severe form of juvenile-onset osteoporosis is associated with decreased sex steroid receptor expression in bone, which is probably genetic in origin.

18

Clinical features and investigation Osteoporosis remains asymptomatic unless structural collapse or fracture of bone occurs. Damage is particularly likely to occur in the dorsal vertebrae, femoral neck and distal radius. The precipitating trauma may be trivial, but fracture frequently follows a fall. There are usually no abnormalities on routine plasma biochemistry, except in cases secondary to specific endocrine disorders (see below). In some patients urinary calcium excretion may be increased, consistent with net bone resorption, and plasma calcium may be marginally elevated. This is particularly common when immobility is a contributory factor. Plain radiographs demonstrate decreased bone density due to loss of trabeculae, and reduced cortical thickness of long bones in severe cases. The dorsal vertebrae may show the classic 'wedge collapse' of osteoporotic crush fracture and herniation of the disc into the centre of the vertebra to give a 'codfish' vertebra, and exaggeration of the normal curvature of the thoracic spine (Fig. 18.5). Detection and assessment of osteoporosis is best achieved by measurements of bone mass by dual X-ray absorptiometry or CT scanning of the spine. Normality of plasma calcium, phosphate and alkaline phosphatase helps to exclude osteomalacia, but in doubtful cases further investigation may be necessary, including

959

TABLE 18.3 Causes of generalized osteoporosis and their mechanisms* Cause

Bone resorption

Common causes Corticosteroid excess Early menopause Immobility Multiple myeloma1 Rare causes Hypogonadism (primary or secondary) Thyrotoxicosisf Type 1 diabetes mellitus Pregnancy Primary hyperparathyroidism* Adult GH deficiency

Bone formation

Normal

Normal

Normal

4

Normal

* Primary liver disease is omitted as this is associated with a combination of osteoporosis and osteomalacia. Conditions in which the plasma calcium may be increased. FIG. 18.5 X-ray of lateral thoracic spine of a 60-year-old woman with severe postmenopausal osteoporosis Note gross loss of density, collapsed and 'codfish' vertebrae and kyphosis.

Management measurement of plasma 25-hydroxyvitamin D and transiliac bone biopsy. In most instances no specific underlying cause for osteoporosis will be found. However, careful consideration should be given to the causes outlined below, and further investigation initiated if the overall clinical picture so warrants.

Specific causes of osteoporosis

960

Osteoporosis is common in the elderly and it may be difficult to decide whether a vertebral collapse is due to osteoporosis or to another, coincidental, underlying disease, such as secondary carcinoma. Cancers of the breast, lung, kidney and thyroid typically cause lytic lesions in bone. Prostatic cancer may also do so, though sclerosis is more common. Breast cancer in particular has a predilection for secondary spread to the axial skeleton. Myeloma (Ch. 24, p. 1435) may be accompanied by generalized thinning of bones and/or focal lytic areas. Breast cancer will often be detected by physical examination, but further investigation may be needed to exclude other forms of cancer and myeloma. Chronic liver disease (especially primary biliary cirrhosis) may give rise to a mixed picture of osteoporosis and osteomalacia (see below). Hereditary abnormalities in bone formation are discussed on page 978. Conditions that may be complicated by significant generalized osteoporosis are listed in Table 18.3.

Prevention Prevention of the development of osteoporosis in individuals at risk is of the utmost importance. This is most obviously the case in females with primary or premature ovarian failure, who should be treated with either oral oestrogen, e.g. oestradiol valerate l-2mg/day, or transdermal oestradiol 50-100 ug patches twice weekly. Progesterone supplements should also be given cyclically, or at low dose continuously (p. 948), to avoid unopposed endometrial stimulation with its risk of neoplastic change. However, the use of prophylactic oestrogen supplements after the physiological menopause is more controversial. Oestrogen alone undoubtedly reduces the rate of bone loss after the menopause (and even increases bone mass in many patients). This is probably not associated with an increased risk of atheroma development, at least at low dosage but is associated with an increased relative risk of breast cancer. Appropriate exercise, e.g. regular swimming, should be encouraged in patients with relative immobility, and regular non-strenuous exercise in the elderly. Treatment Osteoporosis is a heterogeneous condition that arises most commonly in postmenopausal women, but also as a consequence of specific conditions, including thyroid disease, parathyroid disease, renal disease, chronic liver disease, hypogonadism and corticosteroid excess (both endogenous and iatrogenic). The role of growth hormone (GH) deficiency in the development of osteoporosis in the hypopituitary patient, and possibly also in the elderly,

18

CASE STUDY 18.1 ACUTE BACK PAIN A 68-year-old woman presented with a history of sudden onset of localized dorsal spinal pain. She experienced severe pain on attempted movement and examination demonstrated localized tenderness over the sixth dorsal vertebra. There were no breast masses and no neurological signs. Previous medical history was unremarkable, apart from a relatively early menopause at the age of 44. She had not received postmenopausal oestrogen replacement therapy, was a non-smoker, and consumed occasional alcohol only. Initial investigations indicated haemoglobin 12.9 g/dL, ESR 13mm/h, normal serum electrolytes, urea 5.2mmol/L, serum creatinine 86 umol/L, serum calcium 2.22 mmol/L, serum phosphate 0.98 mmol/L and alkaline phosphatase 78IU/L. Questions 1. What is the most likely cause for her pain? 2. How might this be confirmed? 3. What treatment options are available for long-term management?

Discussion The presentation with acute dorsal spinal pain in a patient of this age is strongly suggestive of a vertebral crush fracture. Pre-exisiting osteoporosis is the most likely precipitant, and the early onset of post menopausal oestrogen deficiency would be considered a significant predisposing risk factor for this. However, vertebral metastasis is a common manifestation of malignancy, with cancer of the breast, kidney and lung and myeloma being common underlying causes. The high risk of spinal cord compression makes this an important differential diagnosis. The normal breast examination is an important negative finding in this respect, as is the normal blood count. Osteomalacia, or rarer causes of metabolic bone disease, are improbable given the normal serum concentrations of calcium, phosphate and alkaline phosphatase. Plain radiological imaging confirmed a wedge-shaped fracture of the body of the sixth dorsal vertebra against a background appearance of reduced bone mineral density. Formal assessment of bone mineral density by dual-energy X-ray absorptiometry demonstrated T-scores (standard

in whom GH secretion declines with age, is increasingly recognized. The treatment of established osteoporosis includes: • Avoiding immobility and encouraging exercise; • Adequate dietary calcium; • Treating sex steroid deficiency (particularly when this has occurred pathologically or physiologically at an early age); • Treatment of underlying disorders; • Bisphosphonate therapy. Patients should be encouraged to take regular, but not over-strenuous exercise: walking and activities such as golf are convenient for many. Dietary calcium should be generous (at least 20mmol/day); if this is not achieved, the diet may be supplemented with calcium carbonate tablets. The evidence that calcium supplements are of benefit is con-

deviation scores compared to the young adult reference range) of -2.6 and -2.1 at the lumbar spine and femoral neck, respectively, indicating osteoporosis. An isotope bone scan showed no evidence of metastases at other sites. Long-term management of this patient's condition should include dietary assessment, with calcium and vitamin D supplements (400IU/day) if her diet is deemed to be inadequate. Specific treatment in the form of bisphosphonate therapy (e.g. cyclical etidronate with calcium supplementation or alendronate) should be commenced. It is uncertain whether the addition of oestrogen replacement, in addition to bisphosphonate therapy, would have any substantial effect in a patient of this age. However, there is emerging evidence of an additive effect of these two modalities of treatment, so that in the absence of any specific contraindication, oestrogen and progesterone replacement in the form of a continuous 'no bleed' preparation (either oral or transdermal) might be given. Followup bone densitometry measurements at biannual intervals will be indicated.

flicting, but the therapy is easy and does no harm. There is now strong evidence that in the very old, calcium (1.5g) and vitamin D (400 units daily) supplementation reduces hip fracture incidence. Preparations of 'hydroxyapatite compound' are of no proven benefit above that of simple calcium salts, and anabolic steroids are of doubtful value. Other measures currently under investigation include the use of calcitonin, PTH and sodium fluoride. Oestrogen replacement therapy should be considered as for prevention. Accumulating experience of the use of selective oestrogen receptor-modulating agents (SERMS, e.g. raloxifene) in the management of osteoporosis is promising. Oral bisphosphonate therapy given as an interval therapy in conjunction with calcium supplements (Didronel PMO) has been shown to reduce the rate of bone loss in postmenopausal osteoporosis. Recent large trials of the bisphosphonate alendronate have confirmed substantial increments in bone mineral density and a reduction in fracture rates. Calcitonin, by s.c. injection or intranasal

961

SUMMARY 1 Risk factors and prevention of osteoporosis Risk factors • • • • • • •

>65 years Premature menopause (<45 years) Personal/family history of low-trauma fractures History of amenorrhoea >6 months, excluding pregnancy Body mass index <19kg/m2 Immobility Predisposing conditions, e.g. thyrotoxicosis, glucocorticoid excess and primary hyperparathyroidism

Prevention • • • • • •

Lifestyle measures Avoid excessive alcohol Stop smoking Regular weight-bearing exercise Avoid excessive dieting and exercise resulting in amenorrhoea Maintain adequate calcium and vitamin D intake: Calcium: 1500mg daily in postmenopausal women, teenagers and pregnant women; 1000mg in other patient groups Vitamin D: 400 units daily • Minimize risk of falls by consideration of problems with gait, eyesight, medication etc.

administration, has a significant effect but is relatively inconvenient and expensive. Oestrogen, calcitonin and bisphosphonate therapy all exert their major benefit by inhibiting bone resorption; the only agent that has a predominant effect on bone formation in simple osteoporosis is fluoride. Endemic fluorosis, in which excessive fluoride is ingested in certain parts of South Africa and China, is associated with bone of increased mineral density and, paradoxically, increased fragility. However, early concerns that fluoride therapy might produce bone of poor quality have not been borne out by recent studies using lower doses. GH replacement in hypopituitary adults has been shown to produce a sustained increment in bone mineral density after 18 months of therapy. The long-term effects of GH therapy on bone density in the elderly are under investigation. The prevention and management of corticosteroidinduced osteoporosis is shown in Figure 18.6. In all studies of the effectiveness of treatment for osteoporosis it is important to monitor the effect of treatment on fracture risk rather than simply on bone mineral density, because these two endpoints may not always be concordant. Furthermore, we should be cautious about extrapolating data obtained from studies in women to the population as a whole. None the less, the emergence of effective therapies which modify the bone remodelling

1 962

MCQ18.2

cycle has gradually extended the treatment of established osteoporosis to the prevention of bone loss, particularly in situations in which it may be predicted (e.g. long-term corticosteroid therapy). Therapeutic nihilism in the management of reduced or threatened bone density is clearly no longer justified. 1

RICKETS AND OSTEOMALACIA Rickets and osteomalacia are conditions characterized by pathological defects in bone matrix mineralization. Rickets refers specifically to osteomalacia, where the defect occurs in growing bone. The aetiological factors are diverse, but the end result is an increased quantity of unmineralized bone matrix (osteoid). Histologically this is characterized by widened osteoid seams, morphologically flat 'inactive' osteoblasts, and an absence of tetracycline uptake, indicating impairment of mineralization. The increase in osteoid must be distinguished from that seen in thyrotoxic bone disease, Paget's disease or hyperparathyroidism. In all these conditions there is high bone turnover with increased formation of osteoid, but mineralization of matrix is normal or increased.

Aetiology The aetiology of osteomalacia (or rickets) is shown in Table 18.4. The conditions may arise in three distinct situations: • Deficiency or abnormal metabolism of vitamin D • Phosphate depletion • Chronic metabolic acidosis. Vitamin D deficiency is numerically by far the most important cause of osteomalacia (or rickets): it leads to a combined deficiency of calcium and phosphorus. There is intestinal malabsorption of calcium and, to a lesser extent, phosphorus. Secondary hyperparathyroidism results from the fall in plasma calcium, and the ensuing phosphaturia accelerates the phosphate deficiency. Lack of vitamin D is often due to a combination of poor exposure to sunlight and dietary deficiency. The elderly are particularly at risk. In the UK there is a high incidence of osteomalacia in the Asian immigrant population (p. 122). The reason for this is multifactorial and incompletely understood, but dietary deficiency of vitamin D and calcium, together with binding of calcium by the phytate contained in chapati flour, may contribute. In addition, deeply pigmented skin generates less vitamin D3 in response to a given amount of ultraviolet exposure, and this may be disadvantageous in temperate climates. Vitamin D deficiency may complicate fat malabsorption from any cause. Finally, calcium deficiency enhances the rate of degradation of 25(OH)D in the liver, further exacerbating the shortage of vitamin D and its metabolites.

TABLE 18.4 Causes of osteomalacia/rickets Cause

Mechanism

Vitamin D deficiency

Defective intake/formation Lack of solar exposure (D3) and lack of dietary intake or fat malabsorption (D2) Severe chronic liver disease Increased/wasteful metabolism Liver-enzyme-inducing drugs, e.g. phenytoin, phenobarbitone, carbamazepine, rifampicin Failure of activation, 25(OH)D->1,25(OH)2D Chronic renal failure Hereditary vitamin D-dependent rickets type I Target organ defect Hereditary vitamin D-dependent rickets type II

Phosphate deficiency

Poor intake Aluminium hydroxide - excess binding of dietary phosphate Increased loss X-linked hypophosphataemia (vitamin D-resistant rickets) Fanconi's syndrome Renal tubular acidosis Ureterocolic anastamosis Tumour-associated

Chronic metabolic acidosis

Renal tubular acidosis (may be phosphate depletion also) Ureterocolic anastamosis

Osteoblast/ mineralization defect

Hypophosphatasia Etidronate therapy Aluminium intoxication (in renal disease, seep. 1055)

Failure of 25-hydroxylation of vitamin D (p. 955, Fig. 18.3) This may be a feature of severe hepatocellular dysfunction. It may also result from hepatic enzyme induction consequent upon long-term anticonvulsant medication (leading to enhanced degradation). Reduced 1-hydroxylation of25-hydroxyvitamin D by the kidney occurs in chronic renal failure, and also as a rare recessively inherited condition in which levels of the 1hydroxylase enzyme are low (vitamin D-dependent rickets type I). Failure to hydroxylate 25-hydroxyvitamin D also occurs in hypoparathyroidism and pseudohypoparathyroidism. Another rare form of rickets is associated with an

1MCQ18.3

964

abnormality of the 1,25(OH)2D receptor in target tissues (vitamin D-dependent rickets type II). Phosphate depletion is a less common cause of rickets than deficiency or altered metabolism of vitamin D. It arises as a result of a reduced maximal capacity for renal tubular phosphate reabsorption. This may either be inherited as an X-linked recessive characteristic (vitamin Dresistant rickets, X-linked hypophosphataemia) or occur as part of Fanconi's syndrome - a more generalized hereditary or acquired renal tubular defect typified by aminoaciduria and glycosuria (p. 1057). Rarely, excessive ingestion of antacids may restrict intestinal phosphate absorption to the point where clinically important phosphate depletion, and even osteomalacia or rickets, may appear. Chronic metabolic acidosis, usually resulting from renal tubular disorders, may lead to rickets or osteomalacia. The bone lesions usually heal satisfactorily following treatment to correct the acidosis. Oncogenic osteomalacia Severe osteomalacia may rarely be a non-metastatic manifestation of tumours. The tumours are almost always sarcomas or connective tissue tumours of borderline malignancy (neurofibromatosis, haemangioendothelioma), and all produce a humoral factor that mediates phosphate. The osteomalacia is cured by removal of the tumour. The humoral factor responsible has been termed phosphatonin.

Clinical features In children, failure of bone mineralization gives rise to classic bone deformities, which include widening of the metaphyses, prominence of costochondral junctions (socalled 'rickety rosary'), and varus or valgus abnormalities of the knee joints. The bones are painful and statural growth is reduced. In adults, bone pain and tenderness are the most prominent features. A characteristic proximal myopathy may develop, and is more common in cases due to vitamin D deficiency.

Investigation Biochemical In osteomalacia due to deficiency or abnormal metabolism of vitamin D, investigations typically demonstrate a low normal plasma calcium (leading to secondary hyperparathyroidism), a low plasma phosphate due to phosphaturia as a result of increased PTH secretion, and a raised alkaline phosphatase indicating increased numbers of osteoblasts (Table 18.5). Vitamin D deficiency may be confirmed by measurement of 25-hydroxyvitamin D in plasma. X-linked hypophosphataemia (vitamin D-resistant rickets), on the other hand, is characterized by normocalcaemia, hypophosphataemia, and little or no alteration in plasma alkaline phosphatase level. Radiological Features of rickets include widening and irregularity of the metaphyses with an increased width of the growth plate. In

18

TABLE 18.5 Plasma measurements in various causes of osteomalacia/rickets Plasma measurements Cause

Calcium

Vitamin D deficiency Familial hypophosphataemia Chronic renal failure Renal tubular acidosis

Nor

Phosphate

N

N or N or

Nor

Alk. phos.

Nor Nor Nor

Urea

Creatinine

PH

N N

N

N

N

N

N

N

14

N = normal.

FIG. 18.7 Radiological features of osteomalacia Arrow indicates Looser's zone.

osteomalacia the bones appear less dense than normal, and the long bones may show localized areas of decalcification on the concave surface, referred to as pseudofractures or Looser's zones (Fig. 18.7). Decreased bone strength may result in biconcave deformity of the vertebrae. In hypophosphataemic rickets there may be a paradoxical apparent increase in bone density.

(500-1000 units/day). This can conveniently be given as combined calcium with vitamin D tablets. Alternatively, a single i.m. does of 150 000IU of D2 in oil is effective prophylaxis and treatment for at least 6 months. The possibility of intestinal malabsorption should be borne in mind, and appropriate investigation initiated if necessary. Osteomalacia associated with anticonvulsant therapy requires higher doses of vitamin D (3000-6000 units/day) unless a change in anticonvulsant is clinically feasible. Treatment of X-linked hypophosphataemia is generally unsatisfactory. The logical treatment is with oral phosphate supplements, but the doses needed are large and they may be poorly tolerated because of gastrointestinal side-effects. Oral phosphate works better when combined with pharmacological doses of vitamin D, or (and probably best) in combination with 1,25(OH)2D3. When 1,25(OH)2D3 or laOHD3 is given, great care must be taken to adjust the dose against the serum calcium, as hypercalcaemia may easily be induced. The management of renal tubular acidosis and renal osteodystrophy is dealt with in Chapter 20, pages 1065-1068. 1

HYPERCALCAEMIA Elevation of total plasma calcium is usually associated with a concomitant increase in ionized calcium concentration.

Causes

Histological Occasionally a bone biopsy may be needed (following double tetracycline labelling) to confirm a diagnosis of osteomalacia when biochemical investigations are equivocal. This is particularly the case when seeking possible treatable factors in elderly patients with decreased bone density.

Table 18.6 gives the proximate cause of the elevated blood calcium concentration (increased ingress from bone or intestine, or decreased egress via the kidneys), together with the key hormonal changes associated with the major causes of hypercalcaemia. Note that the appropriate response of PTH and of 1,25(OH)2D during hypercalcaemia is to be low - elevation of either or both of these hormones may therefore be presumed to be causally related to the hypercalcaemia.

Management

Clinical features

Dietary vitamin D deficiency is treated by oral replacement of vitamins D2 or D3 in physiological quantities

Mild hypercalcaemia is generally asymptomatic. Severe hypercalcaemia, however, leads to a life-threatening

965

CASE STUDY 18.2 PROXIMAL MUSCLE WEAKNESS A 64-year-old Bangladeshi woman presented to her GP with a 4-month history of generalized aching discomfort. Through an interpreter she described muscular discomfort, particularly in her shoulders, present at rest. The symptoms were not worse in the morning and there were no specific relieving factors. Her weight was stable. She had a previous history of gestational diabetes in two consecutive pregnancies, and 2 years previously had been diagnosed as having diabetes mellitus. This had been controlled with dietary measures alone. On examination she exhibited mild proximal weakness in both arms and legs, but no muscle tenderness and the reflexes were preserved. There was no temporal artery or scalp tenderness, and the remainder of the examination was unremarkable. Subsequent investigation demonstrated haemoglobin 10.9 g/dL, ESR 16mm/h, plasma glucose (postprandial) 7.6mmol/L, glycated haemoglobin 5.6%, plasma urea and electrolytes within normal limits, plasma calcium 2.04mmol/L, phosphate 0.78 mmol/L and plasma alkaline phosphatase 198IU/L (normal <110IU/L).

Questions 1: What is the most likely explanation for her symptoms and what other causes should be considered? 2. How might this problem arise? 3. What is the mechanism underlying the abnormalities in serum calcium, phosphate and alkaline phosphatase?

Discussion This patient's presentation with myalgia associated with a combination of mild hypocalcaemia, hypophosphataemia and elevation of alkaline phosphatase is strongly suggestive of vitamin D deficiency. This would be confirmed by measurement of serum 25-hydroxyvitamin D. This condition is very much more common in Asian patients than in whites. Important precipitants include dietary deficiency, limited exposure to sunlight and a calcium chelating effect of phytates present in chapati flour. A finding of proximal muscle weakness with preservation of reflexes raises the possibility of either glucocorticoid or thyroid hormone excess, which will usually (but not invariably) have other clinical features evident. Malignant ectopic ACTH syndrome may present, with muscle wasting and weakness as the sole clinical features of glucocorticoid excess (see Ch. 17, p. 933). Hypokalaemia is very frequent in this situation and pigmentation may occur. These were not evident in this case. Thyrotoxicosis may present with an indolent clinical picture dominated by proximal myopathy. However, brisk reflexes are usually found and weight loss is the rule. Furthermore, the proximal myopathies of glucocorticoid excess and thyrotoxicosis are not painful. Diabetic femoral radiculopathy (amyotrophy) is a relatively uncommon but characteristic complication of poorly controlled diabetes which presents with proximal muscular pain and profound weakness; reflexes are reduced or absent, and therefore this patient's

acute illness, treatment of which constitutes a medical emergency. The principal symptoms and signs are renal, gastrointestinal and neurological.

966

Renal Thirst and polyuria are early symptoms of nephrogenic

signs are atypical for this diagnosis. Polymyalgia rheumatica should always be considered in any older patient presenting with proximal muscle pain and discomfort. However, the absence of a diurnal variation in symptoms and the normal ESR effectively exclude this diagnosis in the present case. Although dietary deficiency of vitamin D is the most likely explanation for this condition, conditions causing malabsorption of fat-soluble vitamins (A, D, E, K) should be considered and serum tested for antiendomysial antibodies as an initial screen for gluten enteropathy. In dietary vitamin D deficiency, supplementation in the form of calcium and ergocalciferol tablets (two daily, supplying 800IU calciferol) provides satisfactory therapy. Deficiency of vitamin D results in reduced intestinal calcium absorption and a secondary stimulation of parathormone secretion, the latter promoting phosphaturia and hypophosphataemia. Increased production of alkaline phosphatase reflects increased osteoblast activity, stimulated in response to bone resorption.

diabetes insipidus. The increased ionized calcium level impairs the ADH-mediated augmentation of water reabsorption in the distal nephron. Sodium and potassium excretion is increased, leading to volume contraction, prerenal reduction of GFR (Ch. 20, p. 1035) and a further decrease in renal calcium excretion. The situation is com-

18

CASE STUDY 18.3 HYPERCALCAEMIA A 64-year-old man presented with nocturia and hesitancy of micturition. Initial clinical assessment demonstrated prostatic enlargement, and investigation showed elevation of serum prostate-specific antigen (PSA, 12ng/mL; reference range <4.1ng/ mL). In his previous medical history he described a single episode of acute left loin pain radiating to the left groin 6 months previously, which had been treated with intramuscular pethidine; these symptoms had not recurred. Prostatic biopsies confirmed adenocarcinoma of the prostate (T3, NO, MO Gleason grade 3-4). He was treated with radical radiotherapy and commenced on GnRH agonist therapy in the form of goserelin subcutaneously every 4 weeks. During follow-up 1 year later he complained of fatigue and generalized aches and pains. Investigations demonstrated serum calcium 2.83 mmol/L, phosphate 1.12mmol/L, alkaline phosphatase 96IU/L, sodium 140 mmol/L, potassium 4.5 mmol/L, chloride 109 mmol/L, bicarbonate 25 mmol/L, urea 8.5 mmol/L, creatinine 130(imol/L and serum PSA 2.5 ng/mL. Plain radiology demonstrated multiple calculi in both kidneys. Examination at that stage indicated a well-nourished man with clinically normal thyroid function. He was normotensive, there was no corneal calcification, and the remainder of the examination was unremarkable. Investigation confirmed hypercalcaemia (corrected serum calcium 2.73 mmol/L) and serum parathormone 172pg/mL (reference range 10-50 pg/mL). Urine calcium excretion was 8.7mmol/24h. Full blood count was unremarkable and ESR was normal at 17mm/lst

hour; serum albumin was 42 g/L and total globulin 26 g/L.

Questions

1. Wliat is the most likely diagnosis? 2. What features in the clinical presentation and investigations,point towards a benign cause of hypercalcaemia? 3. How might the increased fatiguability and musculoskeletal symptpms be explained?

Discussion The hypercalcaemia is most likely to be due to primary hyperparathyroidism. This is supported by the unequivocally raised serum parathormone level, and the pattern of serum electrolytes, with a high normal serum chloride, and low normal serum bicarbonate, is consistent with this diagnosis. The serum phosphate concentration was in the mid-normal range and was therefore atypical for primary hyperparathyroidism. However, the modest elevation of serum creatinine, suggesting some reduction in glomerular filtration rate, might explain this. Importantly, urine calcium excretion was not decreased, thereby excluding a diagnosis of hypocalciuric hypercalcaemia, an improbable diagnosis given the previous history suggesting renal colic and the radiological evidence of renal calcification. Benign hypercalcaemia due to hyperparathyroidism is more likely

than a malignant aetiology in the context of non-specific symptoms and only modest serum calcium elevation. Furthermore, the normal haemoglobin and serum albumin provide some circumstantial evidence against disseminated malignancy. The finding of renal calculi also suggests hypercalcaemia of longer duration than is usually the case with malignant hypercalcaemia, whether due to bone metastases or to parathormone-related peptide (PTHrp) secretion. The normal ESR and serum total globulin concentration provide good evidence against multiple myeloma, but a serum protein electrophoresis would be required to definitely exclude this, especially in view of the musculoskeletal symptoms. Although humoral hypercalcaemia of malignancy (PTH-rp secretion) may produce some biochemical features in common with hyperparathyroidism, PTH-rp is not detectable in PTH assays in current use, and this diagnosis is effectively excluded by the documented elevation of serum PTH, which would be suppressed in hypercalcaemia from causes other than hyperparathyroidism. Although frequently absent, nonspecific symptoms of increased fatigue and musculoskeletal discomfort are well recognized in primary hyperparathyroidism and do not necessarily indicate parathyroid bone disease. The normal serum alkaline phosphatase argues against a substantial degree of bone resorption in this case, although this is a common finding in more advanced disease. Myeloma and bone metastasis should be considered as a cause of skeletal symptoms, but are effectively excluded in this instance.

pounded if the patient is ill and confined to bed, because immobilization accelerates bone resorption and the mobilization of skeletal calcium.

Later, particularly when prerenal uraemia develops, the patient develops nausea and vomiting.

Gastrointestinal

Symptoms commonly include lassitude, anxiety and depression, or any psychosis to which that person may

Common early symptoms are anorexia and constipation.

Neurological

967

CASE STUDY 18.4 ACCELERATING HYPERCALCAEMIA A 54-year-old man presented to his GP with a 48-hour history of nausea and vomiting. He described weight loss over the preceding 6 weeks and had experienced several episodes of haemoptysis. He had no significant previous medical history but had smoked heavily throughout his adult life. Examination demonstrated obvious weight loss, pallor of the mucous membranes, and dullness to percussion and reduced breath sounds at the right lung base; abdominal examination was unremarkable. Further investigation revealed haemoglobin 9.8g/dL, MCV 88, serum urea 11.7mmol/L, creatinine 138umol/L, sodium 133mmol/L, potassium 3.4 mmol/L, calcium 4.1 mmol/L, phosphate 0.82 mmol/L and alkaline phosphatase 163IU/L (NR < 120). He was referred urgently to the local hospital for further management. On admission, chest X-ray demonstrated collapse of the right lower lobe, hypercalcaemia was confirmed, and serum sent for parathormone measurement.

Questions 1. What clinical features point towards a malignant cause for the hypercalcaemia? 2; What is the most likely mechanism for the hypercateaemia? 3. What is the immediate management?

Discussion The patient's acute symptoms at presentation are attributable to accelerating hypercalcaemia. The severity of the hypercalcaemia, the associated normochromic anaemia and the preceding history of weight loss are all suggestive of a malignant cause. Furthermore, haemoptysis and the presence of focal chest signs in a patient with a long smoking history points towards bronchial carcinoma as the underlying aetiology. Hypercalcaemia as a complication of bronchial carcinoma is most commonly due to the production of parathormone-related peptide (PTHrP) by squamous carcinoma cells. PTH-rP promotes calcium absorption in the intestine and renal tubules and also increases net calcium loss from bone. In contrast to primary hyperparathyroidism, the production of this humoral factor is not subject to negative feedback control. Continued PHT-rP release, the progressive salt and water depletion which occur as a combined result of reduced renal concentrating capacity, and vomiting induced by hypercalcaemia, result in accelerating and eventually life-threatening hypercalcaemia and salt and water depletion. The emergency management of this condition depends primarily on the restoration of extracellular fluid volume by means of isotonic (0.9%) saline infusion, which should be

be otherwise predisposed. In severe hypercalcaemia (calcium levels above 4mmol/L), confusion and drowsiness may be followed by coma. Regardless of cause, hypercalcaemia is liable to be complicated by ectopic calcification. This is particularly likely

1

968

Fig. 18.7

administered at an initial rate of 1.0 L every 1-2 hours. Restoration of ECF volume allows the further promotion of calciuresis by means of the addition of a loop diuretic, e.g. furosemide (frusemide), with continued saline infusion. It is obviously important that frusemide should not be given until the patient has been adequately rehydrated. In parallel with these emergency measures, an infusion of a bisphosphonate (e.g. pamidronate) should be commenced. The effects of this will not be evident for 24 hours or more, but it is essential in limiting subsequent relapse of the condition and can be repeated as required. The availability of parenteral bisphosphonate therapy has eliminated the use of other measures, such as intravenous phosphate and mithramycin. Ultimately the successful treatment of malignant hypercalcaemia will depend on management of the primary lesion, but this is frequently not possible, so continuing oral or interval intravenous bisphosphonate therapy is an important component of palliative care.

if the hypercalcaemia is of long standing, as in primary hyperparathyroidism or sarcoidosis. The renal tract is particularly susceptible, and renal stones or nephrocalcinosis (Fig. 18.8) commonly occur. Frequently there is calcification of blood vessels, especially the media of arteries. Calcification bands on the medial or lateral edge of the cornea may be visible (band keratopathy), 1 and conjunctival inflammation may develop if hypercalcaemia is severe. Acute pancreatitis is an important complication, particularly if other predisposing factors, such as high alcohol intake, are also present.

18

TABLE 18.6 Causes of hypercalcaemia Intestinal absorption

Bone resorption

Urine calcium

PTH

1,25(OH)D

Malignancy - osteolytic metastases (carcinoma of breast, lung, thyroid. Myeloma) - humoral - PTH-rp (squamous carcinomas, renal cell carcinoma) - 1,25(OH)D produced by tumour (lymphoma) Primary Hyperparathyroidism - adenoma - hyperplasia - carcinoma Renal hyperparathyroidism (severe) Extrarenal1,25(OH)D synthesis ' - sarcoidosis - tuberculosis - glucocorticoid deficiency (Addison's) Vitamin D intoxication Milk alkali syndrome Familial hypocalciuric hypercalcaemia Drugs - thiazide diuretics - lithium *only when hypercalcaemia overwhelms the renal tubular reabsorptive capacity. ** suppressed by hypercalcaemia. Specific assay for PTH-rp shows elevated PTH-rp in plasma. ***high if 1a-hydroxylated compound ingested (e.g. calcitriol). Low or normal if native vitamin D ingested.

Finally, the cardiovascular system is also susceptible through the increased incidence of hypertension associated with chronic hypercalcaemia. The ECG shows shortening of the QT interval.

Management

Investigation

General measures Symptomatic hypercalcaemia requires measures directed towards rehydration and appropriate sodium and potassium repletion. A forced saline diuresis (normal saline given i.v. at 4-6L/day) substantially and immediately augments the urinary excretion of calcium. A further increase

Potentially useful investigations in hypercalcaemia are listed in Table 18.7. The steroid suppression test aims to identify those cases in which pharmacological doses of glucocorticoids significantly reduce the elevated plasma calcium. Hypercalcaemia due to extrarenal 1,25(OH)2D synthesis usually suppresses fully, and this is likely to be the case with underlying granulomatous disease (e.g. sarcoidosis, tuberculosis), some malignancies (e.g. leukaemia, lymphoma) and glucocorticoid deficiency (Addison's disease). Prednisolone is given at 30mg/day for 10 days, with measurement of plasma calcium before, after 5 days and after 10 days of treatment.

Asymptomatic hypercalcaemia requires no therapy except appropriate management of the underlying condition.

SUMMARY 2 Treatment of severe hypercalcaemia • • • •

Correct salt and water deficit with normal saline Begin forced saline diuresis + frusemide Give bisphosphonate, e.g. pamidronate, clodronate Treat underlying cause if possible

969

TABLE 18.7 Diagnostic yield in hypercalcaemia Investigation

Potential information

Full blood count

Anaemia suggests non-parathyroid aetiology if renal function normal Usually over 80 in myeloma Upper normal in hyperparathyroidism Low normal in hyperparathyroidism Significantly raised in the tertiary autonomous hyperparathyroidism of renal failure. Moderately raised in renal impairment secondary to hypercalcaemia Monoclonal band suggests myeloma Presence of light chains suggests myeloma

ESR Plasma chloride Plasma bicarbonate Creatinine

Plasma protein electrophoresis Bradshaw's test and urinary immunoelectrophoresis Plasma phosphate Plasma PTH

Steroid suppression test FIG. 18.8 Nephrocalcinosis in a patient with severe primary hyperparathyroidism 24 h urine calcium Chest X-ray

in calciuria may be induced by the addition of frusemide (20 mg/h), provided that the saline infusion is fast enough to keep the patient volume expanded. With such treatment alone, the plasma calcium will usually come down to safe levels (3.5mmol/L or less). Specific measures Specific treatment depends on the cause. Glucocorticoids are only effective for the hypercalcaemia of sarcoidosis, vitamin D intoxication, lymphomas and Addison's disease. For the first two, a high dose rapidly tapering down is used. The treatment of choice in hypercalcaemia of malignancy and hyperparathyroidism is usually the bisphosphonate pamidronate, which reliably restores calcium levels to normal within 5 days when infused as a single dose of 60 mg over 4 hours. Clodronate is equally effective, and can be given orally as maintenance treatment. The drugs act by inhibiting osteoclastic bone resorption, but because of a delayed action (24-72 hours) saline diuresis remains an important part of early management. Calcitonin given at the same time accelerates the fall in calcium by up to 24 hours. Sodium phosphate infusion (l0mmol/h for a maximum of 10 hours) is very effective in the short term, but is dangerous as it lowers calcium principally by causing microscopic precipitation of calcium phosphate in soft

1MCQ18.4

970

2

Fig. 18.8

Isotope bone scan X-ray hands

Low normal in primary hyperparathyroidism unless renal function impaired Upper normal or raised in hyperparathyroidism Suppressed with other causes of hypercalcaemia Sarcoid hypercalcaemia always suppresses; hyperparathyroid never; malignant causes occasionally Decreased in familial hypocalciuric hypercalcaemia Look for primary neoplasm; metastases; hilar lymphadenopathy suggesting sarcoid 'Hot spots' suggest metastatic malignant disease Subperiosteal erosions suggest hyperparathyroidism

tissues; this is particularly marked if initial plasma phosphate is high. 1

SPECIFIC CAUSES OF HYPERCALCAEMIA Primary hyperparathyroidism Primary overactivity of one or more parathyroid glands is common, especially in elderly women. In the UK, routine measurement of serum calcium in the hospital population reveals an incidence of 1 in 100. In the general population the incidence is 1 in 500 to 1 in 1000. Of these, approximately three-quarters are associated with a single adenoma. In the others there is either more than one adenoma, or four-gland hyperplasia. In about 1 % of cases a parathyroid carcinoma is the cause. Adenoma or hyperplasia may be associated with one of the syndromes (I or II) of multiple endocrine neoplasia (MEN) (p. 938). In MEN I, inactivation of a tumour suppressor gene (meniri) on 11ql3 is the underlying mechanism. In primary hyperparathyroidism PTH is elevated, usually as a result of a parathyroid adenoma, and stimulates osteoclastic bone resorption, tubular reabsorption of

calcium and the activity of 1-hydroxylase in the proximal renal tubular cells. This in turn promotes conversion of 25(OH)D to 1,25(OH)2D, which acts on the upper small intestine to increase active calcium transport. Because PTH also enhances renal phosphate excretion, the serum phosphate level falls. Clinical features The clinical features of primary hyperparathyroidism are variable. Patients commonly remain asymptomatic for long periods, and the finding of a mild elevation of serum calcium in an elderly asymptomatic patient requires no active intervention. In more severe cases the symptoms may be those of hypercalcaemia (p. 965). Immobilization is a potent precipitating factor and may lead to a rapid rise in serum calcium. Renal complications include renal calculi, which may present with renal colic and nephrocalcinosis (Fig. 18.7). This is particularly likely in warm climates, where insensible fluid and sodium loss are excessive and urine flow rate is low. Hypertension is common, even in the absence of overt renal damage. It may respond to surgical treatment of hyperparathyroidism. The bone disease of hyperparathyroidism is characterized by high turnover and remodelling rates. Bone is lost progressively, mainly from cortical sites - osteoporosis and increased fracture risk may ensue. Very infrequently, osteitis fibrosa cystica occurs, with the development of socalled 'brown tumours' of the mandible or long bones. Again, the finding of such a lesion is an indication to check the serum biochemistry. The typical radiological lesion is subperiosteal erosion in the phalanges of the hands (see Fig. 18.4). Management Management depends on the severity of the hypercalcaemia and the presence of complications. In cases with severe clinical hypercalcaemia, emergency treatment should be initiated along the lines outlined on page 970. The only curative therapy is parathyroid surgery. Modern diagnostic techniques which may help to localize a tumour include 99mTc sestamibi 2 and ultrasound scanning of the neck, and (rarely) catheterization of the veins draining from the neck for measurement of PTH in blood from different sites. High-resolution CT or MRI scanning may also be of value. In recent years there has been a shift away from the conservative approach to asymptomatic primary hyperparathyroidism. This has been based on the recognition of osteopenia as a long-term risk, especially in females, and the fact that subtle deficits in wellbeing may be present. For these reasons, we believe that surgical therapy should be the general rule, unless there are obvious contraindications. Surgery Surgical expertise in identifying all four parathyroids will determine the success of the operation. Adenomas of the lower parathyroids (derived from the upper pair of branchial arches) are sometimes retrosternal. Adenoma-

tous tissue should be removed and a least one other gland biopsied. With four-gland hyperplasia the safe course is probably to remove all four, and then to maintain the hypoparathyroid patient on 1,25(OH)2D3 or 1(OH)D3 (alfacalcidol), a synthetic precursor of 1,25(OH)2D3. Preand postoperative treatment with 1,25-(OH)2D3 or 1(OH)D3 may be helpful in preventing postoperative hypocalcaemia, which is especially likely to occur when there are bone lesions.

18

Other measures In some patients surgical cure is not achieved. Failure to remove enough parathyroid tissue may leave the patient with 'incurable' hyperparathyroidism. Some patients refuse surgery, and because primary hyperparathyroidism is principally a disease of the elderly, some will be deemed unfit for surgery. Conservative management of these patients may entail observation only, with monitoring of plasma calcium concentration and checking for complications. More active measures include the use of oral phosphate supplements, oestrogens, or the oral bisphosphonates such as clodronate (p. 970). These measures aim to protect the skeleton from the excess PTH. However, although these measures will alleviate hypercalciuria and hypercalcaemia in the short and medium term, surgery remains the preferred treatment in most younger and symptomatic older patients. In the future it is likely that calcimimetic drugs will provide the best non-surgical option. These agents bind to the extracellular calcium receptor on the parathyroid cells and increase the sensitivity of the receptor to extracellular calcium. Thus the parathyroid cells respond as if the plasma calcium is higher than it really is - both PTH and calcium fall as a result.

HYPERCALCAEMIA OF MALIGNANCY Malignancy is an important cause of hypercalcaemia. There are several mechanisms (Fig. 18.9): • Direct invasion of bone and the production of local factors, which either destroy bone directly or stimulate local osteoclasts to do so, lead to accelerated bone resorption. Breast cancer usually causes hypercalcaemia by this mechanism. If the serum calcium is elevated in a patient with breast cancer, it almost always indicates that bony metastases are present. • Haematological malignancies may produce cytokines which act as osteoclast-activating factors. Hypercalcaemia in the presence of multiple myeloma is often by this mechanism. • Some solid tumours (such as renal carcinomas and squamous carcinomas of the lung and oesophagus may cause hypercalcaemia by secreting PTH-related peptide (PTH-rp), a factor which binds to the PTH receptor and activates bone resorption and/or enhances renal tubular calcium reabsorption and the renal production of

971

FIG. 18.9 Mechanism of malignancy in hypercalcaemia

1,25(OH)2D. This peptide has partial sequence homology with PTH (accounting for the hypercalcaemia) but is not detected by PTH immunoassays.

Clinical features These are of hypercalcaemia normally associated with those of the underlying cancer. The hypercalcaemia is usually rapidly progressive and severe, with salt and water depletion and disturbance of renal and CNS function.

in the form of calcium carbonate, to relieve dyspepsia. The hypercalcaemia is accompanied by metabolic alkalosis and is due to a combination of increased intestinal calcium absorption and decreased renal calcium excretion. Addison's disease may present with hypercalcaemia through extrarenal production of 1,25(OH)2D. Glucocorticoid replacement leads to a prompt response. Thyrotoxicosis. Thyroid hormones increase bone turnover and, in excess, cause net efflux of calcium from bone. Some 30% of patients display hypercalciuria and/or hypercalcaemia, although the latter is not severe. These abnormalities are corrected by treatment of the underlying thyrotoxicosis. Familial benign hypocalciuric hypercalcaemia is a syndrome that may be detected at any age. It is usually asymptomatic and is not associated with renal damage (although it may cause pancreatitis). It is inherited as an autosomal dominant condition, and so can usually be excluded if both parents are available for blood sampling. It is due to a defect in the parathyroid gland calcium sensor - an inactivating mutation reduces the sensitivity of the receptor to calcium, and the parathyroid gland therefore sets the serum calcium at a supranormal level. The same mutation affects the calcium receptor in the renal tubular cells, and there mediates the hypocalciuria. 1 Thiazide diuretics. These drugs are anticalciuric. The hypercalcaemia, when present, is mild unless other factors coexist favouring the elevation of calcium, for example Paget's disease, mild hyperparathyroidism or immobilization.

Management This is as for hypercalcaemia generally (p. 971). Saline rehydration and diuresis and bisphosphonates are the main methods. The cancer is treated when possible.

HYPOCALCAEMIA

Causes and pathophysiology MISCELLANEOUS HYPERCALCAEMIC SYNDROMES In sarcoidosis, and some other granulomatous diseases, macrophages in the diseased tissue produce 1,25(OH)2D outside the kidneys in amounts that are sufficient to produce hypercalcaemia. Calcium is absorbed by the intestine in increasing amounts, PTH is suppressed, and the normal feedback of the vitamin D-endocrine system is effectively sabotaged. The same mechanism occurs in some lymphomas and sarcomas. Treatment is with small doses of glucocorticoids - the extrarenal synthesis of 1,25(OH)2D is exquisitely sensitive to steroids. Milk alkali syndrome, now rarely seen, is associated with excessive ingestion of milk and alkali, particularly if taken 1

972

Case 18.1

The principal causes of hypocalcaemia are listed in Table 18.8. As the major calcium-binding protein in plasma is albumin, hypoalbuminaemia is associated with a proportionate fall in plasma total calcium (see correction formula on p. 957), though not in ionized calcium. The ionized calcium concentration is affected by the blood pH, as the avidity of albumin for calcium falls at acid pH and rises with alkalosis. Hence, respiratory alkalosis due to overbreathing can lead to depression of ionic calcium and tetany. Rarely, infusion of citrate (e.g. in repeated blood transfusions) leads to a serious fall in ionized calcium by complexing the ion. In hypoparathyroidism (p. 973) the major abnormalities result from failure of PTH-dependent renal synthesis of 1,25(OH)2D, leading to a reduction of intestinal calcium absorption (see Fig. 18.3). This is compounded by failure of osteoclastic bone resorption, by enhanced (inappropriate) urine calcium excretion and by hyperphosphataemia. These are all due to loss of the effects of PTH on bone and kidney. Secondary hyperparathyroidism is a consequence of hypocalcaemia (Table 18.9). In chronic renal failure

TABLE 18.8 Causes of hypocalcaemia

TABLE 18.9 Causes of secondary hyperparathyroidism

Cause

Comments

Cause

Example

With low levels of PTH Hypoparathyroidism Postsurgical Congenital Idiopathic (autoimmune)

Vitamin D deficiency Phosphate high Alkaline phosphatase and renal function normal

Lack of solar exposure Dietary deficiency Increased catabolism - drug-induced (e.g. anticonvulsants, rifampicin)

Dietary calcium deficiency

Low calcium intake High phytic acid intake (chapatis, wholemeal bread)

Gastrointestinal causes of low calcium or vitamin D intake

Gastric surgery Partial gastrectomy Gastroenterostomy Vagotomy and pyloroplasty Jejunal diverticulosis Crohn's disease Small bowel resection Gluten enteropathy

Renal

Chronic renal failure Phosphate retention Reduced 1a-hydroxylation of 25(OH)D

With normal or high levels of PTH Rickets/osteomalacia Sunlight/dietary deficiency Malabsorption Increased D metabolism Tumour-induced Chronic renal failure Pseudohypoparathyroidism

Phosphate low Alkaline phosphatase high Calcium deficiency Drug-induced Urea, creatinine, phosphate high Phosphate high, somatic manifestations (type I only)

(CRF), hypocalcaemia results from phosphate retention and loss of renal substance, with failure to produce normal amounts of 1,25(OH)2D. The rise in serum phosphate depresses the serum calcium, partly by leading to microprecipitation of calcium phosphate in soft tissue. The associated bone disease is discussed on page 1055. In CRF, metabolic acidosis is commonly present and this diminishes the fall in ionized calcium. Tetany in CRF is rare in adults (but more common in children), at least until the acidosis is corrected. Hypocalcaemia in vitamin D deficiency is usually not severe, because secondary hyperparathyroidism enhances bone resorption and maintains the serum calcium level. The typical biochemical changes are borderline low serum calcium, a low phosphate, and elevated alkaline phosphatase levels. If the parathyroid glands do not respond, or calcium intake is very low, hypocalcaemia may be more profound. In CRF and vitamin D deficiency, and in other conditions (see Table 18.9), the parathyroid hypersecretion (secondary hyperparathyroidism) is an appropriate response to threatened or actual hypocalcaemia. In some patients, however, long-standing appropriate hypersecretion (secondary hyperparathyroidism) transforms to inappropriate autonomy and consequent hypercalcaemia. This is called tertiary hyperparathyroidism, and histologically the glands show nodular hyperplasia. The nodules consists of clonally proliferating cells and are therefore resistant to the normal suppressive influence of 1,25(OH)2D.

Specific causes of hypocalcaemia Hypoparathyroidism There are three principal types of hypoparathyroidism, all of which are associated with hypocalcaemia and hyperphosphataemia despite normal renal function.

18

Postsurgical hypoparathyroidism Postsurgical hypoparathyroidism is the commonest cause of hypocalcaemia and follows thyroid or parathyroid surgery. Estimates of its frequency vary from 0.2% to 22%, depending on diagnostic criteria and the length of followup after surgery. Vascular or mechanical trauma to the parathyroid glands may contribute to the subsequent parathyroid insufficiency. Idiopathic hypoparathyroidism This is usually acquired in adolescence or early adult life, and may be associated with autoantibodies to parathyroid tissue, and with other organ-specific autoimmune diseases such as Addison's disease and immune thyroiditis. Some cases are familial. An autoimmune aetiology has been postulated but not proven. Patients may suffer from widespread Candida infections and dystrophic nails, alopecia, calcification in the basal ganglia and cataracts, in addition to other symptoms and signs of hypocalcaemia. PTH is undetectable despite hypocalcaemia, and the lack of its effect on the kidney is reflected by renal phosphate retention leading to hyperphosphataemia and low conversion of 25(OH)D to 1,25(OH)2D. Administration of PTH corrects all the biochemical abnormalities, indicating that target organ responsiveness to PTH remains normal in this condition. However, long-term treatment with PTH is impractical and management is with 1,25(OH)2D3 or 1(OH)D3. Most cases respond well to 1-3 ug daily of either. There is, however, little margin for error, and the dose should be titrated carefully against the serum calcium

973

levels. Oral calcium supplements enhance the effect of any given dose. 1 Pseudohypoparathyroidism Pseudohypoparathyroidism is a rare disorder thought to be due to a defect in the PTH receptor or its coupling to adenylate cyclase. The clinical and biochemical features of hypoparathyroidism are associated with characteristic somatic manifestations; these include short stature, mental deficiency, and unusually short fourth and fifth metacarpals and metatarsals. There is a partial or complete failure of response (urinary cAMP or phosphaturia) to PTH, and high serum PTH levels. As with true hypoparathyroidism, 1,25(OH)2D levels are low. Some patients have the somatic manifestations without the hypocalcaemia (pseudopseudohypoparathyroidism). In some cases the biochemical changes are associated, paradoxically, with hyperparathyroid bone disease, possibly reflecting isolated renal resistance with normal skeletal responsiveness to PTH. 2 Hypomognesaemia Hypomagnesaemia is a rare but important cause of hypocalcaemia. It is refractory to calcium, but responds to magnesium therapy (p. 956).

Clinical features The clinical manifestations of hypocalcaemia are principally in the nervous system. Tetany, painful cramps and tingling in the extremities may occur, and muscle spasm may lead to a characteristic position of the hand, the socalled 'main d'accouncheur. These symptoms may be precipitated by exercise. Laryngeal spasm may lead to stridor and obstructed respiration. Increased neuromuscular excitability in the presence of hypocalcaemia may be demonstrated by gentle percussion over the facial nerve at its point of exit from the parotid gland. This produces ipsilateral twitching contraction at the angle of the mouth (Chvostek's sign). Hypocalcaemia lowers the epileptic seizure threshold, and fits may be the presenting symptom. Hypocalcaemia, particularly when chronic, may present with psychiatric manifestations varying from general malaise to overt psychosis. Clinically, the differential diagnosis is principally of tetany due to anxiety, leading to hyperventilation and respiratory alkalosis, or hypokalaemic metabolic alkalosis.

Differential diagnosis of hypocalcaemia Correction for a low serum albumin concentration will allow the distinction between true (ionized) hypocal-

caemia and the spurious fall due to hypoalbuminaemia. As with other calcium disorders, the combination of clinical context and routine biochemistry will usually establish the cause. Uraemia and elevated serum creatinine concentration point to CRF. Hypoparathyroidism and pseudohypoparathyroidism are associated with normal renal function and an elevated serum phosphate level. There may also be clinical pointers, e.g. recent thyroid or parathyroid surgery for the former, skeletal abnormalities for the latter.

Management Hypocalcaemic tetany responds rapidly to the intravenous infusion of 10mL of 10% calcium gluconate. The response, however, is short-lived. The underlying cause should be treated, and in the case of hypoparathyroidism this has been greatly helped by the availability of 1,25(OH)2D3 (calcitriol) or 1-hydroxyvitamin D3 (alfacalcidol). The latter is then 25-hydroxylated in the liver to yield 1,25(OH)2D3. By virtue of their high potency, rapid onset and short duration of action, these compounds have converted a very difficult therapeutic problem into an easy one. Maintenance treatment of hypoparathyroidism should aim to keep the serum calcium in the lower part of the reference range, in order to avoid the increased risk of renal calculi consequent upon the increased urine calcium for a given filtered load of calcium. Hypocalcaemia in osteomalacia due to vitamin D deficiency is seldom marked; here, correct treatment is with vitamin D itself, rather than the active metabolite (p. 955).

PAGET'S DISEASE OF BONE In 1879, Sir James Paget described a bone disease which he named 'osteitis deformans'. Paget's disease is common and its frequency increases with age. In the UK as many as 4% of the population over the age of 40 may have evidence of the disease, although most have no associated symptoms. It is a chronic focal disease which may affect virtually any bone in the body. It varies widely in extent, with individual patients having anything from one to 20 or 30 bones affected. Even in a patient with extensive disease, the structure of unaffected bones appears to be quite normal. In general, the axial bones and large long bones are more commonly affected than the small bones of the hands or feet, or the ribs. Within an individual, the distribution appears to be random. In most patients the majority of affected bones are asymptomatic. This does not mean that they will always remain so.

Aetiology and pathophysiology 974

1

MCQ 18.5

2

3

Figs 18.9-18.11

Case 18.2

The primary disease process in Paget's disease appears to be in the osteoclasts of affected bone, and there is much circumstantial evidence pointing to a viral aetiology. These

18

CASE STUDY 18.5 HYPOCALCAEMIA A 67-year-old woman was admitted for an elective thyroidectomy for a large multinodular goitre. There was no significant previous medical history, she was euthyroid and was not on any medication. At surgery, a near total thyroidectomy was performed. Two parathyroid glands were identified and preserved. Twenty-four hours postoperatively she complained of tingling in her hands and feet. Examination demonstrated positive Chvostek and Trousseau signs and serum calcium was measured at 1.73mmol/L, with serum phosphate 1.49mmol/L. She was treated with a slow infusion of intravenous calcium gluconate and her symptoms resolved. Maintenance treatment with oral calcium was commenced, and the following day serum calcium was 1.83mmol/L. At that stage she was commenced on Icchydroxycholecalciferol 2 ug daily, with subsequent normalization of serum calcium (2.23mmol/L), and thyroxine replacement therapy. Histology of the

resection specimen confirmed multinodular goitre and two normal parathyroid glands. Her surgical recovery was otherwise uncomplicated and she was discharged from hospital. On outpatient review some 4 weeks later her serum calcium was 2.73 mmol/L; the dose of loc-hydroxycholecalciferol was reduced to 1 ug daily. Repeat serum calcium 1 week later was 2.56 mmol/L; lochydroxycholecalciferol was discontinued. Subsequent serum calcium measurements were within normal limits.

Questions 1, What was the mechanism for the hypocalcaemia in the immediate postoperative period? 2. What features suggested a likelihood of reversal of this phenomenon?

osteoclasts are abnormal in both structure and function, being larger than normal with more nuclei, and resorbing bone apparently without regard to normal anatomical restraints. Affected osteoclasts contain numerous cytoplasmic and nuclear particles thought to be viral nucleocapsids. Immunological and other evidence from molecular biology techniques suggests that a virus similar to the measles virus may be responsible. There is also evidence that at least some examples of Paget's disease may have a primarily genetic basis, with autosomal dominant inheritance linked to mutations of the gene for receptor activator of NFKB (RANK) on chromosome 18q21-22. Clearly the viral and genetic hypotheses are not mutually exclusive. Even if there is a genetic predisposition, other factors must determine which bones are affected. The apparent decline in the incidence of the condition in recent years points towards a role for an infective agent. Residual osteoclasts in bone behind the leading edge of advancing disease remain abnormal in size and shape, and continue to resorb bone chaotically. Wherever it occurs, bone resorption is followed by bone formation which, after an initial lag, catches up and keeps pace with resorption. However, the new bone laid down follows the chaotic

Discussion The occurrence of acute symptomatic hypocalcaemia after thyroid surgery is strongly suggestive of hypoparathyroidism. The severity of this phenomenon may be increased by pre-existing thyrotoxicosis, which was not evident in this case, by dietary vitamin D deficiency, or by rarer causes of metabolic bone disease. The fact that two parathyroid glands were identified and preserved at surgery suggests trauma/ischaemia as the underlying mechanism, and would suggest (but not confirm) subsequent improvement. This case highlights the importance of frequent monitoring of serum calcium after initiation of lahydroxycholecalciferol therapy in order to avoid overtreatment, which might delay the recovery of residual parathyroid function because of the inhibitory effect of modest hypercalcaemia.

pattern of bone resorption, and the gross structure of the whole bone therefore becomes progressively distorted. Affected bones are greatly weakened and become expanded and bent - this is particularly the case for the major long weight-bearing bones, such as the tibia and femur, as well as the humerus and radius, which may develop partial fractures. The latter are unusual in extending from the inner cortex outwards, and being concentrated on the convexity of the bowing bone (Fig. 18.10). Vertebrae and the pelvis become distorted. The skull thickens and its front and back sink, as the vertebral column selectively supports its centre. 3 Secondary bone changes include osteoarthritis, e.g. of the hips and knees. Within an individual the disease usually appears to be at. the same stage of development at different sites, and unaffected bones remain so throughout the disease. This possibly reflects a single episode of widespread viral infection of bloodborne osteoclast precursors, which were then carried into affected bones. Continuing viral replication within affected cells, combined with transfection by cell fusion (cell fusion being a normal property of the osteoclast), appears to explain why the disease is rampant in affected bones while initially unaffected bones generally seem to remain normal.

975

• A sharp pain, worse on movement or pressure, at the site of a partial or complete fracture; • The pain of secondary osteoarthritis; • Pain due to pressure on nerve roots. The patient may complain of deformity of an affected bone or bones, or enlargement and irregularity of the skull. There may be pressure effects, including cranial nerve lesions - of which tinnitus and deafness (nerve, as well as conductive) are very common - if the disease involves the skull. Occasionally, Paget's disease of the spine leads to paraparesis. Platybasia of the skull may lead to obstruction of the aqueduct and internal hydrocephalus. Pagetic bone is highly vascular, and the patient is often aware of the warmth of an affected limb, especially of the tibia. The limb is obviously hot to touch. Increased vascularity of the skull leads to prominence of the superficial temporal arteries and, occasionally, to distended external jugular veins. Cardiac output is modestly increased in extensive Paget's disease, as a result of vascular shunting of blood through diseased bone. This may occasionally lead to high-output congestive cardiac failure. Patients with extensive Paget's disease commonly have generalized malaise and lethargy, which may improve with effective treatment. An uncommon but serious complication of the condition is the development of bone sarcoma (usually osteogenic sarcoma), which is marked by expansion of the affected bone and further elevation of serum alkaline phosphatase activity. This complication may arise simultaneously at more than one pagetic site, and probably develops in less than 1 % of cases. FIG. 18.10 Gross expansion and deformity of right tibia in longstanding Paget's disease

Clinical features Since the natural time course of Paget's disease is very long, and many patients have had it for 20-40 years or more before diagnosis, virtually nothing is known of its early clinical features or whether there is an acute illness at the onset of the putative initial viral infection. Many patients present fortuitously, either because a biochemical screen has revealed a high serum alkaline phosphatase level, or because an X-ray has revealed an affected bone. Typically the disease starts in one or more specific areas, e.g. the epiphysis of a long bone or the skull base, and spreads through it with a lytic advancing edge at a rate of about 1 cm per year. In the skull this gives rise to the picture of osteoporosis circumscripta, and in long bones to the 'blade of grass' appearance. This is the leading edge of 'vandal' osteoclasts, which resorb bone in an uncontrolled way. Bone pain is of at least four types:

976

• A deep, boring dull pain, probably due to increased vascularity, which is present at rest and not relieved by analgesics;

Investigation Bone resorption is reflected by increased excretion of collagen-derived hydroxyproline or deoxypyridinoline in urine (best measured as the fasting morning urine hydroxyproline (or deoxypyridinoline)/creatinine ratio). Deoxypyridinoline has greater specificity than hydroxyproline. Bone formation is reflected by the level of serum alkaline phosphatase. These two markers, which together provide a measure of bone turnover rate, are a function both of the extent of skeletal involvement and of the disease activity within affected bones. In addition to these measurements, a bone scan is performed using 99mTc-labelled bisphosphonate (Fig. 18.11), followed by radiology of 'hot' bones. It is important to document the baseline state of disease before instituting therapy.

Management Until the late 1960s there was no effective treatment for Paget's disease. Although most patients have a mild condition which is rarely fatal, in some the disease is extremely disabling. Traditional criteria for instituting treatment, hitherto only for symptomatic disease, will change once an effective and simple treatment can be shown to arrest the progress of advancing disease or cure the condition.

18

FIG. 18.11 Bone scan from a patient with extensive Paget's disease A In skull. B In several dorsal vertebrae, sternum and left humerus. C In the whole pelvis.

Most authorities currently recommend treatment for pain, progressive disease with actual or threatened deformity, neurological complications and evidence of widespread disease. The bisphosphonates (diphosphonates) The bisphosphonates are analogues of pyrophosphate that bind to hydroxyapatite, and are therefore taken up by newly laid-down bone collagen matrix. Pamidronate and clodronate are highly effective, especially parenterally. Figure 18.12 illustrates the dramatic response to a short course of intravenous infusions of pamidronate in severe Paget's disease. Remineralization of lytic Paget's disease may be evident radiologically as early as 4 months after the start of a 1-3-month course of therapy, and in many cases is followed by prolonged remission. In the early stages, pamidronate treatment is commonly followed by a transient increase in bone pain and mild fever. This is quickly followed by symptomatic improvement and a reduction in bone pain. Intravenous pamidronate (three 60 mg infusions 2 weeks apart) is now licensed for Paget's disease and is the preferred treatment. Treatment is continued for up to 3 months, but may be repeated at intervals thereafter for recurrence of symptoms. Mithmmycin, although also quite effective, is much more toxic than pamidronate and cannot now be generally recommended for the treatment of Paget's disease. Calcitonin Calcitonin given by repeated injection was the first treatment to have any beneficial effect on bone pain or the

FIG. 18.12 Response to pamidronate treatment in Paget's disease A Paget's disease in a 71-year-old woman before (left) and 9 months after a course of i.v. pamidronate (right). Note restoration of bone in areas of osteolysis. Bone distortion remains. B Decline in alkaline phosphatase level (normal less than 130IU/L) and hydroxyproline/creatinine (HOP/Cr) ratio (normal less than 20umol/mmol) during and after a single course of 12 weekly 30 mg infusions of i.v. pamidronate (APD). Note decline in activity. Upper limit of normal indicated by blue dotted line.

977

course of the disease. The hormone inhibits osteoclasts. There is an abrupt reduction in bone resorption, but bone formation continues. For this reason, the onset of treatment in a severely affected patient with extensive disease leads to a transient fall in serum calcium concentration. Synthetic salmon or human calcitonin is given in doses of 50-100IU by subcutaneous injection three times a week. The hydroxyproline and plasma alkaline phosphatase levels fall, but rarely dramatically, and an improvement in bone structure and recalcification of affected bone occurs. Unfortunately, 'escape' is commonly seen within a year of treatment, usually as a result of antibody production. Side-effects are frequent and include flushing, palpitations, headache and malaise shortly after each injection. 1

MISCELLANEOUS RARE BONE DISEASES HYPOPHOSPHATASIA Hypophosphatasia is characterized by failure of mineralization, possibly as a result of inhibition by increased concentrations of pyrophosphate. Its inheritance is autosomal recessive and it takes its name from the low levels of plasma alkaline phosphatase usually found. The fully expressed severe form of the disease may be lethal neonatally, but later a clinical picture similar to that of severe rickets is found, with abnormal dentition and premature fusion of the cranial sutures. Urinary phosphoethanolamine excretion is increased, but the mechanism is not known. There is no specific therapy.

OSTEOGENESIS IMPERFECTA Osteogenesis imperfecta, which is also termed fragilitas ossium or brittle bone syndrome, is a collection of rare inherited disorders in which there appear to be specific defects in the formation of type I collagen, the predominant collagen in mature bone. There are four main forms: • A dominantly inherited mild form in which structural abnormalities are unusual; • A dominant form with fragility and short stature; • A recessive form with early multiple fractures and skeletal deformity; • A severe, often lethal, form due to new mutations in collagen genes.

1MCQ18.6

978

Abnormal dentition may occur in all forms and in forms 1, 3 and 4 the sclerae are often blue. Treatment is directed towards the prevention of trauma and correction of posttraumatic deformity. Some patients with osteoporosis and a strong family history of the disease may in fact have mild forms of osteogenesis imperfecta.

OSTEOPETROSIS (MARBLE BONE DISEASE, ALBERS-SCHOENBERG DISEASE) The term osteopetrosis describes a number of very rare conditions characterized by increased bone mass throughout the skeleton. Of these, a dominantly inherited and relatively mild form, and a recessively inherited severe form, are numerically the most important. The increase in bone mass is thought to be due to a defect in osteoclast activity. Radiological appearances are characteristic, with patchily increased bone density throughout the skeleton. In severe disease, encroachment of bone on the marrow and neural canals may lead, respectively, to bone marrow failure with leukoerythroblastic anaemia (which may be fatal) and cranial nerve palsies. Hepatosplenomegaly may be evident, indicating extramedullary haemopoiesis. The dominantly inherited mild disease is usually symptomless, but classic radiological changes are found in adults, with increased bone density and alternating dense and lucent bands in the vertebral column. Severe osteopetrosis presents in childhood and is, predictably, associated with a markedly positive calcium balance. Some benefit has been claimed for the treatment of severe disease by calcium restriction. Treatment by marrow transplantation has been tried with some success: transplanted osteoclast precursors, formed in the bone marrow, are able to re-establish normal bone resorption.

FURTHER READING Altmann P, Cunningham J 1987 Postgrad Med I 63: 77. Anderson D C 1995 Paget's disease of bone. Collected reports on the rheumatic diseases. Arthritis and Rheumatism Council, pp. 126-131. Bikle D D, Negro-Vilar A (eds) 1995 Hormonal regulation of bone mineral metabolism. Endocrine Reviews Monographs, vol 4. Cohen R D, Lewis B, Alberti K G M M, Derman A M (eds) 1990 Metabolic and molecular basis of acquired disease. London: Bailliere Tindall. Favus M J (ed) 1993 Primer on the metabolic bone diseases and disorders of mineral metabolism, 2nd edn. American Society for Bone and Mineral Research. Kanis J A 1991 Pathophysiology and the treatment of Paget's disease of bone. London: Martin Dunitz. Mallette L E 1991 The parathyroid polyhormones - new concepts in the spectrum of peptide hormone action. Endocr Rev, 12: 110-117. Mundy G R 1990 Calcium homeostasis: hypercalcaemia and hypocalcaemia, 2nd edn. London: Martin Dunitz. Nordin EEC 1993 Metabolic bone and stone disease, 2nd edn. Edinburgh: Churchill Livingstone.

19

Diabet sMelitus andDisordes ofLipdand Intermediary Metabolism John Wilding and Gareth Williams

Regulation of metabolism 979

Porphyria 1026

Diabetes mellitus 984

Inherited disorders of carbohydrate metabolism 1028

Hypoglycaemia 1018 Plasma lipoproteins and their disorders 1021

Lysosomal storage disorders 1029

REGULATION OF METABOLISM An appreciation of metabolic processes is essential to understand the causes, effects and treatment of diabetes, which is a common, important and interesting disease. The following is therefore a brief account of the normal control of metabolism.

Insulin Synthesis and structure Insulin is synthesized and secreted by the [3 cells, which form the core of the islets of Langerhans. Translation of the insulin gene yields a precursor, preproinsulin, which is cleaved proteolytically to produce proinsulin. This is further split into equimolar amounts of insulin and C peptide, which are released from the B cell by exocytosis of secretory vesicles (Fig. 19.1). Insulin consists of A and B chains, linked by two disulphide bridges. Human insulin differs structurally from porcine insulin in a single amino acid residue, and both these species differ from bovine insulin in two additional residues (Fig. 19.2). Various point mutations in the molecule ('insulinopathies') have been identified, some of which cause glucose intolerance by interfering with the molecule's

biological activity. Insulin belongs to a large family of structurally similar proteins, conserved during evolution, including the epidermal and insulin-like growth factors. C peptide is biologically inert. It is cleared more slowly than insulin and its circulating levels are therefore a more stable indicator of B-cell secretion than insulin itself. Subnormal C-peptide levels ('C-peptide negative') after stimulation by either glucagon (a powerful insulin secretagogue) or a high-carbohydrate meal can be used to identify diabetic patients who are sufficiently insulin deficient as to need insulin replacement. Insulin secretion Some insulin is secreted all the time and is important in restraining glucose release from the liver and preventing triglyceride breakdown in fat. This 'basal' secretion occurs in regular 10-15-minute pulses, which may serve to maintain the sensitivity of insulin receptors on target tissues. Eating induces an additional, rapid surge of insulin secretion, roughly proportional to the size of the meal; levels rise severalfold after an average breakfast (Fig. 19.3).This is due mainly to the secretagogue actions of absorbed nutrients, and also to stimulation by humoral and autonomic nervous signals. Glucose is the main insulin secretagogue, and arginine and other amino acids are also active. Glucose is first metabolized in the B cell to glucose-6phosphate by the enzyme glucokinase. This is the ratelimiting step, and allows the pancreas to 'sense' the extracellular glucose concentration. The further metabolism of glucose via the glycolytic pathway generates ATP. This closes a potassium efflux channel, causing a rise in intracellular potassium, which depolarizes the (3-cell membrane. Depolarization opens calcium channels in the membrane, allowing calcium to enter the (3 cell, and this in turn activates the microtubules that propel insulin-containing vesicles to the cell surface (Fig. 19.4). After an acute secretagogue challenge (e.g. an intravenous glucose bolus), there is an immediate first phase surge of insulin secretion lasting a few minutes, followed by a slowly rising second phase, which declines when the stimulus ends. This pattern is less clear after meals owing to the relatively sluggish absorption from the gut. Prandial insulin release is also stimulated by the vagal innervation of the islet and by the 'incretins'. These are hormones secreted by the gut into the circulation after eating, such as gastric inhibitory peptide and glucagon-like peptide 1 (7-36) amide, a gut-derived product of the preproglucagon gene. Insulin is secreted into the islet capillaries, which drain into the pancreatic vein, a tributary of the portal vein. It is therefore delivered directly to the liver, where it exerts many of its metabolic effects. Insulin is cleared from the circulation within a few minutes, mostly being removed during its first pass through the liver. Metabolic actions of insulin Insulin is an exclusively anabolic hormone, promoting the storage of carbohydrate, fat and protein. It:

979

FIG. 19.1 Electron micrograph showing secretory granules containing insulin in B cells Higher magnification (inset) shows granules that are released from the cell surface by exocytosis.

FIG. 19.2 Structure of insulin Proinsulin is cleaved to produce insulin and the connecting C peptide. The amino acid residues that differ between human, porcine and bovine insulins are highlighted.

SUMMARY 1 Main metabolic effects of insulin Carbohydrate Promotes glycogen formation reduces hepatic glucose output* Inhibits gluconeogenesis Stimulates glucose uptake into muscle and fat Lowers blood glucose concentration Fat Promotes triglyceride formation Inhibits lipolysis and ketogenesis*

Protein Promotes protein synthesis *Action occurring at low (basal) insulin concentrations The counter-regulatory hormones (glucagon, catecholamines, growth hormone and cortisol) oppose these actions

• causes glucose to be converted into glycogen (glycogenesis) in the liver and muscle; • promotes the formation of triglyceride (lipogenesis) in fat; • stimulates protein synthesis in many tissues; • inhibits the opposing catabolic processes of glycogenolysis, lipolysis and proteolysis.

980

Basal insulin levels are enough to inhibit lipolysis and restrain glucose output by the liver, by inhibiting glycogenolysis and gluconeogenesis. The much higher levels that occur postprandially also stimulate the uptake of glucose into muscle and fat, by activating specific glucose transporter proteins in these tissues. Insulin acts through specific receptors on its target cells (Fig. 19.5). Rare cases of congenital insulin resistance have been attributed to mutations affecting the insulin-binding or phosphorylation domains of the receptor (Figs 19.5 and 19.6).

FIG. 19.3 Diurnal changes in plasma insulin and glucose concentrations in non-diabetic subjects Insulin levels vary severalfold, whereas plasma glucose is maintained between tight limits of 4 and 7mmol/L

Counter-regulatory hormones and somatostatin Counter-regulatory hormones oppose the anabolic effects of insulin, either by direct catabolic actions or by making the tissues insensitive to insulin. Glucagon, secreted by the a cells in the islet periphery, powerfully stimulates glycogenolysis and gluconeogenesis in the liver; together, these actions rapidly increase hepatic glucose output and cause hyperglycaemia. Glucagon also stimulates ketogenesis. All these actions are accentuated by insulin deficiency, when glucagon also enhances lipolysis. Catecholamines, either circulating or released locally in tissue by activation of the sympathetic nervous system, cause glycogenolysis and gluconeogenesis in the liver and lipolysis in fat.

19

FIG. 19.4 Insulin secretion Glucose enters the B cell via the GLUT-2 glucose transporter. Within the B cell, glucose is metabolized intially by glucokinase, generating ATP. This closes ATPsensitive potassium channels, causing membrane depolarization, which triggers calcium influx, resulting in exocytosis of insulin secretory granules. Sulphonylurea drugs promote insulin release by closing the potassium channel, mimicking the effect of glucose.

Cortisol renders tissues insensitive to insulin and stimulates proteolysis, lipolysis and gluconeogenesis. Growth hormone also reduces insulin sensitivity and stimulates lipolysis, but cooperates with insulin in promoting protein synthesis (see also p. 894). The counter-regulatory hormones are secreted during stress and assist glycaemic recovery from hypoglycaemia. Their secretion is also influenced by insulin and vice versa, e.g. glucagon potently stimulates insulin release. In contrast, insulin inhibits glucagon secretion; insulin deficiency therefore allows glucagon levels to rise, aggravating catabolism in diabetes. Somatostatin, secreted by the 5 cells of the islet, is not a counter-regulatory hormone but strongly suppresses the release of both insulin and glucagon. It also inhibits the secretion of growth hormone, gastrin and many other peptides. The islets produce a fourth hormone, pancreatic polypeptide, secreted by PP cells. Its function is unknown.

Carbohydrate, fat and protein metabolism The three branches of metabolism - carbohydrate, fat and protein (Figs 19.7-19.9) - are all interlinked at various levels. Each is finely balanced between anabolism, largely mediated by insulin, and catabolism due to the counterregulatory hormones and activation of the sympathetic nervous system.

FIG. 19.5 The insulin receptor The insulin receptor consists of paired a and (3 subunits. Insulin binds to the a subunit, causing a conformational change resulting in activation of tyrosine kinase enzymes which are part of the B-subunit sequence. These become phosphorylated, resulting in a cascade of intracellular events. Other parts of the receptor are responsible for internalization of the receptor, following which insulin is removed and the receptor is recycled.

FIG. 19.6 Intracellular actions of insulin After insulin binds to its receptor it triggers a complex cascade of phosphorylation of intracellular proteins, beginning with insulin receptor substrate 1 (IRS-1). This results in the movement of glucose transporters (GLUT-4 in muscle and fat) to the cell surface; activation of key enzymes involved in glucose metabolism; effects on gene transcription and cell proliferation, which occur in the nucleus.

Carbohydrate metabolism Glucose is the body's main metabolic substrate and is normally the sole fuel for the brain, which uses about 50% of the 200 g of glucose produced by the liver each day. Glucose is supplied steadily to the brain and other tissues by maintaining blood levels between tight limits (4-7mmol/L; see Fig. 19.3). Between meals, blood glucose is determined mainly by

981

FIG. 19.8 Principal pathways of fat metabolism TG, triglyceride; FFA, free fatty acids; Cort, cortisol; Ins, insulin; Cats, catecholamines; GH, growth hormone; Glcg, glucagon; +, -, reactions stimulated, inhibited (shown in one direction only). FIG. 19.7 Principal pathways of carbohydrate metabolism Ins, insulin; Cats, catecholamines; Cort, cortisol; Glcg, glucagon; FFA, free fatty acids; +, -, reactions stimulated, inhibited. Blue indicates actions stimulated by insulin.

how much glucose is released by the liver into the circulation. The liver contains most of the body's glycogen, the glucose polymer that is the main storage form of carbohydrate. Glycogenesis is promoted by glycogen synthase, an enzyme that is stimulated by insulin and inhibited by glucagon. Conversely, glycogenolysis (under the action of phosphorylase) is inhibited by insulin and stimulated rapidly and powerfully by glucagon and the catecholamines. Hepatic stores of glycogen are soon depleted. The liver also secretes glucose produced by gluconeogenesis from lactate, amino acids and glycerol, either formed within the liver or taken up from the circulation. Lactate is a product of glycolysis in muscle. Amino acids, especially alanine and glutamine, derive from proteolysis or amination of pyruvate. Glycerol results from lipolysis. Gluconeogenic enzymes are stimulated by glucagon, catecholamines and cortisol, and inhibited by insulin; glucagon also stimulates amino acid uptake into the liver (Fig. 19.7). After meals, glucose enters the circulation from the gut and its storage is favoured by the high prevailing insulin levels. Insulin stimulates glucose uptake into skeletal muscle and adipose tissue and its subsequent conversion into glycogen and triglyceride, respectively. 1

982

MCQ19.1

Fat metabolism (Fig. 19.8) Circulating triglyceride, either absorbed from the gut or secreted by the liver, is hydrolysed by lipoprotein lipase in the capillary endothelium of adipose tissue and skeletal muscle (Fig. 19.8). This lipolysis yields free (non-esterified) fatty acids (FFA, or NEFA) and glycerol, which are metabolized mainly in adipose tissue. Their recombination into triglyceride (lipogenesis) is stimulated by insulin. Subsequent lipolysis of triglyceride (mediated by triacylglyceride lipase) is inhibited by insulin at low concentrations and stimulated by catecholamines, growth hormone and, in severe insulin deficiency, by glucagon. Glycerol is exported and in the liver is a substrate for gluconeogenesis. FFA are also exported and are used as a fuel by skeletal muscle and many tissues. FFA taken up by the liver can either be used for lipogenesis or undergo oxidation to ketone bodies (ketogenesis}.

Ketogenesis (Fig. 19.9) This takes place within the mitochondria of the hepatocytes, using fatty acyl-coenzyme A (CoA) derivatives formed in the cytoplasm by the reaction of FFA with CoA. These are carried inwards across the mitochondrial membrane by a 'shuttle' enzyme (carnitine palmitoyl transferase) and then undergo B-oxidation to acetyl-CoA and further conversion to acetoacetate. This and its derivatives, 3-hydroxybutyrate and acetone, are the ketone bodies. Ketogenesis is stimulated by glucagon and suppressed by insulin, which respectively enhance and inhibit the activity of the carnitine shuttle, the rate-limiting step of the pathway. Ketogenesis is also regulated by the supply of

19

FIG. 19.10 Major pathways of protein metabolism Ins, insulin; Glcg, glucagon; Cort, cortisol; +, -, reactions stimulated, inhibited (shown in one direction only).

FIG. 19.9 Ketogenesis Ketone body production is inhibited by insulin (Ins), which reduces both the supply of FFA (from lipolysis) and, indirectly, the activity of the shuttle enzyme (carnitine palmitoyl transferase), which carries the precursor fatty acyl-CoA into the mitochondria. Glucagon (Glcg), catecholamines (Cats) and growth hormone (GH) stimulate ketogenesis by opposing these actions.

FFA, and is therefore greatly increased by the combination of insulin deficiency and glucagon excess - as in untreated type 1 diabetes mellitus - which releases the brake on lipolysis. Normally, ketone bodies are exported from the liver and can be used by many tissues as metabolic fuels; indeed, they are the main source of energy in prolonged starvation, when fat reserves are mobilized. In uncontrolled type 1 diabetes, however, their overproduction greatly exceeds their utilization; 3-hydroxybutyrate and acetoacetate, and their accompanying hydrogen ions, accumulate to cause diabetic ketoacidosis. 1 Protein metabolism (Fig. 19.10) Protein synthesis is promoted by insulin and growth hormone and normally balances proteolysis, which is stimulated especially by cortisol. Circulating amino acids, derived either from proteolysis, amination of pyruvate or absorption of digested proteins, are taken up by the tissues. Alanine and glutamine are important substrates for gluconeogenesis in the liver, where their uptake is stimulated by glucagon. Regulation of metabolism in health, starvation and stress Fasting is characterized by low basal insulin levels without rises in counter-regulatory hormones. This permits controlled glycogen breakdown, gluconeogenesis and glucose production by the liver and so maintains blood glucose

levels. Lipolysis is not increased; FFA and ketone body levels remain low. After eating, insulin release increases in step with the influx of glucose, triglycerides and amino acids from the gut, promoting the storage of nutrients not immediately needed. High insulin levels promptly suppress hepatic glucose secretion, and stimulate glucose uptake and storage in skeletal muscle (as glycogen) and adipose tissue (as triglyceride). This prevents postprandial hyperglycaemia, and lipogenesis and protein synthesis are encouraged. In prolonged starvation, insulin levels fall below the threshold necessary to suppress lipolysis, ketogenesis and hepatic glucose production. These processes are enhanced by increased glucagon levels (stimulated by depressed insulin secretion) and by the other counter-regulatory hormones (pp. 980-981), whose secretion is increased by the stress of starvation. The result is controlled catabolism. Blood glucose levels are maintained initially by hepatic glycogenolysis and then, when this is exhausted, by gluconeogenesis, again predominantly in the liver. The body's fat reserves, its major energy store, are then mobilized by lipolysis and ketogenesis. Blood ketone levels increase greatly (the ketosis of starvation) and are used as fuels by most tissues - even the brain - after several days of starvation. Protein metabolism is defended until relatively late, possibly because ketone bodies inhibit proteolysis. Muscle and other proteins are ultimately broken down under the influence of increased cortisol concentrations to support gluconeogenesis. Severe stress (e.g. sepsis, trauma or surgery) greatly increases counter-regulatory hormone secretion and leads to uncontrolled catabolism. Excess gluconeogenesis, lipolysis and proteolysis lead to rapid loss of body fat and muscle, the latter being exacerbated by impaired protein synthesis. The situation is even more precarious in diabetic patients who are insulin deficient, because they are

983

often either untreated or inadequately treated. Profound ketoacidosis and hyperglycaemia can develop with alarming speed. 1

FURTHER READING ON REGULATION OF METABOLISM Kruszynska Y T 1997 Physiology of fuel homeostasis. In: Pickup and Williams eds Textbook of diabetes. Blackwell Science, Oxford.

DIABETES MELLITUS The simple definition of diabetes mellitus as 'a state of chronic hyperglycaemia' belies the immense medical, scientific, social and economic implications of the disease. Because it is common and affects virtually every system in the body, every medical practitioner must be familiar with it. People with diabetes must learn to cope with a lifelong disease which is treatable but not curable, and which carries the threat of complications that may endanger the future quality of life and even life itself. Society as a whole has to shoulder a considerable burden from diabetes. In the UK, diabetes affects about 2% of the population (over 1 million people) and may absorb as much as 10% of healthcare expenditure. Worldwide, there are probably over 150 million people with diabetes, and this is set to rise to over 240 million by the year 2010. Research and advances in clinical care are steadily making the disease easier to live with, at least in richer countries. The problem of diabetes in the developing world is only now receiving the attention it deserves.

The diagnosis of diabetes Blood glucose concentrations are normally remarkably constant: fasting concentrations in whole blood are 3.5-6.7 mmol/L and peak values after meals rarely exceed 8 mmol/L. Glycosuria therefore never occurs unless the renal threshold for glucose reabsorption is below the normal value of 10 mmol/L. The diagnostic criteria for diabetes have recently been revised. The revision has been recommended because it has recently become evident that the fasting glucose concentration used in the 1985 WHO criteria missed many cases of diabetes that would only be diagnosed using the oral glucose tolerance test (Fig. 19.11). The new American Diabetes Association criteria are based on fasting plasma glucose concentration. A value above 7.0mmol/L indicates diabetes, whereas a concentration above 6.0 mmol/L defines a new category known as impaired fasting glucose (IFG).

1

984

MCQ19.2

2 2MCQ19.3 MCQ19.3

FIG. 19.11 Diagnostic criteria for diabetes A fasting plasma glucose concentration >6.0 mmol/L indicates impaired fasting glucose (IFG). After a 75 g glucose load a 2-hour value of >11.0 mmol/L identifies diabetes. A value >7.0mmol/L but <11.1 mmol/L indicates impaired glucose tolerance. The 1-hour intermediate readings on OGTT are not used for diagnosis.

A single very high value (>15 mmol/L) in a symptomatic patient is diagnostic, but otherwise two separate estimations should be made. The values relate to venous whole blood; as red cells contain less glucose, the corresponding plasma values are about 1.1 mmol/L higher. Because glucose is extracted by the tissues, arterial and capillary (e.g. fingerprick) blood levels are higher than those in venous blood. In borderline cases and for research applications, the oral glucose tolerance test (OGTT) is recommended. After an overnight fast, the subject drinks 75 g of glucose (dextrose) dissolved in water; plasma glucose is measured both before and 2 hours afterwards. The normal values and diagnostic criteria for diabetes and impaired glucose tolerance (IGT) are shown in Figure 19.11. A low renal threshold is diagnosed if glycaemia is normal but glycosuria occurs. Urine should be collected during the test if this is suspected. The diagnosis of diabetes carries considerable personal and social disadvantages and must be certain. Only laboratory measurements of blood glucose should be used for diagnostic purposes. Test-strip readings, urinary glucose levels or glycosylated haemoglobin or fructosamine concentrations must never be used. OGTT results vary considerably between and within normal individuals and the test should be repeated if a marginal abnormality is found. Glucose tolerance is particularly sensitive to changes in nutritional state and is worsened in normal subjects by dieting during the preceding days. An adequate diet must therefore be taken for 3 days before an OGTT. 2

Classification of diabetes As more is understood about the aetiology of diabetes it is likely that the classification will change to reflect this. The latest WHO classification (Table 19.1) reflects the two main types of diabetes that are recognized clinically: type 1 (previously known as insulin dependent) and type 2 (previously known as non-insulin dependent) diabetes mellitus, as well as a number of rarer, but specific, aetiological forms of diabetes. Some of the features distinguishing between

TABLE 19.1 New aetiological classification of disorders of hyperglycaemia (WHO 1998) Type 1 diabetes

Autoimmune Idiopathic

Type 2 diabetes

Predominantly due to insulin resistance Predominantly due to insulin secretory defect

Other specific types

Genetic defects of B-cell function Genetic defects of insulin action Diseases of the exocrine pancreas Endocrinopathies Drug- or chemically induced diabetes Infections Uncommon forms of immune-mediated diabetes Other genetic syndromes sometimes associated with diabetes

19

Glucokinase, hepatic nuclear factor-a mutations Insulin receptor mutations Chronic pancreatitis; fibrocalcific pancreatitis Cushing's syndrome, acromegaly, phaeochromocytoma, glucagonoma Steroid treatment, pentamadine Rubella, mumps, cocksackie B4 Insulin antibodies Haemochromatosis, mitochondrial mutations, myotonic dystrophy

Gestational diabetes

TABLE 19.2 Features distinguishing Type 1 from Type 2 diabetes

Metabolic features Insulin deficiency Spontaneous ketosis Need insulin to survive Insulin insensitivity Aetiology Genetic susceptibility HLA markers Autoimmune features Environmental factors

Type 1

Type 2

Severe (C-peptide-negative) Yes Yes Mild, variable

Moderate, variable (C-peptide-positive) No No Severe, variable

Moderate Yes (DRC3/4) Yes ? Viruses ? Cow's milk

Very strong None known No ? Overeating ? Physical inactivity

Clinical features at presentation Age (not very useful) Most <40 years (peak at 12 years) Body weight BMI* mostly <25; recent loss common Microvascular complications Rare

ketosis and will survive (albeit with hyperglycaemia) without insulin therapy. Many patients with type 2 diabetes are, however, treated with insulin, as this is the only available means to lower glycaemia effectively, but these insulin-treated patients can be distinguished from those who require insulin for survival because they have measurable C-peptide levels and do not become ketotic if insulin is withdrawn. It is also now beginning to be recognized that type 1 and type 2 diabetes do not necessarily define a specific pathology. Not all cases of type 1 diabetes have an autoimmune aetiology. Type 2 diabetes may be predominantly due to insulin resistance or to insulin deficiency, or indeed may progress from the former to the latter during the clinical course of the disease. Some examples of the other specific types are indicated in Table 19.1.

TYPE 1 DIABETES MELLITUS Most >40 years (peak at 70 years) BMI mostly <25; recent loss sometimes Sometimes present

*BMI, body mass index = weight (kg)/height (m)2.

the two are shown in Table 19.2. The critical difference is the degree of insulin deficiency, which is so profound in type 1 diabetes that even the low insulin concentrations that normally prevent lipolysis and ketogenesis cannot be sustained. Without insulin replacement, these patients become ketotic and die. By contrast, patients with type 2 diabetes have enough endogenous insulin to prevent

Epidemiology In Britain, type 1 diabetes accounts for about 10% of all diabetic patients and perhaps 30% of those treated with insulin. The frequency of type 1 diabetes shows striking geographical variation: in Finland it is some 35 times commoner than in Japan. Within the UK the highest frequency is found in Scotland, where, as in several other countries, the incidence appears to have doubled within a decade. Incidence shows a pronounced seasonal variation, with most cases presenting in the autumn and winter. These observations all suggest that environmental factors are important in causing or precipitating the disease. Also unexplained is the age predilection of type 1 diabetes. Any age can be affected, but most cases present before 20 years of age and the incidence peaks at 12 years (Fig. 19.12).

985

Environmental agent (? viruses)

Islet Bcell autoantigen exposed favourable configuration Islet Bcell FIG. 19.12 Incidences of type 1 diabetes and type 2 diabetes with age Note that the scale for type 2 is 10 times that for type 1.

DR3/4

Aetiology and pathogenesis

Genetic factors The genetic component of type 1 diabetes is less important than in type 2 diabetes, as is shown by studies of identical twins of whom one becomes diabetic. In the case of type 1 diabetes the chances of the unaffected twin developing the disease are less than 40%, whereas the overall risk for type 2 diabetes may be as high as 90%. Several markers for type 1 diabetes have now been identified, notably specific HLA class 2 antigens, glycoproteins which are encoded by genes that lie on the short arm of chromosome 6 (p. 73). Possession of either HLA-DR3 or DR4 antigens increases the risk of developing diabetes (60% of the population have one or both alleles, but 95% of patients with diabetes). Possession of HLA-DR2 is protective (less than 5% will develop diabetes). These antigens are encoded by the genes HLA-DR A and DRB1, which encode the a and B chains of the HLA-DR molecules, respectively. In the neighbouring region the HLA-DQA1 and HLADQB1 genes are also highly polymorphic, and recent evidence suggests they may also be important in determining disease susceptibility. Both HLA-DR and HLA-DQ molecules are involved in antigen presentation by T cells, and variations in the proteins expressed by these genes may determine the nature of the response if the B cell is damaged, with activation of lymphocytes and continuing B-

1

986

MCQ19.4

unfavourable configuration Islet Bcell

I-helper lymphocyte

Type 1 diabetes results from autoimmune destruction of the islet B cells, which is partly determined by heredity. The genetic markers for diabetes susceptibility are relatively common in Caucasian populations, whereas type 1 diabetes is relatively rare, again indicating that additional factors, presumably environmental, are superimposed.

aberrant expression of class II HLA antigens

Activation of T-lymphocyte

DR2

T-helper lymphocyte

No activation of T-lymphocyte

B-celI destruction FIG. 19.13 Hypothetical scheme of autoimmune B-cell destruction in type 1 diabetes Viruses or other agents could induce islet B cells to express class II HLA antigens and to 'expose' autoantigens (viral antigens could be expressed on B cells, or a normal B-cell antigen could mimic viral antigens). The autoantigen (indicated in blue), when coexpressed with HLA-DR3 and/or 4, has a 'favourable1 configuration, which activates T-helper lymphocytes and triggers autoimmune damage. In contrast, HLA-DR2 has an 'unfavourable' configuration, which prevents this from happening.

cell destruction only occurring if certain configurations of the protein molecules are present (Fig. 19.13). Other diabetes susceptibility genes include the region of the insulin gene on chromosome 11, and at least eight other genes. The ways in which these gene products interact with each other, and with possible environmental factors, to cause autoimmune destruction of the B cells is still not well understood. Autoimmune phenomena An association of type 1 diabetes with organ-specific autoimmune diseases has long been recognized, e.g. with Schmidt's syndrome (adrenal and thyroid failure) in type 2 polyglandular endocrinopathy syndrome (p. 930). Studies in human type 1 diabetes and in rodents that develop spontaneous type 1 diabetes-like syndromes have documented convincing evidence of autoimmune disease. Furthermore, B-cell damage in animals and man, if detected early enough, can be limited by immunosuppressive drugs. Like thyroid, adrenal and other tissues that are the target of autoimmune disease, the pancreatic islets of newly diagnosed patients with type 1 diabetes are infiltrated with immunologically active cells (mostly cytotoxic/suppressorT lymphocytes), an appearance termed 'insulitis'. With the

passage of time the intensity of the insulitis abates, but surviving islets are found to be specifically depleted of (3 cells, implying that these were the victims of the autoimmune attack. B cells may attract immune damage by their 'aberrant' expression of class II HLA antigens (which they do not normally express). The subsequent events thought to occur are shown in Figure 19.13. As well as insulitis, newly diagnosed type 1 diabetic patients show other evidence of ongoing autoimmune processes, with activated T cells and autoantibodies directed against islet cells and insulin, in the circulation. Although most islet-cell antibodies (ICA) react with all islet cell types, some complement-fixing ICA may recognize the B-cell membrane specifically. The crucial B-cell autoantigen is still unknown, although recent work has implicated the enzymes glutamate decarboxylase (GAD), protein tyrosine phosphatase B, and insulin itself. Viruses and environmental factors Environmental factors must account for a large part of the susceptibility to type 1 diabetes. Many viruses target the pancreas and some selectively attack the B cells, where they could persist and alter the cells' activity or antigenicity. Type 1 diabetes presents when seasonal viral infections are common, and some newly diagnosed patients have serological evidence of previous (sometimes recent) encounters with viruses such as mumps, coxsackie B, and retroviruses. Because type 1 diabetes has a long subclinical course, such infections may represent a final insult which tips the patient into critical insulin deficiency. Type 1 diabetes develops in up to 20% of children with congenital rubella, but there is little evidence that rubella is diabetogenic if acquired after birth. There is an apparent

protective effect of breastfeeding on the later development of type 1 diabetes. It has been hypothesized that early exposure to cow's milk may trigger an immune reaction which damages B cells because of 'molecular mimicry' between bovine serum albumin and certain B-cell surface proteins. Various chemical toxins destroy B cells and cause type 1 diabetes-like syndromes in animals (e.g. the cytotoxic agent streptozotocin in the rat) but no toxins have been convincingly implicated in human type 1 diabetes. 1

19

Natural history of type 1 diabetes Long-term prospective studies of subjects at high risk of developing type 1 diabetes (especially siblings of recently diagnosed children) indicate that type 1 diabetes is the culmination of many years of low-level autoimmune damage, which is manifest clinically when a critical mass (>95%) of (3 cells has been destroyed. This is in striking contrast to the acute presentation of type 1 diabetes: symptoms are present in most patients for less than 4 weeks. ICA and insulin autoantibodies appear during the prodromal phase, up to 10 years before type 1 diabetes presents (Fig. 19.14). This is asymptomatic, as is the subsequent prediabetic phase, in which subtle disturbances of insulin secretion (loss of first-phase release) can be identified. Later, the OGTT becomes abnormal and, finally, frank hyperglycaemia supervenes and symptoms appear. Interestingly, the process may abort without leading to type 1 diabetes in a proportion of susceptible subjects.

RECENT ADVANCES IN THE PREDICTION AND PREVENTION OF TYPE 1 DIABETES In families in whom one member has developed diabetes, it is possible to predict with reasonable accuracy whether other family members (e.g. siblings) are at risk of developing the disease. The risk for a sibling is about 1 in 20 and for a monozygotic twin 1 in 3. Accuracy can be improved to over 90% by a combination of HLA-DR phenotyping and measurement of islet-cell antibodies. This ability to predict with reasonable accuracy subjects at risk of developing type 1 diabetes now allows attempts to interfere with the disease processs to be tried in those at very high risk. Trials are under way with various agents, including immunosuppression with agents such as cyclosporine. Less toxic alternatives also under investigation include the vitamin B3 derivative nicotinamide, which may interfere with the immune-mediated (3-cell destruction, and low-dose insulin, which could modify the immune response. It is still too early to tell whether any of these interventions will be effective.

FIG. 19.14 Natural history of type 1 diabetes Islet-cell antibodies (ICA) may be present many years before the onset of clinical diabetes. Intercurrent illness may cause an acute decline in B-cell function and patients may present at this stage. Recovery from illness allows temporary improvement in function, accounting for the 'honeymoon period' before the inevitable development of true insulin dependence. Interventions interfering with immune destruction might therefore be possible before too many B cells have been lost (see Recent Advances box).

987

Biochemical disturbances of type 1 diabetes Insulin deficiency, with the glucagon excess it causes, results in hyperglycaemia due to unrestrained hepatic glucose output, lipolysis and proteolysis. These catabolic processes can be exacerbated by counter-regulatory responses to infection and other stresses. The insulindependent tissues, notably skeletal muscle and fat, 'starve in the midst of plenty', as lack of insulin decreases glucose entry despite its very high extracellular concentrations. An osmotic diuresis, polyuria and dehydration occur when blood glucose levels exceed the renal reabsorption threshold (usually l0mmol/L). Severe insulin deficiency greatly accelerates lipolysis, which fuels ketogenesis (p. 982). Proteolysis causes muscle breakdown and wasting.

Clinical features of type 1 diabetes Presentation Young patients typically have a short history (a few days or weeks) of the classic symptoms of polyuria, thirst, polydipsia, and weight loss. Polyuria is due to an osmotic diuresis caused by high glucose and ketone body concentrations in the urine, and may reach several litres per day. Sleep is disturbed by nocturia, and previously continent children may develop enuresis. Thirst is often intense. Weight is lost rapidly, often despite increased appetite, owing to both dehydration and catabolism of muscle and fat. General tiredness and malaise are common. Other features include blurred vision (due to osmotic disturbances in the lens), muscle cramps and infections, such as boils, urinary tract infections, or candidiasis causing balanitis or pruritis vulvae. Abdominal pain may occur, especially in children. Patients with impending ketoacidosis may have nausea, acidotic (Kussmaul) breathing and other features described on page 1003. Chronic diabetic complications (retinopathy, neuropathy) are exceedingly rare in newly diagnosed type 1 diabetes patients because significant hyperglycaemia is usually only a recent event. Very few cases are detected incidentally in asymptomatic subjects. A full examination should be performed, as it is important to document the presence of any complications and to identify precipitating illnesses. Examination may be unremarkable, and this makes it mandatory to check the blood glucose concentration in any patient who is severely unwell.

SUMMARY 2 Presentation of type 1 diabetes

988

Polyuria, nocturia (enuresis) Thirst, polydipsia Weight loss Tiredness, lack of energy Blurred vision Infections (including candidiasis) Abdominal pain (in children) Features of ketoacidosis

triad of major symptoms

Investigations and diagnosis The diagnosis is confirmed by demonstrating hyperglycaemia (defined on p. 984). The urine must be tested for ketones. Routine blood biochemistry may show features of dehydration and ketoacidosis. Other tests should aim to identify any infection that could have precipitated diabetes (full blood count, urine and blood culture, chest X-ray) and, in older patients especially, any coexisting risk factors for diabetic complications, notably vascular disease (ECG, fasting lipid concentrations).

Management of type 1 diabetes Insulin replacement is mandatory, but other aspects diabetes education, blood glucose monitoring, diet and other lifestyle changes - are essential for the day-to-day and long-term management of type 1 diabetes Insulin treatment Insulin was originally extracted from porcine and bovine pancreases. Insulin of human sequence is now synthesized on an industrial scale by inserting a cloned synthetic gene into either bacteria or yeasts. Biosynthesis mimics the natural process, yielding proinsulin, which can be cleaved in vitro. Human insulins are unlikely to provoke antibody formation. In the UK, many preparations of bovine insulins have been phased out and porcine insulins are becoming less popular; over 90% of patients in the UK and Europe now use human insulin. Human insulin preparations are absorbed slightly more rapidly and have somewhat shorter action profiles than the porcine equivalents, but these differences are unimportant in clinical practice. Human insulins are less antigenic than porcine and especially bovine insulins, and allergic reactions (see below) are very rare. The suggestion that human insulin is associated with decreased awareness of hypoglycaemia is unsubstantiated, although porcine insulin should be prescribed if the patient requests it. Insulin preparations in the UK, North America and most European countries contain a standard concentration of l00U/mL. Insulin preparations Insulin replacement aims to reproduce the normal pattern of insulin secretion, with a background basal supply and additional peaks at mealtimes. Two basic types of insulin, short- and long-acting, are therefore required (see Table 19.3). Short-acting (soluble) insulins (e.g. Actrapid, Humulin S) are simple solutions of unmodified insulin, which appear clear. When injected subcutaneously they have a relatively short action profile, with a peak hypoglycaemic effect after 2-4 hours and a duration of 5-8 hours. Soluble insulins are therefore used to cover major meals, and are injected 30 minutes before eating so that their maximal effect coincides with the postprandial glycaemic peak. Soluble

19

TABLE 19.3 Types of insulin preparation Approximate action profile (hours)* Type

Examples

Onset

Short-acting insulin analogues

Lispro Insulin Insulin Aspart

15-30min

1-2

4-6

Immediately before, with or after meals

Soluble

Actrapid, Humulin S

0.5-1

2-4

5-8

30-40 minutes before eating

Intermediate Isophane (NPH)

Duration

Insulatard Humulin I 6-10

12-24

Monotard, Humulin Zn

'Long-acting' Ultralente

Ultratard (human) 30:70 50:50

Mixtard 30 Humulin M5

Schedule

b.d. with soluble insulin before meals 2-4

Lente (IZS)

Biphasic Soluble: isophane

Peak

or

o.d. in the morning or evening

3-4

12-18

<24f

o.d. or b.d., as for intermediate

0.5-1

2-10

12-18

b.d. or o.d., 30-40 minutes before meals

* Action profiles vary widely between and within individuals. Intermediate and even 'long-acting' insulins based on human insulin often last less than 24 hours, and twice-daily injections are needed in many cases. NPH, neutral protamine Hagedorn; IZS, insulin-zinc suspension; o.d., b.d., once, twice daily.

f

insulins are also used for intravenous infusion (e.g. in treating ketoacidosis) and in insulin pumps. Short-acting insulin analogues (e.g. Lispro Insulin, Insulin Aspart) are a recent development. These insulins bind the insulin receptor and have identical action in vivo to soluble insulin, but their altered amino acid structure prevents them from forming hexameric complexes in solution. This means that they are absorbed much more rapidly after subcutaneous injection, with an onset of hypoglycaemic effect within 30 minutes, and a duration of action of about 4 hours. They can therefore be given immediately before, during or even after meals, which is preferred by some patients. In addition, late postprandial hypoglycaemia is less common in patients using these insulin preparations as part of a 'basal-bolus' regimen, and some studies suggest improved control as well. Prolonged-action insulins are chemically modified to delay absorption. These suspensions must be shaken gently before injection and appear cloudy. There are two main types: • Insulin-zinc suspensions are complexed with excess zinc to produce either large crystals which dissolve and are absorbed very slowly ('ultralente') or amorphous material with a shorter action profile ('semilente'). These are combined in a 70:30 ratio to produce the commonly used 'Lente' insulins (e.g. Humulin Zn and Monotard). • Isophane insulins (e.g. Humulin I, Insulatard) are combined with protamine, and are also known as NPH ('Neutral Protamine Hagedorn').

An ideal long-acting insulin would provide constant basal levels for 24 hours, so that a single daily injection would suffice. In practice, none of the available preparations - including the 'long-acting' human ultralente - reliably lasts a full day, and most severely insulin-deficient (C-peptide negative) type 1 diabetes patients therefore require two daily injections of prolonged-action insulin, Their sluggish onset of action means that these insulins can be injected at any time of day, but for convenience they are usually injected before meals, together with any short acting insulin. Lente insulins contain excess zinc, which combines with soluble insulin and retards its action. If the two are mixed in the same syringe, soluble insulin should be drawn up before lente to avoid contaminating the soluble bottle with zinc, and the mixture must be injected immediately. Insulin Glargine is a new, long-acting insulin preparation that has recently become available. It has an amino acid substitution resulting in slow absorption and has a 24-hour duration of action. Premixed (biphasic) insulins consist of isophane and soluble insulins mixed in 70:30 (e.g. Mixtard 30),50:50 (e.g. Humulin M5) or other proportions. Their action profiles reflect both components. Recent biphasic preparations include faster-acting analogues instead of standard soluble insulin, and so can be injected immediately before eating, Insulin injection regimens The normal pancreas produces about 40 U of insulin each day. A diabetic patient needs less than this if some endoge-

989

nous B-cell function remains (e.g. during the 'honeymoon' period of type 1 diabetes, described below), and more if insulin sensitivity is impaired by increased counterregulatory hormone secretion during intercurrent illness, or by obesity in type 2 diabetes. If little endogenous insulin remains, insulin replacement must provide both the basal and the prandial components of normal insulin secretion. Most C-peptide-negative patients (i.e. those with little residual insulin secretion) ultimately require twice-daily injections of prolongedaction insulin, together with soluble insulin before breakfast and the evening meal. Midday meals (if not large) are usually covered by the morning dose of prolonged-action insulin. About two-thirds of the total daily dose is usually given before breakfast and one-third in the evening, each dose consisting roughly of two-thirds of prolonged-action and one-third soluble insulin. Some patients can be controlled by premixed insulins (two-thirds of the daily requirement before breakfast and one-third before the evening meal), whereas others prefer the flexibility of adjusting the soluble and prolonged-action components separately. 'Intensified' insulin regimens use more injections to try to achieve very tight glycaemic control. These are becoming more widely used in the light of the results of the Diabetes Control and Complications Trial (see below). Soluble insulin is injected 30 minutes before each meal, the dose being determined by the amount eaten and previous postprandial glucose measurements; for convenience, this is often given using a pen injection device (see below). The basal needs are covered by long-acting insulin injected once (before breakfast or at bedtime) or twice daily. Patients with enough residual insulin, for example during the 'honeymoon' period, will be able to fill in gaps in insulin replacement and can achieve good glycaemic control with fewer doses, such as one or two injections of prolonged-action insulin. Adjustment of insulin dosages on the basis of glycaemic monitoring is discussed on page 1000.

990

Insulin injections, pen devices and pumps Insulin injections are almost painless. Disposable plastic syringes are now in general use and can be reused several times, if kept socially clean. Insulin should be injected at an angle into a skinfold raised between finger and thumb, to avoid intramuscular injection, which can hasten absorption unpredictably. The skin does not need to be cleaned. Suitable injection areas are the abdomen (where absorption is fastest and most reliable), upper arm and thigh. Sites should be used in rotation to avoid lipoatrophy or lipohypertrophy. 'Pen' injection devices contain a prefilled vial of soluble or isophane insulin, of which metered doses can be selected and injected by pressing a button. They are easily portable and popular with active and younger patients, especially for intensified injection regimens. Continuous subcutaneous insulin infusion (CSII) is delivered by a miniaturized battery-driven pump. Insulin

in a prefilled cartridge is infused continuously through a butterfly cannula implanted subcutaneously in the abdomen; the infusion site is changed daily. The basal rate can be varied and is often reduced at night, when insulin requirements fall. Prandial boluses, judged according to the size of the meal, are given about 30 minutes before eating, by pressing a button on the pump. Pump failure is rare, but interruption of insulin delivery, e.g. by blockage or disconnection of the cannula, can rapidly precipitate hyperglycaemia and ketoacidosis because the subcutaneous insulin depot is small. This has led to the development of pumps that sound an alarm if insulin delivery is interrupted. Other problems include infection at infusion sites and a transient deterioration in retinopathy, which can accompany any rapid tightening of glycaemic control (p. 1008). CSII is currently seldom used in the UK, but is more widely used in the USA and some European countries. These differences in prescribing practice do not appear to be based on pharmacological or clinical evidence. Carefully selected patients may be treated by insulin infused intraperitoneally, using a pump implanted subcutaneously. Intraperitoneal insulin is absorbed into the portal system and can achieve good control in patients with 'unstable' diabetes that cannot be controlled by less invasive approaches. Intensive insulin treatment can achieve essentially normal blood glucose levels in selected patients, including some who are poorly controlled by other regimens. However, the techniques require great commitment from the patient and the diabetes care team. Problems with insulin treatment Hypoglycaemia Most insulin-treated type 1 diabetes patients suffer occasional hypoglycaemia and, each year, about one in seven will have an episode severe enough to require help. Fear of hypoglycaemia probably prevents many patients from trying to achieve tight control. In the DCCT trial (see below) there was a threefold increase in the incidence of hypoglycaemia in the intensively treated group, compared with those on conventional treatment. Hypoglycaemia is a recorded cause of death in young type 1 diabetes patients, but the chances of a patient dying during a hypoglycaemic episode, even if profound, are very low. The threat of hypoglycaemia can be greatly reduced by education of the patient and his or her family. Causes of insulin-induced hypoglycaemia. Hypoglycaemia occurs when circulating insulin levels are too high and are insufficiently antagonized by counter-regulatory hormones. Contributory factors are: • Too much circulating insulin: wrong dose (errors in drawing up and injecting insulin are surprisingly common) or timing (e.g. injection given too long before next meal); • Not enough food: delayed, missed or inadequate meals; • Exercise: this hastens the absorption of insulin injected

SUMMARY 3 Symptoms and signs of hypoglycaemia Sweating Tremor Tachycardia Pallor Hunger Weakness Visual problems Hemiparesis Aggression, confusion Fitting Coma

sympathetic discharge

neuroglycopenia

into the limbs and increases glucose uptake into muscle; the latter effect continues for several hours after acute strenuous exercise in order to replenish muscle glycogen, and may cause hypoglycaemia the following day; • Alcohol: this is a potent cause of hypoglycaemia. Ethanol directly inhibits gluconeogenesis and glucose output by the liver; • Impaired awareness of hypoglycaemia; • Previous episodes of hypoglycaemia (see below). The widely variable absorption of insulin probably contributes to the many episodes that cannot be explained. Episodes recurring at the same time of day may be due to basic mismatching to the insulin, dietary and exercise regimens. Other possible causes of severe and repeated hypoglycaemia include loss of counter-regulatory hormones (especially in hypothyroidism and Addison's disease), alcohol abuse, the loss of awareness of hypoglycaemia that affects patients with long-standing type 1 diabetes, and factitious hypoglycaemia (see below). Patients with diabetic nephropathy clear insulin more slowly from the circulation and so are prone to hypoglycaemia. Clinical features of hypoglycaemia. Hypoglycaemia triggers sympathetic activity and counter-regulatory hormone secretion. This corrects hypoglycaemia by reducing insulin sensitivity in its target tissues; glucagon and adrenaline also stimulate gluconeogenesis, glycogen breakdown and hepatic glucose production. If this response fails to correct hypoglycaemia, neuroglycopenia may follow. The main symptoms of hypoglycaemia are attributed to either sympathetic stimulation (sweating, tremor, tachycardia and pallor) or neuroglycopenia, whose many features include hunger and weakness; irritability, confusion and aggression; clumsiness; blurred vision and diplopia; aphasia and hemiparesis (Summary 3). Profound neuroglycopenia can cause fitting, coma and finally death. Repeated, severe hypoglycaemia may cause intellectual impairment. Nocturnal hypoglycaemia may not wake the patient but may be suggested by sweating and restlessness (often apparent to the patient's partner), nightmares, and a 'hung-over' feeling with headache the following morning. Hypoglycaemia unawareness. Patients with longstanding type 1 diabetes (>10 years' duration) may lose

SUMMARY 4 Management of hypoglycaemia

19

• Check blood glucose to confirm hypoglycaemia and subsequent recovery • Mild episode (patient able to swallow): Oral carbohydrate (20g) • Severe episode (patient unconscious or uncooperative): Glucagon 1 mg i.m. or s.c. or Glucose 50 ml of 50% solution i.v. (use a large vein) Admit to hospital if recovery is not immediate

their perception of hypoglycaemic symptoms and suddenly develop neuroglycopenia (confusion, fitting, coma) without warning. Most long-standing patients also display blunted glucagon and noradrenaline responses to hypoglycaemia. Recent research has highlighted an important clinical message: a single episode of hypoglycaemia can itself delay the onset and blunt the severity of the autonomic warning symptoms; avoidance of hypoglycaemia can partly restore warning symptoms of hypoglycaemia. Diagnosis of hypoglycaemia. After a few episodes, most patients claim to be able to recognize hypoglycaemia. However, some tolerate very low glycaemia without symptoms. Insulin-treated patients must therefore be educated to check their blood glucose levels immediately if they feel unwell in any way, and their relatives should also be able to do this if the patient becomes confused or unconscious, or behaves atypically. Patients with possible nocturnal hypoglycaemia (see below) should test themselves during the night (especially at 1-\ am). Using test-strips, patients can diagnose hypoglycaemia quickly. Test-strip readings of <2.5 mmol/L on fingerprick samples require treatment. If severe features are present, treatment should be started immediately and levels checked as soon as possible afterwards, as they will remain low for some minutes. Management of hypoglycaemia. Because of the small but significant risks of cerebral damage and death, all episodes must be treated without delay. Most features, including fitting (which is often resistant to anticonvulsants) and coma, are rapidly reversible when glycaemia is normalized. Mild episodes, where the patient is conscious, cooperative and able to swallow safely, are treated with oral carbohydrate: 20-30 g should be taken initially (for example a glass of milk or 6 dextrose tablets), and again if the blood sugar has not started to rise after 10-15 minutes. Severe hypoglycaemia requires intravenous glucose (dextrose): 50mL of a 50% solution should be given over a minute into a large vein; small veins will be painfully thrombosed by the hypertonic solution. The patient may improve immediately, but the intravenous cannula should be left in place until glycaemic and clinical recovery are complete. An alternative is glucagon (Img; 0.5 mg in children), which mobilizes liver glycogen and acts within a few minutes after intravenous injection. Glucagon injected i.m. or s.c. produces recovery within 10-15 minutes. Prefilled syringes are available for emergency use outside hospital,

991

and the technique should be taught to the relatives of hypoglycaemia-prone patients. Patients should always carry easily swallowed carbohydrate, glucose test-strips and documentation that they are receiving insulin (e.g. a clinic or shared-care cooperation card, Medic Alert bracelet or necklace). All episodes should be recorded in the patient's self-monitoring diary and discussed with the diabetes care team. Injection-site problems Minor bleeding or bruising after injection sometimes worry the patient but are unimportant. Lipoatrophy (pitting of the skin due to loss of subcutaneous tissue) is apparently an allergic response to injected insulin and is uncommon with human insulin. Lipohypertrophy is local thickening of subcutaneous tissue, possibly due to the lipogenic and growth-promoting effects of high ambient insulin concentrations, which may form prominent lumps. Both conditions can be unsightly and may impair the absorption of insulin injected into lesions. They are more common at frequently used injection sites and can be avoided by using several sites in rotation. Insulin allergies Allergic reactions to insulin are now rare. Allergy can cause local inflammation, either acute or delayed. Anaphylaxis, mediated by IgE antibodies, is rare. Immune insulin resistance is due to absorbed insulin combining with IgG antibodies, forming large immune complexes which are cleared rapidly from the circulation by the reticuloendothelial system. Daily insulin requirements of several thousands of units have been recorded. Established insulin allergy may be circumvented by changing to an unrelated species of insulin (e.g. human, tunafish), and may respond to immunosuppressive drugs or desensitization with repeated small doses of the allergenic species. Specific problems with insulin treatment Starting insulin treatment If the patient is clinically well, insulin therapy can be started at home by the diabetes specialist nurse. The optimal regimen for an individual depends on the clinical

SUMMARY 5 Side-effects of insulin treatment • Hypoglycaemia • Injection-site lipoatrophy and lipohypertrophy • Allergies (rare with human insulin) local inflammation anaphylaxis immune insulin resistance

992

1

MCQ19.5

3

Figs 19.1-19.3

2

Case 19.1

situation, and discussion with the patient. Although some patients prefer to start with the flexibility of a basal-bolus regimen, good control can often be achieved with a simpler regimen of twice-daily injections, taken as approximately two-thirds of the daily dose in the morning before breakfast and the remaining one-third before the evening meal. Initial doses should be low (e.g. 8-12 U) to avoid severe hypoglycaemia, which can demoralize the newly presenting patient. Except for young children, the elderly and the partially sighted, patients should be encouraged to give their own injections as soon as possible. After starting insulin treatment, many patients with type 1 diabetes enjoy a period of high-quality glycaemic control, achieved relatively easily with low insulin requirements; dosages may fall to a few units per day, and very rare patients can manage without insulin. This 'honeymoon period' is due to the transient recovery of insulin secretion following correction of hyperglycaemia. This ultimately ends because remaining (5 cells are destroyed by continuing autoimmune processes. Both the 'honeymoon period' and its end have to be anticipated and explained to the patient, because failure to reduce insulin dosages can cause hypoglycaemia, and at the end, worsening control and increasing dosages can be dispiriting. Once residual insulin secretion is lost, the average daily insulin requirements for type 1 patients are usually about 0.7-0.8 U/kg/day, but this will vary according to the age, sex, weight and fitness of the patient. Poor glycaemic control overnight Many type 1 diabetes patients have fasting early-morning hyperglycaemia, sometimes preceded by nocturnal hypoglycaemia. The 'dawn phenomenon' is a rise in glycaemia during the 2-4 hours before waking. It is apparently due to delayed insulin insensitivity of the tissues, caused by surges of growth hormone secreted after falling asleep, which are accentuated in type 1 diabetes. Rebound hyperglycaemia may also occur after hypoglycaemia and is attributed to discharge of counter-regulatory hormones, but is probably unimportant in clinical management; a much commoner cause of 'rebound' hyperglycaemia is taking too much carbohydrate to treat a hypoglycaemic episode. Nocturnal hypoglycaemia followed by fasting hyperglycaemia is often due to the time-course of prolonged-action insulin. When injected in the early evening its hypoglycaemic action often peaks at 2-^ am, but fades thereafter. Simply increasing the evening dose of long-acting insulin may worsen nocturnal hypoglycaemia; the insulin does not last longer and so fasting hyperglycaemia may not improve. Both problems can be improved by delaying the evening injection of long-acting insulin until just before bedtime, so that its peak hypoglycaemic effect is postponed and helps to cancel out the hyperglycaemia of the dawn phenomenon. This regimen requires an extra injection, as shortacting insulin is still needed to cover the evening meal. Erratic control and 'brittle' diabetes Few patients enjoy continuously good glycaemic control. This is not surprising, as subcutaneous insulin absorption

is extremely variable, both between and within individuals, and because the short-acting preparations act too slowly and for too long, whereas long-acting insulins do not last long enough (Table 19.3). Patients whose metabolic control is so unstable as to prevent them from living a normal lifestyle (so-called 'brittle' diabetes) are, fortunately, rare. Unpredictable glycaemic swings are sometimes due to intercurrent infection, endocrine disease (e.g. hypopituitarism) and poor education or treatment regimens. When identifiable causes are excluded, most of the remaining patients are young women who are typically obese and apparently have high insulin requirements. They suffer recurrent ketoacidosis and/or hypoglycaemia. The causes of brittleness in these patients are unknown, but some deliberately manipulate their own treatment (e.g. by omitting insulin) to induce poor control. Management can be extremely difficult. They should be admitted to exclude known causes of unstable control and for intensive education and evaluation of control while following a strictly supervised intensive regimen. Continuous subcutaneous insulin infusion (CSII) may be useful in some patients with recurrent hypoglycaemia and/or recurrent ketoacidosis. Sympathetic counselling and psychotherapy Sympathetic psychotherapyofofthe thepatient patient

and her family may also help. 1 2

Insulin resistance

Most type 1 diabetes patients need less than 1 U/kg/day of insulin, but obese patients with type 2 diabetes are insulin insensitive and often need more. Clinical insulin resistance has been arbitrarily defined as a daily requirement of >200 U. Causes include excess counter-regulatory hormones (e.g. in septicaemia and glucocorticoid treatment), immune resistance due to insulin antibodies, and defects of the insulin receptor or postreceptor sites. The latter are either inherited (type A syndromes) or acquired (type B syndromes, due to autoantibodies to the receptor). Both syndromes are associated with acanthosis nigricans. 3 Intercurrent illness The stress of intercurrent illness stimulates the secretion of counter-regulatory hormones. Basal insulin requirements therefore increase and the total daily insulin dosage will tend to rise, even if the patient is unable or unwilling to eat. Failure to increase the insulin dosage therefore causes hyperglycaemia and ketoacidosis. Some patients reduce or even stop their insulin when they fall ill, mainly because they fear hypoglycaemia while not eating normally. This basic mistake, all too often endorsed by doctors, still causes avoidable deaths in the UK each year. Vomiting is particularly worrying, as it may herald ketoacidosis. Patients must be given clear rules stating what to do if they become ill (see Summary 6). Exercise Regular exercise should be part of the diabetic person's normal lifestyle. Exercise increases glucose uptake into muscle, so favouring hypoglycaemia. Insulin-treated patients can help to avoid hypoglycaemia by reducing

SUMMARY 6 'Sick-day rules'

19

• Increase monitoring and test for ketones. • NEVER stop insulin. • Call doctor if vomiting or ketonuria occurs.

TABLE 19.4 Current dietary recommendations for diabetic patients Carbohydrate: 50-55% of total calories

Choose complex carbohydrates (starchy foods) Avoid foods and drinks sweetened with sucrose or glucose Fat: 30-35% of total calories

Choose poly- or monounsaturated fats (vegetable oils) Avoid animal fats, dairy products and cholesterol-rich foods (eggs, fatty meats) Avoid fried foods Protein: 10-15% of total calories

Choose fish or vegetable protein (e.g. pulses)

Dietary fibre Increase intake of vegetables, bran, pulses, etc. Sweeteners Avoid 'diabetic' foods (which contain sorbitol or fructose) Aspartame is an acceptable calorie-free sweetener Salt: <6g/day (<3g/day if hypertensive) This is not a 'diabetic' diet, but a guide to healthy eating suitable for the general population.

insulin dosages before exercise, by checking blood glucose levels before and at intervals during exercise, and by taking carbohydrate (e.g. a small chocolate bar) if the blood glucose is <5 mmol/L. They must be told that delayed hypoglycaemia may occur up to a day after strenuous exertion. Dietary advice for patients with type 1 diabetes Food eaten by type 1 diabetes patients must balance insulin injections and exercise so as to limit both hyper- and hypoglycaemia. Total calorie requirements. These can be calculated from a person's sex, body weight and level of activity. Most patients with type 1 diabetes are not obese and calorie restriction is not needed, although it is in type 2 diabetes. Indeed, some type 1 diabetes patients with marked weight loss may need to increase their energy intake. Dietary composition. Advice for diabetic patients has changed radically in recent years. Current recommendations are essentially those for the general population (Table 19.4). Choosing foods from which the carbohydrate is absorbed more slowly ('low glycaemic index'), such as pulses, rye bread and pasta, may have beneficial effects

993

on postprandial glucose levels. The traditional 'diabetic diet', low in all types of carbohydrate and with a relatively high fat content, may have predisposed to macrovascular disease and is no longer recommended. Timing and size of meals. Meals must be coordinated with insulin injections and exercise. Twice-daily injection regimens fit best with a larger breakfast and evening meal, a small lunch, and snacks at mid-morning and bedtime. At present, judging insulin doses to cover meals is a matter of trial and error rather than a science. Hyperlipidaemic patients should reduce their fat intake, and hypertensive patients their salt. A low-protein diet may slow the progression of diabetic nephropathy. Diabetic education ana1 lifestyle changes The diabetic patient has to live with the disease every day and cope with treatment and self-monitoring procedures that many doctors and nurses do not fully understand. Careful education about diabetes, which generally is best provided by the diabetes specialist nurse, can greatly improve diabetic control and quality of life. Newly diagnosed patients need to learn a great deal, and every patient's understanding of diabetes should be checked and reinforced whenever possible. Some key points are: • Causes and effects of hyper- and hypoglycaemia; • Treatment: doses of insulin or tablets; injection technique; diet; • Monitoring: techniques; targets; how to respond to poor control; • 'Sick-day rules': when and how to call for help (Summary 6), General suggestions for healthy living apply particularly to diabetic patients. Stopping smoking reduces the risks of macrovascular disease and probably of nephropathy and retinopathy. Alcohol, which is both a source of 'invisible' calories and a cause of hypoglycaemia, should be limited to 21 units/week in men and 14 units/week in women, and less in hypertensive patients. Regular exercise should be advised. There are no employment restrictions on patients not receiving insulin and without complications, but because of the risk of hypoglycaemia, insulin treatment is a bar to joining the armed, police or fire-fighting forces and to driving public-service or medium-/heavy-goods vehicles. Insulintreated patients must inform the driving licence authorities and their insurance companies, and may need medical confirmation of fitness to drive. Special life insurance policies are available. Membership of a patient organization such as Diabetes UK (previously the British Diabetic Association (BDA)) may provide valuable support.

994

Pancreas and islet transplantation Segments of whole pancreas can be transplanted, with the exocrine ducts either occluded or drained into the bladder to prevent release of harmful enzymes into the peritoneal cavity. At present, however, pancreas transplants require

RECENT ADVANCES FUTURE TREATMENTS FOR TYPE 1 DIABETES Modified insulins. The use of monomeric short-acting insulin has rapidly gained acceptance in clinical practice, and the race is now on to develop a true 'basal' insulin, with a full 24-hour duration of action. The search is also continuing for alternative modes of insulin delivery, the latest idea being an 'inhaler' that delivers insulin via the lungs. Transplant of encapsulated islets. At present islet and pancreas transplantation require immunosuppression. However, it might be possible to encapsulate islets with a special semipermeable membrane; this would allow the entry of glucose and other insulin secretagogues and the exit of insulin, while preventing immunological attack. An alternative would be to genetically modify the patients' own fibroblasts to produce insulin in response to glucose. immunosuppression and so are generally performed in patients undergoing renal transplantation, who will receive long-term immunosuppressive therapy. The best results indicate that insulin dosage can be reduced, and in some cases withdrawn, for many months and even years. Successful islet transplantation has recently been reported in a few patients, with injection of isolated and purified islets into the liver via the portal vein, using a novel immunosuppressive regimen to prevent rejection, but it remains to be seen whether this approach will be widely adopted.

TYPE 2 DIABETES MELLITUS Ninety percent of British diabetic patients have type 2 diabetes. This is often incorrectly described as 'mild' diabetes, but these patients are nevertheless susceptible to chronic diabetic complications and most die prematurely, mainly from myocardial infarction. Type 2 diabetes is a common disease in well-fed societies, affecting perhaps 14% of Britons and North Americans over 70 years of age. There is a striking geographical and racial variation, which is totally different from that in type 1 diabetes. Type 2 diabetes is several times commoner in Asian immigrants to the UK than in the indigenous British population, and the highest recorded prevalence is in the Pima Indians of Arizona, of whom 50% are affected. Except for MODY (maturity onset diabetes of the young), most patients are over 40 years of age at presentation, and males predominate. There is, however, a worrying trend of type 2 diabetes occurring more commonly in younger patients, owing mainly to the rising prevalence of obesity.

Aetiology and pathogenesis Type 2 diabetes appears to be due to two separate defects - insensitivity of the tissues to insulin and defective insulin

Insulin resistance is associated with a constellation of metabolic abnormalities and conditions, including hypertension, obesity, dyslipidaemia, hyperuricaemia and diabetes (syndrome X). The mechanisms underlying these associations are poorly understood; possible mechanisms suggested by experimental data in animals and humans include increased activity of the sympathetic nervous system, induction of insulin resistance by the fat-derived cytokine tumour necrosis factor-a, and the fat-derived hormone leptin.

FIG. 19.15 Pathogenesis of type 2 diabetes mellitus Type 2 diabetes is associated with insulin resistance, obesity and defective insulin secretion. Various genetic and environmental factors may be involved. Hyperglycaemia may exacerbate the impaired insulin action and secretion (glucose toxicity). Eventually, impaired glucose tolerance (IGT) and, later, overt diabetes develop.

secretion - both of which may be partly inherited (Fig. 19.15). Genetic factors The concordance rate for type 2 diabetes in identical twins exceeds 90%. Inheritance appears to be polygenic, except in MODY (maturity onset diabetes of the young) and certain other families, where type 2 diabetes is transmitted as an autosomal dominant. About 20% of patients with MODY have a defect in the enzyme glucokinase, resulting in alteration of the 'set point' of insulin secretion in response to glucose. Two other genes for MODY have been identified, namely, those encoding transcription factors called hepatic nuclear factors (HNF) la and 4oc; this indicates that MODY is a heterogeneous condition. Despite intense research, no genetic markers have been found for the more common forms of type 2 diabetes. Insulin resistance This signifies the inability of insulin at physiological concentrations to exert its normal metabolic actions. Investigation shows that insulin's action is reduced by about 40% in type 2 diabetes. Liver, skeletal muscle and fat are all insensitive, so that increased hepatic glucose output and reduced peripheral uptake both contribute to hyperglycaemia (Fig. 19.15). The defect(s) causing insensitivity to insulin are unknown and apparently lie beyond its receptor. Possible contributory factors include obesity (which is strongly associated with type 2 diabetes), an inherited predisposition, and increased glucagon levels; hyperglycaemia itself also reduces insulin sensitivity ('glucose toxicity').

19

Defective insulin secretion Resting insulin levels, measured by conventional insulin radioimmunoassay, may be normal, high or low in patients with type 2 diabetes. Some of this 'insulin' immunoreactivity is now known to represent biologically inert cleavage products of proinsulin, which cross-react in the assay. True insulin levels (measured by a highly specific assay) are subnormal in many type 2 diabetes subjects. Furthermore, non-diabetic people exposed to hyperglycaemia would respond with far higher insulin levels than occur in type 2 diabetes, confirming that patients with type 2 diabetes are relatively insulin deficient. In addition, the spontaneous pulses and the first phase of insulin secretion are absent in type 2 diabetes. The cause is unknown. (3-cell numbers are reduced, but there are no signs of autoimmune damage. The islets become infiltrated with amyloid deposits containing a peptide named amylin, or islet amyloid polypeptide. Amylin is synthesized by the |3 cells, but its possible aetiological role in type 2 diabetes remains uncertain. Hyperglycaemia per se may also impair insulin secretion ('glucose toxicity'; Fig. 19.15). Recent epidemiological research has suggested a relationship between low birthweight and the prevalence of type 2 diabetes. It is possible that factors involved in intrauterine nutrition may either impair p-cell reserve or result in insulin resistance, but the precise mechanisms are not known. Biochemical abnormalities There is hyperglycaemia, but enough insulin secretion remains to prevent lipolysis. Spontaneous ketosis and ketoacidosis therefore do not occur, even in extreme hyperglycaemia (hyperosmolar, non-ketotic coma). Clinical features of type 2 diabetes Two-thirds of type 2 diabetes patients in the UK are obese. At diagnosis, hyperglycaemic symptoms (polyuria, thirst, polydipsia) have usually been present for months or even years. An acute presentation and sudden marked weight loss are unusual but may occur, especially with intercurrent illness, the introduction of diabetogenic drugs or carcinoma of the pancreas, which is associated with type 2 diabetes in older people. Women often have troublesome pruritus vulvae due to candidiasis. Type 2 diabetes often has a long subclinical course and many asymptomatic patients are diagnosed incidentally when screened routinely for glycosuria or hyperglycaemia.

995

RECENT ADVANCES THE UKPDS STUDY The UKPDS study was the largest study ever conducted into type 2 diabetes, including nearly 4000 patients studied for an average of 10 years. It showed that good control of blood glucose (HbAlc of 7%, compared with 7.9%) was effective at reducing the risk of microvascular complications of diabetes (25% risk reduction), and that using insulin or sulphonylureas as primary therapy did not affect outcome. A cohort of obese subjects were treated with metformin, and both microvascular and macrovascular disease were reduced in this group. The study also highlighted the progressive nature of type 2 diabetes, and the problems of weight gain and hypoglycaemia with sulphonylureas and insulin, but not with metformin. A blood pressure control study was later embedded into the UKPDS study design. This showed that tight control of blood pressure (mean 144/82 vs 154/87) with either captopril or atenolol significantly reduced both micro- and macrovascular complications of diabetes. The effects of blood glucose and blood pressure control are additive, and the benefits occur at all levels of HbAlc.

Others present with chronic diabetic complications, such as myocardial infarction, retinopathy (especially maculopathy), or foot ulceration. It has recently become clear that type 2 diabetes is a progressive disease, with early insulin resistance, often associated with obesity and physical inactivity, moving on to impaired glucose tolerance, through to predominantly insulin-resistant type 2 diabetes. There is a progressive failure of (3-cell function, and ultimately a requirement for insulin in many patients. It is important to remember that the condition is very heterogenous and progression can occur at different rates; some patients may never require insulin, whereas others (particularly lean patients) need insulin early in the course of the disease. There is evidence that some patients with apparently typical type 2 diabetes (either lean or obese) have islet-cell autoantibodies and other markers of autoimmune diabetes and may in fact have a late-onset, slowly evolving form of type 1 diabetes. Maturity onset diabetes of the young (MODY) By definition, MODY presents in patients under 25 years of age, who do not need insulin for at least 5 years. MODY is transmitted as an autosomal dominant trait. Hyperglycaemia may be mild, particularly in glucokinase-deficient MODY, and some affected families appear to be relatively protected from chronic diabetic complications; however, MODY should not be considered a benign disease, as severe complications may develop.

1

996

MCQ19.6

FIG. 19.16 Flow-chart for treating type 2 diabetes Glycaemic targets are shown in Table 19.5.

Type 2 diabetes is diagnosed and distinguished from type 1 diabetes as described on page 984 and in Table 19.2.

Management of type 2 diabetes Type 2 diabetes is often harder to treat effectively than type 1 diabetes. Diet is usually described as the 'cornerstone' of management, but unfortunately few patients are able to follow dietary advice in the long term, and the progressive nature of the disease often supervenes; in the recently reported United Kingdom Prospective Diabetes Study (UKPDS) only 5% of patients were able to maintain good control on diet alone for more than a few months. Oral hypoglycaemic agents, including metformin, sulphonylureas, acarbose, and perhaps thiazolidenediones, may be effective initially, but there is a progressive need to increase the dose and add-in a second or even a third agent. Eventually, insulin treatment is often needed, but this introduces its own problems. A flow-chart for treating type 2 diabetes is shown in Figure 19.16.

The importance of glycaemic control The UKPDS study has now demonstrated that there is a clear relationship between long-term metabolic control

and the microvascular complications of type 2 diabetes. Despite the limitations of many treatments, early and effective use of lifestyle measures, oral hypoglycaemic drugs and insulin is likely to improve long-term outcome. Careful attention should also be paid to other risk factors, particularly hypertension. Diet and general lifestyle changes All type 2 diabetes patients should eat a 'healthy' diet (Table 19.4) and obese patients will also need to reduce their total intake of calories, especially of fat. This often causes an acute fall in blood glucose levels, with symptomatic improvement; continued weight loss will produce a sustained reduction, and even normalization, of glycaemia. An acceptable rate of weight loss is about 0.5-1 kg per week, which corresponds to an energy deficit of at least 500 kcal/day. Weight loss, especially if combined with exercise, increases insulin sensitivity and also improves hyperlipidaemia and hypertension. Timing and distribution of meals are not critical in patients who are not taking insulin or sulphonylureas, as hypoglycaemia does not occur. A sympathetic dietitian, possibly reinforced by a slimming club, stands the best chance of success. Very lowcalorie diets (<700kcal/day), which are very expensive, and weight-reducing drugs such as orlistat or sibutramine may be helpful in some cases. Weight-reducing diets must be given an adequate trial of 1-3 months, but less than 20% of type 2 diabetes patients will achieve either their target weight or acceptable glycaemic control (Table 19.5). The rest, given the gloomy label of 'dietary failures', need oral hypoglycaemic agents or insulin. 1

It is essential to attempt to diminish the risk factors for cardiovascular disease, which is the main cause of pre-

TABLE 19.5 Targets for diabetic control Ideal Blood glucose (mmol/L) Fasting Postprandial

4.4-6.7 4.4-8.9

Glycosylated haemoglobin (%) **(HbA1c) (with normal range of 3.5-5.5%)

<7%

Other targets Few hypoglycaemic episodes Body mass index <25* Normal plasma lipid concentrations Normal growth (in children) No diabetic complications Normal lifespan

ultimate aims

* Body mass index = weight (kg)/height (m)2. ** Depends on local reference range.

Acceptable

<7.8 <10.0

7-8%

mature death in type 2 diabetes (see Fig. 19.29). General measures include stopping smoking, moderating alcohol intake, taking regular exercise and controlling hyperlipidaemia and hypertension. Monitoring of type 2 diabetes is generally assessed from HbA1c supplemented by the patient's own once- or twice-weekly fasting glucose levels, although insulin-treated patients will need to test more frequently (p. 1000).

19

Oral hypoglycaemic agents The major oral hypoglycaemic agents (Table 19.6) are the sulphonylureas, metformin and acarbose. Thiazolidenediones (e.g. rosiglitazone, pioglitazone) have recently been introduced have been found to be effective when combined with either sulphonylureas or metformin. Other drugs with minor hypoglycaemic effects include soluble vegetable fibres such as guar, which interfere with intestinal glucose absorption. Sulphonylureas These are derivatives of sulphonamides. 'First-generation' drugs include chlorpropamide and tolbutamide, and the newer 'second-generation' compounds include glibenclamide, gliclazide and glipizide (Table 19.6). Secondgeneration sulphonylureas are more potent, but seem to have few other specific advantages. Mechanism of action and pharmacology. Sulphonylureas act primarily as insulin secretagogues (see Fig. 19.4). They restore the missing first phase of insulin release and enhance basal and prandial insulin secretion. As sulphonylureas depend on releasing endogenous insulin they are ineffective in insulin-deficient type 1 diabetes patients. Sulphonylureas may exert an additional 'extrapancreatic' action by slightly enhancing insulin sensitivity. Overall, sulphonylureas reduce fasting and postprandial glycaemia by about 25% in moderately hyperglycaemic patients, but usually have little effect if fasting glycaemia exceeds 14 mmol/L. They are ineffective at the outset in 5% of type 2 diabetes patients, and subsequently fail in many of those who respond initially; poor compliance with treatment or progressive deterioration ('exhaustion') of (3 cells may be responsible. Most sulphonylureas are cleared through the kidneys, although glipizide is metabolized in the liver. Chlorpropamide has a very long plasma half-life (36 hours) and glibenclamide has prolonged biological effects which outlast its survival in the circulation. These long-acting drugs are taken once daily. Neither should be used in patients with renal impairment. The others, especially the very short-acting tolbutamide and glipizide, are usually taken before breakfast and the evening meal. Newer insulin secretagogues Two new drugs, glimepiride and repaglinide, have recently become available. These are chemically and pharmacologically distinct from the sulphonylureas, although their action in inducing insulin secretion through stimulation of the glucose-potassium ATPase on the (3 cell is similar, and they have comparable efficacy and side-effects, apart from

997

TABLE 19,6 Oral hypoglycaemic agents Drug

Dosage schedule

Comments

Chlorpropamide

250-500 mgo.d.

Tolbutamide

500 mgb.d.,t.d.s.

Hypoglycaemia +++ (long half-life) Avoid in elderly and in renal failure Alcohol-induced flushing Acceptable in mild renal impairment

Glibenclamide

2.5-1 5 mg o.d.

Gliclazide Glipizide

40-1 60 mgo.d., b.d. 2.5-15 mg o.d., b.d.

Biguanide

Metformin

500-850 mg b.d. , t.d.s.

Contraindicated in renal, hepatic, cardiac or respiratory failures Does not cause hypoglycaemia

a-Glucosidase inhibitor

Acarbose

50-100mg t.d.s.

Start at low dose (Gl side-effects) Does not cause hypoglycaemia Not absorbed systemically

Rosiglitazone

4-8 mg o.d.

Pioglitazone

15-30 mgo.d.

Can be used in combination with sulphonylureas or metformin Contraindicated with insulin; liver enzymes should be monitored Can be used in renal impairment

First-generation sulphonylureas

Second-generation sulphonylureas

Thiazolidinediones

a lower incidence of hypoglycaemia. Glimepiride is long-acting and can therefore be given as a single daily dose. In contrast, repaglinide is very short-acting, with a rapid onset of action, and may be particularly useful at controlling postprandial hyperglycaemia. Adverse effects of sulphonylureas • Hypoglycaemia is the main hazard, especially with the long-acting drugs in renal failure (when they accumulate) and in older people. Chlorpropamide and glibenclamide must therefore be avoided in patients over 60 years of age, and in renal failure. The short-acting agents are relatively safer in mild renal impairment but patients with nephropathy should be transferred to insulin when serum creatinine concentration rises above 250 |imol/L. • Weight gain further reduces insulin sensitivity and damages morale. It is probably due to the lipogenic effects of increased insulin levels, perhaps with relaxed dietary compliance in some cases. • Allergic skin rashes are unusual. Alcohol-induced facial flushing is an idiosyncratic reaction in patients taking Chlorpropamide. Rarer side-effects include water retention and hyponatraemia owing to potentiation of ADH action and blood dyscrasias. Their hypoglycaemic effects are potentiated by drugs such as the sulphonamides, cotrimoxazole, fibrates, salicylates and probenecid, which compete for binding to plasma proteins and reduce the clearance of sulphonylureas. 0MCQ19.7

998

2

Case 19.2

Hypoglycaemia +++ (long biological action) Avoid in elderly or in renal failure Acceptable in mild renal impairment Acceptable in mild renal impairment

Clinical use of sulphonylureas. They are indicated in type 2 diabetes if hyperglycaemia persists after a trial of dietary treatment (usually 3 months). They are used particularly in non-obese patients. Obese subjects may benefit more from metformin, which does not cause weight gain. In general, sulphonylureas will reduce HbAic by about 1 %. Sulphonylureas are Contraindicated in type 1 diabetes (C-peptide-negative) patients, or in type 2 diabetes patients during intercurrent illness or surgery, because insulin is required. Metformin Metformin (Table 19.6) reduces hyperglycaemia by stimulating glucose uptake into peripheral tissues and by inhibiting gluconeogenesis in the liver. It does not stimulate insulin secretion and so does not cause hypoglycaemia in normoglycaemic subjects. Metformin reduces glycaemia by a similar degree to the sulphonylureas, but is less likely to be effective if fasting glycaemia exceeds 14mmol/L. Adverse effects of metformin • Lactic acidosis. Blocking lactate utilzation through gluconeogenesis inevitably causes lactate accumulation. Lactic acidosis occurs rarely with metformin, except in the 'organ failures', which either increase lactate production or prevent its utilization (p. 1007). • Gastrointestinal symptoms - notably bloating, nausea, dyspepsia and diarrhoea - are common; anorexia may help to prevent weight gain during metformin treatment. Metformin impairs vitamin Eu absorption and occasionally causes a megaloblastic anaemia.

Clinical use of metformin. Metformin is indicated in patients who fail to respond to diet, either alone or together with a sulphonylurea, especially if obese. It is contraindicated in renal, hepatic, cardiac and respiratory failures because the risks of lactic acidosis are greatly increased. It is also contraindicated in insulin-deficient patients, although some recent studies have suggested that combination therapy with metformin may be useful in some obese type 2 diabetic patients who require insulin. This may be done conveniently by giving night-time NPH with metformin continued at conventional doses during the day. There is good evidence that lower insulin doses are needed with this regimen, which is associated with less weight gain. Acarbose Acarbose is an oc-glucosidase inhibitor and acts at the intestinal brush border by preventing breakdown of polysaccharides. Its main effect is therefore to reduce postprandial hyperglycaemia, and reductions of HbA1c have been reported in clinical trials. The main side-effects are gastrointestinal distubance with borborygmi and diarrhoea, caused by fermentation of unabsorbed sugars in the large intestine. These can be limited by gradual introduction of the drug over several weeks. It may be used in patients who have failed to respond to diet, but is usually reserved for those who are intolerant of, or have poor responses to, the other agents.

RECENT ADVANCES FUTURE TREATMENTS FOR TYPE 2 DIABETES Although the fundamental defects causing type 2 diabetes remain unknown, several agents that have beneficial effects on both insulin sensitivity and insulin secretion are under development. Insulin sensitizers. The thiazolidinediones have recently been introduced into clinical practice, but a range of compounds with similar actions are being investigated. These include agents that are agonists at the peroxisome proliferator-activated receptor (PPAR) a and y, combining the beneficial effects of fibrates (PPAR-a agonists) with the thiazolidenedione effects. Incretin hormones. Glucagon-like peptide 1 (7-36) amide (p. 979) is a powerful stimulus to insulin secretion and also delays gastric emptying. Preliminary studies suggest it may normalize fasting and mealrelated hyperglycaemia in type 2 diabetes, and the development of orally acting analogues is awaited. Anti-obesity drugs. Obesity is a major factor in the development of type 2 diabetes, and several agents that may be of benefit are being investigated. These include centrally acting appetite suppressants such as sibutramine, pVadrenergic agonists (which increase metabolic rate), and orlistat, which impairs digestion of fat by blocking the actions of intestinal lipases.

Thiazolidinediones This is a new class of drugs that are agonists at the peroxisome proliferator-activated receptor (PPAR)-y. They act as insulin sensitizers, and are effective at reducing HbAic by about 1 % when used as monotherapy, or in combination with sulphonylureas, metformin or insulin. Currently only combinations with sulphonylureas or metformin are licensed in Europe. Side-effects include weight gain and a small degree of fluid retention. Deranged liver function and, very rarely, fulminant hepatic failure have been reported with troglitazone, but not with other drugs in this class. Hypoglycemia does not occur, although the action of insulin or sulphonylureas may be potentiated. 1

19

Insulin treatment in type 2 diabetes Diet combined with one or both types of oral agent will only achieve satisfactory long-term control (Table 19.5) in 50% of patients with fasting glycaemia of <10mmol/L, and in few with higher glucose levels. Insulin is needed in patients with unacceptable hyperglycaemia or features of insulin deficiency (weight loss and spontaneous ketonuria). Many type 2 diabetes patients will be satisfactorily controlled by a single daily injection of ultralente or isophane insulin, as their remaining insulin secretion can cover mealtime requirements, but those with worsening insulin deficiency (low C-peptide levels) may ultimately need multiple injections, as for type 1 diabetes. Insulin treatment often causes weight gain, which further aggravates insulin insensitivity and increases insulin requirements, but the risks of hyperglycaemia probably outweigh those of obesity. Combination of insulin with metformin may be helpful in some cases. ©

IMPAIRED GLUCOSE TOLERANCE (IGT) This diagnostic category defines subjects whose blood glucose concentration lies between normality and diabetes, i.e. between 7.8 and 11 .Ommol/L 2 hours after a 75 g OGTT (p. 984 and Fig. 19.11). IGT is asymptomatic but must be distinguished from normal because long-term follow-up studies indicate that these subjects carry a considerably increased risk of cardiovascular disease (not microvascular disease, which only develops with truly diabetic glucose levels; see p. 1006). Many are obese and display the hyperinsulinaemia, insulin insensitivity, hypertension and hyperlipidaemia that strongly predict cardiovascular disease (see p. 1015). After 10 years most IGT subjects will remain in this category, but 25% will revert to normal and 25% progress to type 2 diabetes. IGT has no specific management, but diabetogenic drugs (e.g. thiazides) should be withdrawn if possible and the patient strongly encouraged to lose weight (which tends to normalize glucose tolerance), stop smoking and reduce other cardiovascular risk factors. Patients and their blood glucose levels should be reviewed annually, and the OGTT repeated if circumstances change.

999

SECONDARY DIABETES MELLITUS Malnutrition-related diabetes mellitus (MRDM). Malnutrition is associated with pancreatic damage (particularly stone formation) and diabetes, although it is not clear whether this represents a distinct disease entity. MRDM is reported from many developing countries, but even there seems to account for a minority of cases of diabetes. It has been subdivided into: • Fibrocalculous pancreatic diabetes with chronic pancreatitis, causing recurrent abdominal pain, large pancreatic calculi (easily visible on plain radiography or ultrasound scanning) and steatorrhoea; and the less clearly denned • Protein-deficient pancreatic diabetes (PDPD) Both affect young adults, who usually need insulin, sometimes in large doses in PDPD, but ketosis is rare. Pancreatic disease. Diabetes may be caused by chronic pancreatitis, usually due to ethanol abuse. Haemochromatosis causes diabetes, cirrhosis and, occasionally, bronzing of the skin owing to iron deposition in the islets, hepatocytes and dermis. Carcinoma of the pancreas is associated with type 2 diabetes in older people. Endocrine diseases causing diabetes are acromegaly (60% of patients have IGT or diabetes), Cushing's syndrome (including iatrogenic causes), phaeochromocytoma and the very rare glucagonoma. These conditions produce excessive counter-regulatory hormones, whereas Conn's syndrome may cause IGT and sometimes diabetes through potassium depletion, which impairs insulin secretion. Gestational diabetes is described on page 1003. Diabetogenic drugs include the glucocorticoids, pY adrenergic agents (e.g. salbutamol or ritodrine) given intravenously, thiazide diuretics (which cause potassium depletion) and diazoxide (which blocks insulin release directly). Other rare conditions associated with diabetes include the DIDMOAD syndrome (diabetes insipidus, diabetes mellitus, optic atrophy and deafness), myotonic dystrophy and haemochromatosis; see pp. 1404 and 868. Lipoatrophic diabetes (with partial or total loss of subcutaneous fat) and 'leprechaunism' are due to insulin receptor defects. Mitochondrial gene defects may result in diabetes with a maternal pattern of inheritance.

ASSESSING DIABETIC CONTROL Diabetic treatment aims to keep glycaemia close to normal, to avoid the acute and long-term complications. Various measures for immediate and medium-term glycaemic control are available (Table 19.5). 1 1000

MCQ19.8

Blood glucose monitoring The development of test-strips suitable for self-monitoring of blood glucose levels has revolutionized diabetes management. Most test-strips (e.g. BM sticks and Glucostix) contain glucose oxidase and a dye that reacts with the hydrogen peroxide generated by the oxidation of glucose in the blood sample. The colour developed can be read either visually against a colour chart, or electronically by a reflectance meter. Newer 'sticks' measure glucose oxidation electrically. Blood is usually obtained from the sides of the pulp of the finger and not the tip (which is a sensory organ), using various lancet devices. A generous drop is lowered carefully on to the reagent area, which must be covered completely, left in contact for exactly the reaction period stipulated, and then wiped off. The colour is allowed to develop for the correct time before reading. Inaccurate readings (sometimes dangerously misleading) are common in patients' own records and on ward charts. All doctors and nurses and all insulin-treated patients must be able to measure blood glucose with test-strips; the equipment, and the patient's technique and ability to interpret the results should be checked regularly. For type 1 diabetes patients taking twice-daily insulin or on intensified regimens the key time-points for testing are before breakfast (when levels are determined by the previous evening's long-acting dose), before lunch and before the evening meal (determined by the morning's dose), and before bed (influenced by the evening's dose). Generally, readings can be made at different time-points on different days, so that a pattern is built up. Patients should test whenever they feel unwell, and during the night if nocturnal hypoglycaemia is suspected (p. 992). Prandial doses of soluble insulin can be adjusted from the glucose level 90-120 minutes after eating. In type 2 diabetes patients treated with oral agents and diet, occasional fasting values are a useful index of control.

Urinalysis for glucose and ketones Urinalysis for glucose (using glucose oxidase test-strips) is inconvenient and usually unhelpful, as it reflects the average glycaemia above the renal threshold (which can lie anywhere between 7 and 13 mmol/L) since the bladder was last emptied. A crucial failing is that normo- and hypoglycaemia cannot be distinguished. Urinary glucose measurements are acceptable only in rare patients who cannot be persuaded to prick their fingers, and in type 2 diabetes patients with a normal renal threshold who are not receiving insulin and who are not subject to hypoglycaemia. Urinary ketone measurements provide useful warning of ketoacidosis and can be useful during intercurrent illness.

Glycosylated haemoglobin and fructosamine Valine residues on the (3 chain of the adult haemoglobin (HbA) molecule can react with glucose, and the result-

ing glycosylated haemoglobin (designated HbA t ) can be separated from native HbA and measured by electrophoresis. HbA! can be further resolved into subcomponents, of which HbAlc is the most stable and clinically useful. The proportion of HbA that is glycosylated depends on its total exposure to glucose and is proportional to the average glycaemia during the half-life of the red blood cell, i.e. the preceding 6-8 weeks. Non-diabetic HbA lc values are about 3.5-5.5% of total HbA; well and poorly controlled diabetic values are about 6-7% and >8%, respectively. HbAic measurements are a useful index of medium-term glycaemic control, but are invalidated if red-cell turnover is disturbed or if abnormally migrating Hb variants are present. One problem with HbAlc measurement is the wide variety of techniques available for measuring it, with the effect that results are rarely directly comparable between laboratories. Fortunately, the two major trials of diabetes control and complications, the DCCT and the UKPDS, used the same technique, and many laboratories are now moving over to report 'DCCT-aligned' HbAlc results. Serum albumin is also glycosylated and can be measured by the 'fructosamine' reaction. Albumin turns over faster than HbAlc and indicates mean glycaemia during the previous 1-2 weeks. Current assays are cheaper, but less reliable and reproducible, than those for HbAlc.1

ORGANIZATION AND DELIVERY OF DIABETES CARE The development of the diabetes care team, consisting primarily of doctors, specialist diabetes nurses, dietitian and chiropodist, has greatly improved efficiency by allowing effective division of labour. The diabetes specialist nurse is especially important, for educating patients about diabetes and for starting and adjusting treatment. Other specialists, e.g. ophthalmologist, vascular surgeon and obstetrician, should be readily available for referral, and combined clinics, such as for diabetic foot problems, are particularly useful. 'Shared-care' schemes, which involve both the general practice and the hospital team in looking after the patient, are becoming popular, as are community 'miniclinics'. However care is organized, all diabetic patients must be thoroughly reviewed annually, primarily to screen for complications (Table 19.7), and must have easy access to expert help.

INTERCURRENT EVENTS AND DIABETES Infection It is widely held that diabetic patients are generally more susceptible to infection than the non-diabetic population. Diabetic patients are more likely to suffer infections with certain atypical or unusual organisms. These include

TABLE 19.7 Annual review of the diabetic patient

19

Treatment and self-monitoring Check and adjust as necessary Coexistent risk factors and diseases Strongly discourage smoking Examination Body weight and BMI*: - determine target weight Cardiovascular: - blood pressure, lying and standing; signs of heart failure; peripheral pulses (presence, strength, bruits) Peripheral nerves: - examine limbs for neuropathy Feet: - examine carefully Eyes: - visual acuity (before mydriatics, ± glasses or pinhole), fundoscopy or retinal photography through dilated pupils; check for cataracts Investigations HbA1c

Plasma urea, creatinine and electrolytes Fasting plasma lipid concentrations Urinary albumin excretion ECG if >40 years of age *Body mass index; see Table 19.5.

tuberculosis, combined staphylococcal/streptococcal infections that result in a rapidly spreading cellulitis (progressive synergistic gangrene or necrotizing fasciitis), mucormycosis (p. 335), and gas-forming organisms that infect the soft tissues of the feet or urinary tract.

Surgery Surgery presents several potential hazards to diabetic patients. First, the trauma of surgery provokes a counterregulatory stress response, which causes catabolism and, especially in insulin-deficient patients, can rapidly lead to hyperglycaemia and ketoacidosis. Second, insulin and the sulphonylureas can cause severe hypoglycaemia in fasted or anorexic patients, and this is particularly dangerous during general anaesthesia. Finally, poorly controlled diabetes accelerates catabolism and delays postoperative wound healing. Diabetic control must therefore be meticulous. The risks of operating on diabetic subjects can be greatly reduced by agreeing a routine management policy between the diabetes care team, surgeons, anaesthetists and ward staff. Fitness for surgery should be carefully assessed in the light of cardiovascular or other complications. Patients should be admitted some days before surgery to stabilize their treatment if necessary, and blood glucose levels must be monitored regularly throughout the perioperative period. Basic treatment guidelines are as follows: • Patients with type 2 diabetes who are well controlled by diet or oral agents and undergoing minor surgery

1001

TABLE 19.8 Glucose-potassium-insulin ('GKI') regimen for surgery in diabetic patients Indication All diabetic patients except for those with type 2 treated with oral agents or diet alone and undergoing minor surgery only Infusion 15 U soluble insulin + "lOmmoL KCI in 500ml of 10% dextrose Delivery Infuse at 100 mL/hr Check blood glucose level hourly: - if <5mmol/L replace bag with one containing 10 U insulin + KCI - if >10mmol/L replace bag with one containing 20 U insulin + KCI

only. Change long-acting sulphonylureas (chlorpropamide, glibenclamide) to short-acting drugs (e.g. tolbutamide, gliclazide) some days before surgery to reduce the risk of hypoglycaemia; omit oral agents and breakfast on the morning of operation; monitor glycaemia carefully and treat persistent hyperglycaemia with the intravenous glucose-potassium-insulin regimen described below. • All other diabetic patients. Stop regular insulin on morning of surgery, give a continuous intravenous infusion of balanced amounts of glucose, potassium and insulin ('GKI'), which will both maintain satisfactory glycaemic control (5-10mmol/L) and prevent hypokalaemia (Table 19.8). This regimen should be started on the morning of surgery and continued until the patient is able to eat and drink normally, when the usual treatment can be resumed. Alternatively, insulin may be given as a variable rate infusion, which provides greater flexibility.

Pregnancy Pregnancy in diabetes poses problems for both mother and fetus. Two separate aspects need to be considered, namely pregnancy in women with established diabetes, and 'gestational' diabetes, which is precipitated by pregnancy in a previously non-diabetic woman. Pregnancy in established diabetes Diabetes can cause difficulties at all stages of pregnancy; fortunately, most can be avoided by careful diabetic management. Conception Menstrual disturbance, anovulatory cycles and subfertility are common in poorly controlled diabetic women.

1

1002

MCQ19.9

2

MCQ19.10

Risks to the mother during pregnancy Pregnancy worsens hyperglycaemia, especially during the second and third trimesters, apparently because the hormones of pregnancy (e.g. placental lactogen) are diabetogenic. Pregnant diabetic women are at increased risk of hypertension and pre-eclampsia, and both retinopathy and nephropathy may deteriorate rapidly, possibly secondary to improved glycaemic control (pp. 1008-1014). Delivery of overweight (macrosomic) babies also carries risks for the mother. Maternal mortality for diabetic women was previously high but is now comparable to that in the general population. Risks to the fetus Uncontrolled diabetes affects the fetus in several ways. Congenital malformations, especially cardiovascular and skeletal (e.g. agenesis of the sacrum), are three times commoner than in non-diabetic pregnancies and occur in up to 7% of cases. Diabetes is particularly teratogenic during the first 10 weeks of fetal life, when the major organs are being formed. The cause is not known. Macrosomia (increased body size and obesity) occurs because glucose crosses the placenta and high levels stimulate hyperplasia of the fetal P cells and insulin secretion; the resultant hyperinsulinaemia induces growth and lipogenesis in the fetus. Gross macrosomia can cause dystocia and so threaten both mother and fetus. Risks to the neonate Babies of diabetic mothers are liable to hypoglycaemia in the neonatal period (because their hyperplastic P cells continue to secrete excessive insulin) and the respiratory distress syndrome (hyperglycaemia decreases surfactant synthesis and lung maturation). Perinatal mortality has fallen but remains higher than in non-diabetic pregnancies. Diabetic women who intend to become pregnant require specific counselling, and strict diabetic control should be achieved before conception to ensure nearnormoglycaemia during the critical first few weeks of organogenesis. Patients must be encouraged to stop smoking and drinking alcohol, and other medical problems, such as hypertension or retinopathy, must be assessed and treated.

SUMMARY 7 Diabetes and pregnancy Effects on the mother Insulin resistance and increased requirements Microvascular complications may deteriorate Risk of dystocia with macrosomic babies

Effects on the fetus Congenital malformations Macrosomia

Effects on the neonate Birth trauma from macrosomia Hypoglycaemia Respiratory distress syndrome

Very tight glycaemic targets should be set (Table 19.5) and achieved if necessary by intensive insulin treatment with multiple daily injections or CSII. The few women taking oral agents should be transferred to insulin. Fetal growth must be carefully monitored. Ultrasound scanning can identify macrosomia, manifested as an excessive increase in abdominal girth, and major malformations. Fetal lung maturation can be difficult to assess, because diabetes spuriously increases the commonly used lecithin: sphingomyelin ratio in amniotic fluid. Macrosomia can be avoided by careful diabetic control, and full-term delivery (vaginal if possible) is preferred to reduce the hazards of prematurity, especially the respiratory distress syndrome. During labour, diabetes should be controlled by intravenous glucose-potassium-insulin infusion or a variable insulin infusion (see above and Table 19.8). Insulin requirements rise rapidly if dexamethasone is given to encourage fetal lung maturation or ritodrine to delay delivery, and dosages fall dramatically after the placenta is delivered. The neonate needs careful monitoring for hypoglycaemia (especially if macrosomic) and respiratory distress. The mother's usual diabetic treatment can be resumed soon after delivery, but oral agents must be avoided during breastfeeding as they enter the milk. 1 Gestational diabetes This occurs in 1-2% of pregnancies. Diabetes develops during pregnancy, most commonly during the second and third trimesters; it usually resolves after delivery, but is likely to recur in subsequent pregnancies. Up to 70% of women who have had gestational diabetes subsequently develop type 2 diabetes. The risk factors for developing gestational diabetes are similar to those for type 2 diabetes, namely family history, age, obesity and ethnicity. It seems likely that the increased demands on the pancreas during pregnancy (insulin requirements increase threefold) unmask latent type 2 diabetes. Although gestational diabetes is a type 2 diabetes-like syndrome, in that ketoacidosis does not develop, patients frequently require insulin to achieve adequate glycaemic control. The mother is usually asymptomatic, but there is a significant risk to the fetus because of macrosomia (causing dystocia) and neonatal hypoglycaemia if the condition is untreated. Screening policies based on age, family history, previous macrosomia, neonatal hypoglycaemia or unexplained fetal loss will miss many cases of gestational diabetes. All pregnant women should be screened for gestational diabetes between 24 and 28 weeks' gestation. Many screening tests have been advocated, but the most practical is probably the short OGTT. A 75 g glucose load is given, followed by a plasma glucose measurement 1 hour later. Fasting is not required. A plasma glucose >7.8mmol/L should be followed by a full glucose tolerance test. A fasting glucose >5.5mmol/L or a 2-hour glucose >9.0mmol/L probably warrants treatment. A team approach to treatment is essential, with the dia-

betes specialist nurse and dietitian working alongside the physician and obstetrician. Patients should be taught blood glucose monitoring (preferably using a meter) and asked to test blood both fasting and 1 hour after meals. Targets are <5.5mmol/L fasting and <7mmol/L postprandially. Unless hyperglycaemia is severe, diet should be tried initially, but insulin is often required. This is usually given as a short-acting preparation to control postprandial glucose levels, with intermediate-acting insulin at night if there is fasting hyperglycaemia. Insulin can be stopped after delivery, but all women who have had gestational diabetes should have a repeat glucose tolerance test postpartum to exclude persistent diabetes, and be advised about the increased risk of diabetes in the future. ©

19

DIABETIC METABOLIC EMERGENCIES Coma in diabetic patients is commonly due to the disease's metabolic complications or its treatment, but other causes (Table 19.9) are very important and must be considered.

Diabetic ketoacidosis Diabetic ketoacidosis (DKA) is severe uncontrolled diabetes with hyperglycaemia and metabolic acidosis due to high circulating ketone-body levels; acetoacetate and 3-hydroxybutyrate are organic acids. Ketone-body overproduction is due to unrestrained lipolysis, which only occurs with severe insulin deficiency, or under conditions in which a massive counter-regulatory response of stress hormones (e.g. severe infection, or following trauma) overwhelms residual insulin secretion. DKA is therefore the classic hallmark of type 1 diabetes, in which it occurs spontaneously, but it occasionally develops following intercurrent illness in type 2 patients, some of whom can subsequently be treated without insulin. Before insulin, DKA killed most patients with type 1 diabetes. It remains an important problem and the commonest cause of death in such patients under 20 years of age. Overall mortality approaches 10%, and exceeds 20% in older patients with severe intercurrent illness. Many deaths are due to avoidable delays and mistakes in diag-

TABLE 19.9 Causes of coma in a diabetic person Diabetic ketoacidosis Hyperosmoiar, non-ketotic coma Hypoglycaemia Lactic acidosis Others: Strokes (commoner in diabetes) Postictal (hypoglycaemia can cause fitting) Drug overdose (including insulin) Alcohol (can cause hypoglycaemia)

1003

nosis or treatment. The blood glucose level must always be checked in anyone who is ill. Pathophysiology DKA is due to insulin deficiency, compounded by excesses of glucagon and other counter-regulatory hormones. Insulin deficiency occurs in newly presenting or untreated type 1 diabetes patients, and in established patients who omit or reduce their insulin, or fail to increase it during intercurrent illness (see p. 1001). Counter-regulatory hormones are increased by the stress of intercurrent illness, such as infection (sometimes apparently trivial), trauma, surgery, burns or myocardial infarction. About 40% of episodes are unexplained, but patients often fail to take enough insulin during illness. The result is runaway catabolism in all three branches of metabolism (see Figs 19.7-19.9). • Unsuppressed lipolysis in fat yields FFA, which fuel ketogenesis, and glycerol, which, together with alanine from proteolysis, feeds hepatic gluconeogenesis. • When hyperglycaemia exceeds the renal reabsorption threshold, glucose remains in the urine and, together with high ketone-body levels, causes an intense osmotic diuresis which carries sodium and potassium with it. This causes dehydration, hypovolaemia and renal underperfusion, which reduces the clearance of glucose through the kidneys (an important mechanism for limiting hyperglycaemia) and may lead to acute tubular necrosis (Fig. 19.17). • Electrolyte and water depletion are exacerbated by vomiting, a direct central action of hyperketonaemia. DKA typically causes the loss of several litres of water and several hundred millimoles each (perhaps a week's normal intake) of sodium, potassium and chloride. • Acidosis develops when acetoacetate and 3hydroxybutyrate levels exceed the body's buffering capacity. Intracellular acidosis interferes with vital enzymes; death usually results from cardiac arrest. Clinical features The features of DKA are usually superimposed on those of untreated type 1 diabetes. Hyperventilation, driven by acidosis, may be mistaken for cardiac or respiratory dyspnoea. Acetone may be smelled on the breath. Nausea and vomiting are ominous, as water and electrolyte loss can become rapidly worse. Hypothermia, due to peripheral vasodilatation, may mask pyrexia of infection. Drowsiness and coma occur late. Investigations and diagnosis Once suspected, the diagnosis can be confirmed instantly by measuring blood glucose and urinary ketone levels with test-strips. Once diagnosed, treatment must begin immediately. Initial investigations aim to: 1004

• Assess the severity of DKA. Venous blood is taken for glucose, urea and electrolytes; arterial blood for pH and

A

Intercurrent illness

Dehydration

Insulin deficiency

Counter-regulatory hormone excesses

Muscle Proteolysis

Liver

Fat Lipolysis

Ketogenesis

FFA

Alanine

Gluconeogenesis •

Glycerol

tGLUCOSE

Glycosuria

B

Osmotic diuresis Na,K,CI depletion 4. renal losses of glucose

Vomiting

Dehydration Hypovolaemia Renal underperfusion

FIG. 19.17 Diabetic ketoacidosis Biochemical basis A and pathophysiological disturbances

B

gases; and urine is tested for ketones. Typical values are shown in Table 19.10. Blood ketone levels are measured by some laboratories and will exceed 5 mmol/L (normal range <1 mmol/L). Plasma osmolality can be measured formally or calculated approximately (in mmol/kg) as twice the sum of plasma sodium + potassium, plus the sum of urea + glucose (all in mmol/L). The 'anion gap' (plasma [Na + K] - [Cl + HCO3]) will be increased above normal (> 17 mmol/L) by the presence of acetoacetate and 3-hydroxybutyrate. Identify the cause of DKA. Tests should include a full blood count (note that DKA can cause a neutrophil leukocytosis, without infection), chest X-ray, blood and urine cultures, and ECG and cardiac enzymes in older patients.

19

TABLE 19.10 Investigations in diabetic ketoacidosis Test

Normal range

Findings in ketoacidosis

Plasma glucose Plasma Na Plasma K Plasma urea Venous HC03 Plasma osmolality

4-8mmol/L 137-1 43 mmol/L 3.5-5.0 mrnol/L 3.5-7.0 mmol/L 18-28 mmol/L 288-298 mmol/kg

>1 7 mmol/L (may be lower, especially in children) >155mmo!/L indicates hyperosmolality. May be depressed by very high glucose levels Normal, high or low (total body K is always depleted) High in dehydration and renal failure Reduced (arterial blood pH and base excess are more useful) >350 mmol/kg indicates severe hyperosmolality

Arterial blood pH Base deficit

7.35-7.45 -2 to +2 mmol/L

<7.0 indicates severe acidosis <-10 indicates severe acidosis

Urinary ketones

Negative

++ or +++ (+ or ++ may occur in starvation without ketoacidosis)

Management of DKA The first priority is to correct fluid and electrolyte depletion, as this is the immediate threat to life. Insulin replacement, to correct hyperglycaemia and restore glucose supply to the tissues, and treatment of any intercurrent illness are also essential. Fluid replacement A large peripheral or central venous cannula is needed for intravenous fluids. Most patients recover rapidly with 1-2 L of saline given in 2 hours, then 2 L in 4 hours, then 4 L in 24 hours. Fluid losses (in urine or vomit) should be added to these volumes. Replacement should be faster in shocked or oliguric patients (who may require plasma expanders), and slower in the elderly or those with cardiac disease or signs of fluid overload, cerebral oedema or respiratory distress. Fluid balance must be monitored continuously from urine output (catheterize if necessary), clinical signs, or measurement of central venous or pulmonary wedge pressures. As sodium, chloride and potassium are depleted, the logical replacement fluid is saline with potassium chloride. Isotonic (0.9%) saline should be used until plasma glucose has fallen to 10-44 mmol/L, when 5% dextrose can be substituted. This is given with insulin, so the glucose is transported into cells and replenishes their stores. In severe hyperosmolality, half-isotonic (0.45%) saline may be given instead, as in hyperosmolar, non-ketotic coma (p. ••). Small amounts of sodium bicarbonate may be used in profound acidosis (see below). Total body potassium is always depleted in DKA, but acidosis causes potassium to leak out of cells, so that plasma levels can be low, normal or high. Plasma potassium often falls rapidly during treatment, as insulin carries it into cells together with glucose. Although potassium replacement is always needed, it must not be given during the period of hyperkalaemia because it may precipitate cardiac arrhythmias. Potassium chloride should be added to each litre of intravenous fluid as follows:

• 20mmol if plasma potassium is normal (3.5-5.0mmol/L); • 40mmol in hypokalaemia (<3.5mmol/L); • none in hyperkalaemia (>5.0mmol/L). Profound hypokalaemia may require additional potassium, which is best given by infusion pump. Patients with severe potassium disturbances should be monitored electrocardiographically. Insulin replacement Soluble insulin given intravenously acts most rapidly and reliably in DKA; intramuscular and particularly subcutaneous absorption are too slow and erratic. Insulin is best given by an infusion pump (50 U soluble insulin diluted in 50mL saline, i.e. lU/mL), or as 50U in a 500mL bag of saline (O.lU/mL). It should be infused initially at 6U/h, i.e. 6mL/h for the pump and 60mL/h for the drip, which should be regulated by a drip-counter or paediatric giving-set burette. Blood glucose must be measured hourly, and when it starts to fall the infusion rate should be titrated to achieve a glucose level of 10-15 mmol/L after 6-8 hours. Faster rates of fall are unnecessary and may cause hypoglycaemia. Most patients need 1-3 U/h. Other measures and problems Frequent monitoring of glucose, electrolytes and clinical state is crucial. Any coexistent illness should be treated. Broad-spectrum antibiotics are usually given prospectively. Myocardial infarction with DKA has a very high mortality. Other problems include: • Shock, which may cause acute prerenal and renal failure. This usually responds to adequate saline replacement, but plasma expanders and inotropes may be needed for severe hypotension (systolic BP < 90mmHg). • Severe acidosis. The use of bicarbonate in DKA is controversial, as it causes hypokalaemia and, although raising blood pH, may paradoxically worsen intracellular acidosis. Small volumes (250-500 mL) of isotonic sodium bicarbonate (1.4%) may be given if arterial pH is 6.9 or less. • Cerebral oedema. Subclinical brain swelling is probably common in DKA and may be exacerbated by water, ion

1005

SUMMARY 8 Management of diabetic ketoacidosis Fluid replacement 2L in 2 hours, then 2L in 4 hours, then 4 L in 24 hours (plus losses) Isotonic saline initially 5% dextrose when blood glucose falls to 10-14mmol/L Potassium replacement Replace according to plasma levels Add 0, 20 or 40mmol to each litre of i.v. fluid if plasma potassium is >5.0, normal or <3.5mmol/L, respectively Insulin replacement Continuous i.v. infusion of soluble insulin at 6U/h until blood glucose falls, then 1-3 U/h according to blood glucose Monitoring Blood glucose (hourly) Fluid balance and plasma electrolytes Clinical condition Identify and treat intercurrent precipitating illness Broad-spectrum antibiotics Identify and treat complications of ketoacidosis Severe acidosis (pH < 6.9) - consider i.v. bicarbonate and ventilation Shock - consider i.v. plasma expanders and inotropes Cerebral oedema - i.v. dexamethasone, mannitol Adult respiratory distress syndrome - ventilate Afterwards Stabilize insulin treatment Education to prevent further episodes

or glucose shifts in the brain during rehydration. It particularly affects children, causing a decline in consciousness, sometimes with papilloedema. It carries a high mortality, but dexamethasone or mannitol given intravenously before brain herniation occurs sometimes helps. Fluid overload should be avoided; a major reduction in the incidence of cerebral oedema is one of the successes of using lower rates of fluid repletion than were traditional. Adult respiratory distress syndrome (ARDS). Alveolar fluid accumulation, perhaps due to ion and water shifts, clinically resembles pulmonary oedema, but with a normal heart size on chest X-ray and low pulmonary wedge pressure. It is usually fatal. Treatment includes ventilation with high-concentration oxygen and correction of fluid overload. Acute gastric dilatation, with vomiting, a succussion splash and a ground-glass appearance on abdominal Xray, requires nasogastric drainage to prevent aspiration, especially if consciousness is reduced.

1

1006

MCQ 19.11

2 MCQ 19.12

• Coma requires standard nursing care, ideally in an intensive care unit. • Hypothermia indicates a poor prognosis. It may respond to rewarming with a space blanket. After the emergency is over Intravenous fluids and insulin should be continued until the patient is able to eat and drink, and ketonuria has diminished. Standard subcutaneous insulin can then be resumed or started in newly diagnosed patients using prolonged-acting insulin. Most patients who have had DKA require insulin for life, as the vast majority suffer from type 1 diabetes; rare patients with type 2 diabetes who develop DKA during severe intercurrent illness may return to oral hypoglycaemic agents. The cause of the episode must be identified if possible and discussed with the patient to prevent it happening again. 1

Hyperosmolar, non-ketotic coma This is distinguished from DKA by the absence of gross hyperketonaemia and metabolic acidosis. Hyperglycaemia is generally greater than in DKA and, together with a rise in urea due to dehydration and prerenal failure, may elevate the plasma osmolality to >350 mmol/kg. Hyperketonaemia does not develop because these patients have enough insulin secretion to suppress lipolysis and ketogenesis, but not enough to prevent the liver from producing glucose. These patients are therefore C-peptide positive with type 2 diabetes, which is often previously undiagnosed. It is commoner in people of AfroCaribbean origin. Precipitating factors include myocardial infarction, stroke, infection, and diabetogenic drugs such as glucocorticoids and thiazide diuretics. Presentation is with severe hyperglycaemic symptoms (polyuria, intense thirst, weight loss, blurred vision), without the features of ketoacidosis. Confusion, drowsiness and coma are commoner than in DKA. Complications include stroke and thromboembolic disease (such as mesenteric artery thrombosis, deep venous thrombosis and pulmonary embolism) owing to increased blood viscosity. Mortality exceeds 30% because these patients are old and often have a serious precipitating illness. The biochemical features are: • Hyperglycaemia: often >50mmol/L, sometimes >100mmol/L; • Hypernatraemia: often >155 mmol/L (may be depressed by high glucose levels); • Uraemia: due to dehydration, with or without renal failure; • Hyperosmolality: >350 mmol/kg; • Blood and urine ketone levels: normal or slightly raised (in starvation); • Arterial pH, venous bicarbonate and anion gap: normal. Management is largely as for DKA. Saline replacement, using half-isotonic (0.45%) solution if plasma sodium exceeds 150 mmol/L or osmolality >350 mmol/kg, must be

CASE STUDY 19.1 DIABETIC KETOACIDOSIS A 23-year-old heavy-goods vehicle driver was taken to the accident and emergency department by his girlfriend. He had been becoming progressively unwell over the previous 2 weeks, with thirst, polyuria, blurred vision and weight loss of about 6kg. On the day of admission he had vomited on two occasions, and although he was intermittently confused she managed to persuade him to seek medical help. He had no relevant past medical history and was taking no medication. On examination he was drowsy, but rousable. He had deep, sighing respiration and smelt strongly of ketones; he was clinically dehydrated and had a pulse of 100 beats per min and blood pressure 100/60 mmHg. The only other abnormality was a gastric succussion splash.

Questions 1. What investigations should be carried out? 2. Describe the initial management.

Discussion Initial investigations confirmed the suspected diagnosis of diabetic ketoacidosis. The plasma glucose was 34 mmol/L and he had glycosuria and +++ketonuria. Serum sodium was 132 mmol/L, potassium 6.3 mmol/L, urea 13.4mmol/L, creatinine 140(imol/L, arterial pH 7.1, Po2 13.5 kPa, Pco2 2.3 kPa, base excess -16. He had a moderate leukocytosis; chest X-ray and ECG were normal. These abnormalities are typical of diabetic ketoacidosis; note that the potassium is raised, although the patient will have a potassium deficit of several hundred mmol. This will fall as acidosis is corrected, and potassium supplementation will be needed soon after treatment has begun. A leukocytosis is common in DKA and does not necessarily indicate infection but, if this is suspected, a careful screen, including urinalysis, CXR and blood cultures, is indicated. An ECG may show changes due to hyperkalaemia. In older subjects with known diabetes a myocardial infarction may present

cautious in older patients, in whom cardiac disease is common. Potassium levels must be carefully monitored and replaced as needed. Most patients respond rapidly to intravenous insulin infusion at low doses. Prophylactic low-dose heparin (5000 U s.c. 8-hourly) should be given, but full anticoagulation should be reserved for proven thromboembolism as the risks of fatal gastrointestinal bleeding are high. Despite often impressive hyperglycaemia at presentation, these patients have type 2 diabetes and many later do not need insulin. Drugs and other precipitating factors must be avoided subsequently if possible. 2 Lactic acidosis Lactate is generated by glycolysis in muscle (Fig. 19.7). Lactate levels rise rapidly during tissue anoxia (e.g. shock, cardiac failure, pneumonia) and when the liver and kidney are prevented from utilizing it (e.g. by hepatic impairment or biguanides). Lactic acidosis is rare with metformin (and even with phenformin, which carries an intrinsically higher risk) as long as other predisposing factors (major organ

with ketoacidosis, but without chest pain. The patient requires intravenous fluids and insulin. Initially he should be given 1L of 0.9% saline rapidly, and an insulin infusion started at 6 units per hour, and adjusted according to bedside measurement of capillary blood glucose. Frequent monitoring of glucose, electrolytes and acid-base status is essential. The fluid deficit (predicted to be 5-6 L) should be gradually replaced over 24 hours. He has been vomiting, has a reduced conscious level, and has a succussion splash indicating gastric paresis, and so requires insertion of a nasogastric tube to reduce the risk of aspiration. Once the plasma glucose falls to 12mmol/L or below, a 10% dextrose infusion (with potassium if necessary) should be substituted for saline. At this stage he will soon be able to eat and drink, and subcutaneous insulin can be started, followed by education in diabetes and its management.

failure) are avoided. It presents as coma with metabolic acidosis (reduced arterial pH and venous bicarbonate) and a wide anion gap (p. 1004) due to lactate. Blood glucose levels are usually within the diabetic range. Treatment often fails to prevent death. Intravenous sodium bicarbonate may paradoxically aggravate intracellular acidosis (see above), which may be limited by vigorous ventilation to blow off carbon dioxide. Haemodialysis may both clear lactate hydrogen ions and remove sodium overload due to bicarbonate administration.

CHRONIC COMPLICATIONS OF DIABETES The chronic tissue complications of diabetes are the single greatest anxiety for most diabetic people. However, over 40% of type 1 diabetes patients survive for over 40 years, half of them without developing significant complications, and there is already encouraging evidence - notably from the results of the Diabetes Control and Complications Trial and the United Kingdom Prospective Diabetes Study and

1007

the falling incidence of nephropathy during the last decade • that continuing advances in diabetic care will reduce the damage caused by diabetic complications. Chronic diabetic complications are usually classified as: • Microvascular (microangiopathic) disease. This is an essential component of retinopathy and nephropathy, and is characteristic of and specific to diabetes. The role in neuropathy is less certain. Common features are basement membrane thickening and abnormal leakiness of the microcirculation. The lesions affect particularly type 1 diabetes patients who developed the disease before puberty but can be devastating in patients with type 2 diabetes. Indeed, significant complications are often found in type 2 patients at diagnosis, reflecting the long unsuspected presence of the disease. • Macrovascular (large-vessel) disease. This is atherosclerosis, distinguished from that affecting non-diabetic people only by its rapid and extensive development.

Microvascular complications Pathogenesis Microvascular lesions in type 1 diabetes and type 2 diabetes are identical, suggesting that hyperglycaemia or some other metabolic disturbance of diabetes is responsible. Epidemiological evidence indicates that microvascular complications are commoner in poorly controlled and rarer in well-controlled diabetic patients. The long-term DCCT (Diabetic Control and Complications Trial) in the USA has now confirmed that tight glycaemic control can more than halve the long-term incidence of retinopathy, nephropathy and neuropathy (Fig. 19.18) in type 1 diabetes. Similar findings were found in the UKPDS study for patients with type 2 diabetes, and, importantly, control of blood pressure (with a target of <150/85) reduced the risk of both microvascular and macrovascular complications of diabetes. Suggested mechanisms of microvascular disease include: • Overactivity of the polyol pathway. This alternative pathway converts glucose and other sugars to their respective polyols or sugar alcohols (e.g. glucose to sorbitol, galactose to galactitol) through the action of aldose reductase. This enzyme is found in the retina, lens, glomerulus and Schwann cell of nerves, which are all damaged in long-standing diabetes. Glucose flux through the polyol pathway increases dramatically in hyperglycaemia, and this may cause accumulation of sorbitol in these tissues. How these changes might cause tissue damage is unknown. • Glycation of proteins. Hyperglycaemia increases nonenzymatic glycation of many proteins, which may impair their structural and metabolic functions.

1

1008

Figs 19.4-19.6

2

Fig. 19.7

FIG. 19.18 Diabetes Control and Complications Trial Diabetic retinopathy is reduced with intensive (mean HbA1c 7%) vs conventional treatment (mean HBA1c 9%) of type 1 diabetes in the DCCT trial. Similar reductions were seen in the development of microalbuminuria, albuminuria and clinical neuropathy, but not of macrovascular disease. These considerable benefits were offset by a threefold increase in the incidence of severe hypoglycaemia and by excessive weight gain.

• Haemodynamic and coagulation disorders. Resting blood flow is increased in the retinal, glomerular and other microcirculations and could cause microvascular damage. Smoking, hypertension and a tendency to enhanced coagulability of the blood may also contribute. • Other factors. As 60-70% of type 1 diabetes patients never develop nephropathy or proliferative retinopathy, an inherited susceptibility may be important.

Ocular complications Diabetic eye disease, especially retinopathy, is the commonest cause of blindness in the British population of working age, and for most patients is the most frightening complication of the disease. Fortunately, it is easily detected and now often treatable. Diabetic retinopathy Retinopathy is a readily visible sign of widespread microvascular damage. Background retinopathy (see below) affects 30% of patients with type 2 diabetes at diagnosis and virtually all patients after 20 years of either type 1 or type 2 diabetes; these changes therefore seem to result from hyperglycaemia or associated metabolic disturbances. In contrast, proliferative changes only ever develop in 30-50% of diabetic people, who may constitute a specific susceptible subset. Early abnormalities include capillary dilatation with increased retinal blood flow, thickening of the capillary basement membrane, loss of the surrounding pericytes and increased leakiness of the capillaries. Focal occlusion of capillaries and later arterioles, possibly due to microthrom-

bus formation, ultimately causes retinal ischaemia and infarction. The progression of retinopathy is apparently accelerated by poor glycaemic control, smoking and hypertension. Retinopathy may deteriorate acutely during pregnancy and after a sudden improvement in previously poor glycaemic control, probably because this reduces retinal blood flow and aggravates ischaemia in critically perfused areas of the retina. Stages and lesions of retinopathy The stages and lesions of retinopathy are detailed in Figure 19.19. Background retinopathy. Capillary dilatation and leakage can only be demonstrated by fluorescein angiography (Fig. 19.20). The first lesions visible on fundoscopy are red dots, which are either microaneurysms (localized dilatations of weakened capillaries) or haemorrhages from ruptured capillaries. Haemorrhages deep in the retina appear as small dots or larger 'blots', whereas superficial ones fan out along nerve fibre bundles into a 'flame' shape. 'Hard' exudates are lipid-rich precipitates of protein extravasated from leaky capillaries. These have a shiny, waxy appearance and are often clustered in arcs and rings around the macula. Background lesions alone do not impair vision, unless extensive exudates cause maculopathy, and in most cases do not progress to new vessel formation. 1 Preproliferative changes. These indicate worsening ischaemia, and 50% of cases develop new vessels 2 within 12 months. Multiple cotton-wool spots ('soft' exudates) are fluffy, greyish-white patches of axoplasm released from retinal infarcts; they also occur in accelerated hypertension. Venous abnormalities include beading and looping. Intraretinal microvascular abnormalities (IRMA) are clusters of dilated capillaries (not new vessels). Vision is not affected by preproliferative changes alone. Proliferative retinopathy. New vessels usually sprout from veins, especially where they bifurcate or enter the optic disc. They advance in leashes or arcades within the retina or into the vitreous, and may extend anteriorly to appear on the iris as rubeosis iridis. New vessels threaten vision because they are fragile and rupture easily, causing haemorrhages within the retina, in the vitreous, or in the potential preretinal (subhyaloid) space between the two. Vitreous haemorrhages appear as greyish opacities obscuring the fundus, whereas preretinal bleeds have a flat top if the red cells in the space are allowed to sediment. Extensive vitreous and preretinal haemorrhages cause sudden loss of vision in the affected eye. Small haemorrhages may resorb almost entirely, but bleeding usually recurs and ultimately impairs vision. Without treatment, 50% of patients with 'high-risk' new vessels (especially those involving the disc, which are particularly liable to bleed) will become blind within 5 years (see Figs 24.6 and 24.7, p. 1293). Advanced diabetic eye disease. Fibrous tissue proliferates with new vessels and following repeated haemor-

19

FIG. 19.19 Stages and lesions of diabetic retinopathy Maculopathy, an important cause of blindness in type 2 diabetes, can accompany any of these stages; macular oedema may be invisible on fundoscopy.

FIG. 19.20 Fluorescein angiogram Severe background retinopathy showing microaneurysms, haemorrhages, and venous beading indicating early preproliferative changes. Diffuse fluorescence outside the vessels indicates extensive capillary leakage. (Photography by courtesy of Mr R. Whitelocke.)

rhages, appearing as opaque, greyish membranes on fundoscopy. It tends to contract, so that strands attached to the retina exert traction, causing retinal detachment and tears. Detachment causes visual loss, which is permanent if

1009

the retina is not reattached within a few weeks. Warning symptoms include distortion of vision and a 'sparkling' sensation, and these therefore need urgent ophthalmological referral. Fibrosis also complicates rubeosis iridis, especially if the vessels bleed into the anterior chamber, and may block the filtration angle. This prevents reabsorption of aqueous and causes glaucoma, with a blind and extremely painful eye. Maculopathy. The macula serves high-resolution central vision, which is essential for daily activities such as reading. Unfortunately, it is particularly sensitive to damage by oedema and exudates from surrounding leaky capillaries and ischaemia. 1 Maculopathy can complicate any stage of retinopathy and is the commonest cause of blindness in type 2 diabetes. Significant maculopathy dramatically impairs vision and must be suspected if visual acuity is 6/12 or worse and cannot be explained by refraction errors, cataracts or vitreous opacities. Fundoscopy may seem deceptively normal; all patients with unexplained visual impairment must therefore be referred to an ophthalmologist. The diagnosis is confirmed by fluorescein angiography, which shows extensive capillary leakage, and slit-lamp examination, which will reveal retinal thickening. Screening and investigation of diabetic retinopathy Visual acuity and fundoscopy should be checked routinely in every diabetic patient at diagnosis, annually thereafter, and 3-6-monthly if changes other than mild background retinopathy are present. Acuity should be tested using the Snellen chart before mydriatics are given, and with any refractive errors corrected by the patient looking through his or her own distance glasses or a pinhole in a piece of card. Fundscopy must be performed through dilated pupils, as retinopathy, especially new vessel formation, often occurs far peripherally. Contraindications to mydriatics (e.g. 1% tropicamide) are glaucoma and previous eye surgery. The lens and vitreous, as well as the disc, major vessels and finally the sensitive macula ('look straight at my light'), must all be examined. Table 19.11 shows guidelines for referring the patient to the ophthalmologist. Specialized investigations include slit-lamp measurements of retinal thickness (which is increased in macular oedema), estimation of intraocular pressure, and fluorescein angiography. Fluorescein injected intravenously binds to albumin and so only leaves the circulation if the vessels are abnormally leaky. After a fluorescein bolus, rapid-sequence photographs using ultraviolet light demonstrate the retinal arteries, capillaries and veins. Fluorescence outside the vessels indicates leakage, notably in macular oedema and around new vessels (Fig. 19.20).

1

1010

Fig. 19.8

2

MCQ 19.13

TABLE 19.11 When to refer diabetic patients to the ophthalmologist • • • • • • •

Extensive or progressive haemorrhages or exudates Preproliterative changes Proliferative changes Unexplained loss of visual acuity (6/12 or worse): possible maculopathy Vitreous haemorrhage Retinal detachment urgent referral (within 24 hours) Glaucoma >

SUMMARY 9 Diabetic retinopathy • The commonest cause of blindness in Britons of working age • Background changes affect almost all cases after 20 years of diabetes • Proliferative changes develop in only 30-50% of cases • Maculopathy commonly causes blindness in type 2 diabetes; it may be invisible on fundoscopy • Laser photocoagulation prevents blindness in 50% of cases with new vessels • Vitreoretinal surgery vision in 50% of cases with advanced eye disease

Management of diabetic retinopathy With advances in laser photocoagulation and vitreoretinal surgery, many cases of severe diabetic retinopathy have become treatable. Laser photocoagulation uses blue-green argon laser light, which is absorbed by vascular structures. The laser can be used either to destroy localized targets such as clumps of new vessels or leaky capillaries, or for panretinal photocoagulation (PRP). PRP ablates the peripheral retina with 2000 or more burns, sparing the macula and maculopapillary bundle, which carries its fibres to the disc. This destroys ischaemic retina and therefore removes the stimulus to neovascularization; it also channels remaining blood flow into the macula to preserve central vision. With PRP the risks of patients with proliferative changes becoming blind within 5 years are reduced from 50 to 25%. Macular oedema can also be treated, using laser burns to seal surrounding leakage points. Vitreoretinal surgery, by removing vitreous haemorrhage and fibrous membranes and repairing detached or torn retina, can now restore vision to over 50% of eyes rendered blind by advanced diabetic eye disease. Prevention of diabetic retinopathy is a major priority. This hinges on regular screening of eyes and early ophthalmologic referral, together with improved glycaemic control, stopping smoking and treatment of hypertension. Visually impaired diabetic people should be registered as partially sighted or blind to obtain appropriate state benefits. Sympathetic counselling and practical devices such as injection aids and speaking blood glucose meters may be very helpful. 2

Other ocular complications of diabetes Diabetes accelerates the formation of senile cataracts and, rarely, 'snowflake' cataracts may develop rapidly during periods of very poor glycaemic control at any age. Cataracts must be distinguished from the blurred vision of hyperglycaemia, usually with hypermetropia, which may be due to osmotically induced changes in lens shape. Normalization of hyperglycaemia may also alter acuity for several weeks.

19

Diabetic nephropathy and other renal diseases Diabetic nephropathy is a specific microvascular disease affecting mainly the glomerulus, although tubular lesions also occur. In the UK it accounts for 25% of patients with end-stage renal failure. It is the commonest cause of premature death in type 1 diabetes, especially in patients who become diabetic before puberty, but is rarer when type 1 diabetes presents after 40 years of age. Only 30-40% of type 1 diabetes patients ever develop nephropathy, and two-thirds of these ultimately enter end-stage renal failure. Nephropathy is rarer in type 2 diabetes, but as this is commoner than type 1 diabetes the numbers of renal failure patients with type 1 diabetes and type 2 diabetes are comparable. Fortunately, the frequency of nephropathy is apparently falling. Diabetes predisposes to urinary tract infections, especially in patients with incomplete bladder emptying owing to autonomic neuropathy. Unusual organisms may be involved, including tuberculosis and gas-forming bacteria; a recognized complication is papillary necrosis, which may cause acute renal failure. Pathophysiology of diabetic nephropathy Nephropathy is one facet of generalized microvascular damage and is almost always associated with retinopathy, often proliferative. An extremely important association is with hypertension. This is present in virtually all patients with clinical nephropathy and, if untreated, accelerates the rate of decline of GFR. Susceptibility to nephropathy may relate to an inherited tendency to hypertension. Other predisposing factors include smoking and poor glycaemic control. As in other capillary beds, the glomerular basement membrane is thickened. The mesangium is expanded by amorphous material and the glomerulus ultimately becomes sclerosed, forming the pathognomonic Kimmelstiel-Wilson nodular lesions. The first functional change is increased permeability of the glomerular capillaries to albumin. Urinary albumin excretion rate (AER) rises initially into the range of 30-300 mg/24 hours, known as 'microalbuminuria'. These albumin concentrations are below the detection limits of conventional dipsticks and are demonstrable only by radioimmunoassay or other sensitive methods. In twothirds of these patients AER continues to rise, and at >300 mg/24 hours the stage of clinical nephropathy

FIG. 19.21 Evolution of albuminuria in diabetic nephropathy 'Microalbuminuria' represents an albumin excretion rate (AER) of 30-300 mg/24 hours, which is detectable by sensitive radioimmunoassay (RIA). In two-thirds of these patients AER rises into the macroalbuminuric range (>300mg/24 hours), detectable by routine dipstick testing. These patients are at risk of developing renal failure or the nephrotic syndrome.

('macroalbuminuria') becomes detectable by dipsticks. In some patients proteinuria ultimately exceeds 3.0-3.5g/day, causing the nephrotic syndrome (Fig. 19.21). GFR is initially normal, or even increased in some cases; this 'hyperfiltration' is associated with renal hypertrophy. Once AER reaches 100-300 mg/24 hours GFR starts to decline linearly, with an average loss of lOmL/min per year (Fig. 19.21). Serum creatinine levels begin to rise when GFR has fallen to 50 mL/min, and end-stage renal failure supervenes on average 5-7 years after macroalbuminuria appears. Investigations and diagnosis Microalbuminuria can now be measured easily and cheaply, and tests suitable for screening spot urine samples are available. Screening should be performed annually. Blood urea and creatinine levels can be misleadingly normal until GFR has fallen by 50% (Fig. 19.22). In about 10% of cases renal failure or proteinuria is not due to diabetic nephropathy. Atypical features, e.g. haematuria, rapid-onset renal failure, small kidney size or the absence of retinopathy, should raise this possibility. The correct diagnosis is vital for therapeutic and prognostic reasons, and may have to be established by renal biopsy. Management of diabetic nephropathy Microalbuminuric stage Much can be done at this stage to slow the pace of renal damage and delay progression to clinical nephropathy. Control of hypertension is crucial (the target is <130/80), as this significantly slows the rise in AER and fall in GFR. ACE inhibitors reduce the AER (possibly by lowering intraglomerular pressure), and should be considered in all type 1 diabetes patients with microalbuminuria. In patients with type 2 diabetes, blood pressure control (with a target

1011

SUMMARY 10 Diabetic nephropathy • The commonest cause of premature death in type 1 diabetes • Microalbuminuria precedes macroalbuminuria (clinical nephropathy) • In clinical nephropathy, GFR declines linearly with time • Treating hypertension slows deterioration in renal function • End-stage renal failure in diabetic patients should be treated as in non-diabetic people

Years FIG. 19.22 Decline in GFR in established diabetic nephropathy If untreated, GFR falls linearly at an average rate of 10ml/min per year; individual rates vary widely and may be slowed by treating hypertension, improving glycaemic control, ACE inhibitors, reducing protein intake and stopping smoking.

of <130/80) is also important. There is less strong evidence for a specific beneficial effect of ACE inhibitors in this group compared with type 1 diabetes, and renal function should be closely monitored because of the high rate of macrovascular disease and associated renal artery stenosis. The DCCT and UKPDS trials have shown the benefits of good glycaemic control in reducing the incidence of microand macroalbuminuria. Stopping smoking may also be of benefit. Clinical nephropathy (macroalbuminuric stage) Effective treatment of hypertension (present in most patients) is again essential, and ACE inhibitors may markedly reduce proteinuria. Moderate restriction of dietary protein and stopping smoking may help to maintain GFR; optimizing glycaemic control has little effect at this stage. End-stage renal failure Renal support therapy is needed when serum creatinine reaches 500u,mol/L. Prediction of this time from plots of inverse creatinine (1000 -s- creatinine in fimol/L) or GFR allows careful planning in advance (Fig. 19.22). The treatment options are the same as for non-diabetic people and are now nearly as successful. Patients must not be denied renal support simply because they have diabetes, although premature death (mainly from cardiovascular disease) is 10 times higher in diabetic than in non-diabetic patients receiving renal replacement therapy.

1

1012

MCQ 19.14

2

Case 19.3 3

MCQ 19.15

Renal transplantation is preferred for younger patients; live related donors are best, when 5-year survival exceeds 60%. Haemodialysis can be complicated by the difficulty of constructing arteriovenous fistulas in calcified arteries (which are common in nephropathy) and by postural hypotension due to fluid shifts. Chronic ambulatory peritoneal dialysis (CAPD) avoids these problems and is suitable for many patients, including the blind. Other aspects of diabetic care can be difficult for renal failure patients. Hypoglycaemia is common, as insulin and most sulphonylureas are cleared partly through the kidneys and so accumulate in renal failure. Sulphonylureas are particularly dangerous and insulin should be substituted when creatinine levels begin to rise; metformin must never be given in renal failure because of the very high risks of lactic acidosis. Proliferative retinopathy commonly coexists and may deteriorate during haemodialysis; early laser treatment is essential. Neuropathy causing foot ulceration is frequent, and autonomic neuropathy aggravates postural hypotension during haemodialysis. Cardiovascular disease ultimately kills most of these patients.12

Diabetic neuropathy Diabetic nerve damage involves both the somatosensory system, causing variable sensory and motor deficits, and the autonomic outflow to various organs (Fig. 19.23). Pathological changes affect both the Schwann cells, with segmental peeling and loss of myelin, and the axons themselves. Axonal degeneration may be accompanied by regeneration and sprouting of nerve endings. Nerve conduction velocity is decreased and smaller fibres may fire spontaneously, a possible basis of the characteristic neurogenic pain. Causative factors in diabetic neuropathy are thought to include: • Hyperglycaemia. Epidemiological studies have associated poor diabetic control with more severe nerve damage, and the DCCT has shown that good glycaemic control more than halves the incidence of clinical neuropathy in type 1 diabetes. Hyperglycaemia could damage