Severe Acute Respiratory Syndrome Edited by
Malik Peiris Department of Microbiology University of Hong Kong Queen Mary ...
87 downloads
1753 Views
2MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Severe Acute Respiratory Syndrome Edited by
Malik Peiris Department of Microbiology University of Hong Kong Queen Mary Hospital, Hong Kong SAR
Larry J Anderson Chief, Respiratory and Enteric Viruses Branch, MS A34 Centers for Disease Control and Prevention Atlanta, USA
Albert DME Osterhaus Department of Virology Erasmus Medical Centre Netherlands
Klaus Stohr Global Influenza Programme CDS/CSR/RMD Switzerland
Kwok-yung Yuen Department of Microbiology University of Hong Kong Queen Mary Hospital, Hong Kong SAR
Severe Acute Respiratory Syndrome
Severe Acute Respiratory Syndrome Edited by
Malik Peiris Department of Microbiology University of Hong Kong Queen Mary Hospital, Hong Kong SAR
Larry J Anderson Chief, Respiratory and Enteric Viruses Branch, MS A34 Centers for Disease Control and Prevention Atlanta, USA
Albert DME Osterhaus Department of Virology Erasmus Medical Centre Netherlands
Klaus Stohr Global Influenza Programme CDS/CSR/RMD Switzerland
Kwok-yung Yuen Department of Microbiology University of Hong Kong Queen Mary Hospital, Hong Kong SAR
© 2005 by Blackwell Publishing Ltd Blackwell Publishing, Inc., 350 Main Street, Malden, Massachusetts 02148–5020, USA Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ , UK Blackwell Publishing Asia Pty Ltd, 550 Swanston Street, Carlton, Victoria 3053, Australia The right of the Author to be identified as the Author of this Work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. 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, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. First published 2005 Library of Congress Cataloging-in-Publication Data Severe acute respiratory syndrome / edited by Malik Peiris ... [et al.]. p. ; cm. Includes bibliographical references. ISBN 1-4051-3031-8 (alk. paper) 1. SARS (Disease) [DNLM: 1. Severe Acute Respiratory Syndrome. 2. Disease Outbreaks—prevention & control. 3. Severe Acute Respiratory Syndrome—prevention & control. WC 505 S498 2005] I. Peiris, Malik. RC776.S27S48 2005 616.2—dc22 2004030180 ISBN-13: 978-1-405-1303-18 ISBN-10: 1-4051-30318 A catalogue record for this title is available from the British Library Set in 8.5/11 pt Stone Serif by SNP Best-set Typesetter Ltd, Hong Kong Printed and bound in India by Replika Press PVT., Ltd. Commissioning Editor: Maria Khan Development Editor: Claire Bonnett Production Controller: Kate Charman For further information on Blackwell Publishing, visit our website: http://www.blackwellpublishing.com The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards.
Contents
Contributors, vii Foreword, xi Acknowledgements, xv 1 SARS: A Historical Perspective from Hong Kong, 1 Kwok-yung Yuen and Nan-shan Zhong 2 SARS: A Global Perspective, 13 David L Heymann 3 Clinical Presentation of the Disease in Adults, 21 JY Sung and Kwok-yung Yuen 4 SARS in Children, 30 CW Leung 5 SARS: Sequelae and Implications for Rehabilitation, 36 David S Hui and Kenneth W Tsang 6 Radiology of SARS, 42 Clara GC Ooi 7 Aetiology of SARS, 50 Malik Peiris and Albert DME Osterhaus 8 Structure of the Genome of SARS CoV, 58 Paul A Rota, Xin Liu, Byron T Cook and Suxiang Tong
10 Pathology and Pathogenesis, 72 JM Nicholls and T Kuiken 11 SARS Coronavirus: An Animal Reservoir?, 79 Yi Guan, Hume Field, Gavin JD Smith and Honglin Chen 12 Comparative Biology of Animal Coronaviruses: Lessons for SARS, 84 Linda J Saif 13 Epidemiology and Transmission of SARS, 100 Angela Merianos, Robert Condon, Hitoshi Oshitani, Denise Werker and Roberta Andraghetti 14 Transmission Dynamics and Control of the Viral Aetiological Agent of SARS, 111 Gabriel M Leung, Anthony J Hedley, Tai Hing Lam, Azra C Ghani, Christl A Donnelly, Christophe Fraser, Steven Riley, Neil M Ferguson and Roy M Anderson 15 The Seasonality of Respiratory Virus Diseases: Implications for SARS?, 131 JC de Jong and WL Lim 16 Public Health Response: A View from Singapore, 139 Chorh Chuan Tan
9 Viral Diagnosis of SARS, 64 C Drosten, KH Chan and LLM Poon
v
vi
Contents
17 Public Health Response: A View from Hong Kong, 165 T Tsang
22 Counting the Economic Cost of SARS, 213 YC Richard Wong and Alan Siu
18 Public Health Response: A View from a Region with a Low Incidence of SARS, 169 James W LeDuc
23 Preparing for a Possible Resurgence of SARS, 231 Umesh D Parashar, Angela Merianos, Cathy Roth and Larry J Anderson
19 Infection Control for SARS: Causes of Success and Failure, 176 WH Seto, PTY Ching and PL Ho
24 Lessons for the Future: Pandemic Influenza, 239 Robert G Webster and David S Fedson
20 Antiviral Agents for SARS, 184 Frederick G Hayden and Mark R Denison
25 Lessons Learnt, 249 Albert DME Osterhaus and Malik Peiris
21 Vaccines, 203 Kanta Subbarao
Appendix, 255 Index, 257 Colour plate facing page 80
Contributors
Larry J Anderson Respiratory and Enteric Viruses Branch, Division of Viral and Rickettsial Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, GA, USA
Byron T Cook Respiratory and Enteric Viruses Branch, Division of Viral and Rickettsial Diseases, Centers for Disease Control and Prevention, Atlanta, GA, USA
Roy M Anderson Department of Infectious Disease Epidemiology, Faculty of Medicine, Imperial College, London University, London, UK
Mark R Denison Department of Microbiology and Immunology, Vanderbilt University School of Medicine, Nashville, TN, USA
Roberta Andraghetti Department of Communicable Disease Surveillance and Response, World Health Organization, Geneva, Switzerland KH Chan Department of Microbiology, Queen Mary Hospital and the University of Hong Kong, Pokfulam, Hong Kong SAR Honglin Chen Department of Microbiology, The University of Hong Kong, Hong Kong SAR PTY Ching Department of Microbiology, Queen Mary Hospital and the University of Hong Kong, Pokfulam, Hong Kong SAR Robert Condon World Health Organization Representative Office in the South Pacific, Suva, Fiji
Christl A Donnelly Department of Infectious Disease Epidemiology, Faculty of Medicine, Imperial College, London University, London, UK Christian Drosten Bernhard Nocht Institute for Tropical Medicine, National Reference Center for Tropical Infectious Diseases, Hamburg, Germany David S Fedson Sergy Haut, France. Formerly, Department of Internal Medicine, University of Virginia School of Medicine, Charlottesville VA, USA Neil M Ferguson Department of Infectious Disease Epidemiology, Faculty of Medicine, Imperial College, London University, London, UK
vii
viii
Contributors
Hume Field Animal Research Institute, Department of Primary Industries, Queensland, Australia
JC de Jong Erasmus University, Rotterdam, Netherlands
Christophe Fraser Department of Infectious Disease Epidemiology, Faculty of Medicine, Imperial College, London University, London, UK
T Kuiken Veterinary Pathologist, Erasmus Medical Centre, The Netherlands
Azra C Ghani Department of Infectious Disease Epidemiology, Faculty of Medicine, Imperial College, London University, London, UK Yi Guan Department of Microbiology, University of Hong Kong, Pokfulam, Hong Kong SAR Frederick G Hayden Division of Infectious Diseases and International Health, Department of Internal Medicine, University of Virginia School of Medicine, Charlottesville, VA, USA Anthony J Hedley Department of Community Medicine, University of Hong Kong, Hong Kong Special Administrative Region, China David L Heymann Representative of the Director-General for Polio Eradication, formerly Executive Director of Communicable Diseases, World Health Organization PL Ho Department of Microbiology, University of Hong Kong and Queen Mary Hospital, Pokfulam, Hong Kong SAR David S Hui Department of Medicine & Therapeutics, Chinese University of Hong Kong, Prince of Wales Hospital
Tai Hing Lam Department of Community Medicine, University of Hong Kong, Hong Kong Special Administrative Region, China James W LeDuc Division of Viral and Rickettsial Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta USA Chi Wai Leung Department of Paediatrics and Adolescent Medicine, Princess Margaret Hospital, Hong Kong Special Administrative Region, China Gabriel M Leung Department of Community Medicine, University of Hong Kong, Hong Kong Special Administrative Region, China WL Lim Public Health Laboratory Centre, Hong Kong SAR Xin Liu Respiratory and Enteric Viruses Branch, Division of Viral and Rickettsial Diseases, Centers for Disease Control and Prevention, Atlanta, GA, USA Angela Merianos Department of Communicable Disease Surveillance and Response, World Health Organization, Geneva, Switzerland JM Nicholls Department of Pathology, University of Hong Kong, Queen Mary Hospital, Hong Kong SAR
Contributors
Clara GC Ooi Department of Diagnostic Radiology, University of Hong Kong, Queen Mary Hospital, Hong Kong SAR
ix
Linda J Saif Food Animal Health Research Program, Ohio Agricultural Research and Development Center, Ohio State University, Wooster, OH 44691, USA
Hitoshi Oshitani Division of Combating Communicable Diseases, World Health Organization Regional Office for the Western Pacific, Manila, Philippines
WH Seto Department of Microbiology, Queen Mary Hospital and the University of Hong Kong, Pokfulam, Hong Kong SAR
Albert DME Osterhaus Department of Microbiology, University of Hong Kong, Queen Mary Hospital, Hong Kong SAR
Alan Siu Faculty of Business and Economics, University of Hong Kong, Pokfulam, Hong Kong SAR
Umesh D Parashar Respiratory and Enteric Viruses Branch, Division of Viral and Rickettsial Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, GA, USA
Gavin JD Smith Department of Microbiology, University of Hong Kong, Pokfulam, Hong Kong SAR
Malik Peiris Department of Microbiology, University of Hong Kong and Queen Mary Hospital, Pokfulam, Hong Kong SAR LLM Poon Department of Microbiology, University of Hong Kong, Pokfulam, Hong Kong SAR Steven Riley Department of Infectious Disease Epidemiology, Faculty of Medicine, Imperial College, London University, London, UK Paul A Rota Respiratory and Enteric Viruses Branch, Division of Viral and Rickettsial Diseases, Centers for Disease Control and Prevention, Atlanta, GA, USA Cathy Roth World Health Organization, Geneva, Switzerland
Kanta Subbarao Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, USA JY Sung Department of Medicine and Therapeutics, Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong SAR Chorh Chuan Tan Provost, National University of Singapore, Singapore Suxiang Tong Respiratory and Enteric Viruses Branch, Division of Viral and Rickettsial Diseases, Centers for Disease Control and Prevention, Atlanta, GA, USA Kenneth W Tsang University Department of Medicine, University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong
x
Contributors
T Tsang Consultant (Community Medicine), Centre for Health Protection, Department of Health, Hong Kong SAR
YC Richard Wong Faculty of Business and Economics, University of Hong Kong, Pokfulam, Hong Kong SAR
Robert G Webster Division of Virology, St Jude’s Research Hospital, Memphis, TN, USA
Kwok-yung Yuen Department of Microbiology, University of Hong Kong, and Queen Mary Hospital, Pokfulam, Hong Kong SAR
Denise Werker Department of Communicable Disease Surveillance and Response, World Health Organization, Geneva, Switzerland
Nam-shan Zhong Guangzhou Institute of Respiratory Diseases, Guangzhou Medical College, Guangzhou, Guangdong Province, People’s Republic of China
Foreword
The participation of infectious disease experts from across the globe in this book addressing a topic unknown just two years ago is witness to the tempo of the 21st Century. This book will be enormously helpful to those who must help to design national policies, and to those contemplating joining the research enterprise on SARS, and the multitude of present and prospective infectious disease threats. If I could think of snappy Greek terms, I’d be talking about some synonyms of travelosis and global-osis as the trajectory we can look forward to in the century opening up before us. In the history of disease, many of its ingredients were anticipated by the AIDS pandemic 20 years earlier. But the long chronicity of HIV infection makes it far less dependent on modern air transport for its dissemination. SARS, for now, takes pride of place with its exploitation of the jetliner as its mode of global dissemination. Prompted by HIV-AIDS, SARS, and a score of other major outbreaks, we have only begun to contemplate the implications of the 21st Century human condition: population crowding often in huge conurbations, typically in close proximity to jet aircraft airports, counting almost 2 million passengers bound for international destinations every day. All this is compounded by intense stratification of material income and health services. What a cauldron!1 If SARS uses high-tech vectors, it has primitive origins in dietary zoonosis, but arguably augmented by urban appetites
nourished by 2003-style levels of affluence. Then the quality of public health infrastructure over the surrounding terrain will be allimportant in determining the level of spread of the new infection. New therapeutic measures are hard to come by, and take years to develop and validate; so it is no surprise that all that was available for SARS was the classic repertoire of case-identification, isolation, and contact tracing and management (e.g. quarantine). Wherever and whenever the political will was consolidated, these were the measures that successfully contained the outbreak of 2002–2003. The response to SARS is beautifully exemplified in the papers in this volume: the extraordinarily rapid and competent identification of the agent of SARS as a novel coronavirus, and the ensuing development of biomolecular technologies for diagnosis and the detective work of tracking the origins and spread of SARS. In this pause we might contemplate the prospects of new emerging infections in general, where they might come from, what if anything we can do to raise our guard. We know virtually nothing of the ultimate evolutionary origins of viruses. Dependent on a host’s metabolism, they are unlikely to be primitive; and likely most viruses are offshoots of host genomes, including their organelles — which in turn may be of symbiotic origin. We then ask, what are the logical possibilities for sources of DNA (read also ‘or RNA’ in all that follows). Viral innovations are more likely to occur as
xi
xii
Foreword
mutations or recombinants or fragmentary additions to existing viruses which have already trodden the tortuous path to adaptation to human, animal, plant or microbial hosts. Where can such DNA be found? Arguably from the hosts themselves, and then by extension also hypothetically from any fragment, plasmid, virus inhabiting those hosts. The cardinal lesson, learned from the practice of biotechnology, is the nearly universal promiscuity of DNA; homologous sequences will pair and recombine no matter what their origin; and more rarely non-homologous segments can also be patched together. What happens in the lab must surely be feasible somewhere in our vast biosphere.
‘Animal–vegetable–mineral’ 1 Anthroponoses 2 Zoonoses (vertebrate or invertebrate) 3 Phytonoses 4 Bacteria and their viruses, and other unicellulars 5 Synthetic DNA
1 Anthroponoses Confining ourselves for the moment to viruses, the majority of anthroponoses are all too familiar in our history, like smallpox and paediatric viruses. Vaccines are available for almost all of the ancient ones. HIVAIDS has rapidly transformed itself from a primate zoonosis to the most devastating viral disease now on the planet. The hepatitides are just being sorted out. The very recent discovery of metapneumovirus, and most recently, yet another novel human coronavirus (NL-63) is a caution that still further discoveries may be in the offing: we have not reached the end of the catalog. So many pneumonias remain undiagnosed even post mortem, we can hardly be reassured. Blood transfusion and organ transplants must now be added to sexual transmission, close personal contact, and arthropods as vectors, all augmented by 21st Century transport. We are also beset with a handful of nearly benign infections, rhinoviruses and other
‘common colds’. Any of these, or all of the above, have the potential, in principle, of mutating to far more aggressive strains, or of recombining with already virulent strains and augmenting their transmissibility. It is some reassurance that this has not come about (at least to our knowledge) during the past hundred years, more or less, of relevant observation. One of the urgent priorities for the genomic revolution will be the meticulous cataloguing of ‘every’ human virus, and more enlightened insight of the prognosis, based on the dissection of pathogenetic pathways. How clumsy are our purported explanations of the seasonality of respiratory disease. These clumsy explanations show how much we still have to learn.2
2 Zoonoses We should not be surprised that the majority of new outbreaks are of zoonotic origin.3 A myriad of host species harbor a matching number of endogenous zooviruses, that have hardly begun to be indexed. These zooviral infections are often nearly benign in their primary host, with coevolution of host and parasite toward a mutually favorable symbiosis. For most zooviruses, human encounter will have no consequence: either the virus is equally well adapted to the human host; else it may be poorly adapted to the cellular environment of the new human host. These then escape our notice, though some may break through the level of detectability by serosurveys. In a few examples, like monkeypox or H5N1 influenza or hantavirus, we may see sporadic human disease, sometimes quite virulent, but little or no human–human transmission that would ignite a pandemic. The barriers to such transmission are ill-understood: do they reflect disease disproportionate to viral shedding? Host idiosyncrasies: genotype or intercurrent infection or cross-reacting immunity? There is no guarantee that such hypothetical barriers will remain unbreached, with intense selection for mutants that can transcend them, as perhaps
Foreword
occurred in the transition of animal coronaviruses to human SARS. It is anachronistic that one must enter a plea, on behalf of global health, to regulate the consumption of ‘bush-meat’ (whether in its African or Asiatic manifestations) and of the international transport of exotic pets. Our zoos must learn to use the best of veterinary science and art. This is not merely to protect the consumers, but to avert the fallout when exotic diseases spread.
3 Phytonosis No animal viruses of plant origin come to mind, though I am unaware of any concerted effort e.g. to elicit expression of TMV RNA in human cell culture. Pseudomonad bacteria might be expected to elicit opportunistic infection on any substrate.
4 Bacteria and their viruses, and other unicellulars Numerous hints prevail of lateral transfer of bacterial and animal host genes, from genomic studies. There are recurrent claims of expression of bacteriophage and plasmid genes in mammalian cell culture. Recently, the Courvalins4 have demonstrated the delivery of genes by Shigella internalized in epithelial cells. This was intended to be a vector for DNA vaccines, but it is easy to imagine recombination between the bacterial DNA with other viral DNA already entrained in the target cells. Similar prospects might apply to other unicellular parasites (Plasmodium, Leishmania).
5 Synthetic DNA Here we enter the ‘mineral’ world, mainly artefacts of the laboratory. Free DNA has been found to be remarkably potent in entering muscle and other cells, and thereby opened the door to DNA vaccines. This mode of entry bypasses the need for specific receptors, applicable to virus DNA as well: witness the infectivity of synthetic po-
xiii
liovirus.5,6 Free DNA may also occur in abundance with lysis of pus, heavily laden with bacterial flora, witness the market for deoxyribonuclease in symptomatic relief of viscous clogging of airways in cystic fibrosis. This discussion inevitably leads to the odious topic of malicious construction of hyperpathogens and their application to biowarfare, in the hands of terrorists or of organized states. The use of contagious agents is an attack on the whole world’s population, insane for any imaginable political agenda. Were it not for our actual experiences of the last 20 years, from AIDS to SARS, my remarks might be put down as overenthusiastic ravings. There has been one alarm after another, and the beginnings of a concerted international response. The technical cooperation elicited by the SARS episodes, and the monumental contributions of researchers from the PRC, Europe, Canada, the US are encouraging for the prospect of more effective responses, but they still pale before the magnitude of the threats from unforeseeable new emergents. The unique position of the Hong Kong SAR, as the pivotal point of entry and exit, of pathogens and of knowledge about them, has been highlighted. A new flu reassortment, matching the 1918–1919 episode, is on everyone’s mind; but we know our current capability to generate stocks of a new vaccine is totally ‘out of synch’ with the speed of its spread in the modern world. We must renew our cooperative planning with a recognition of the magnitude of the threat, and humility about the paucity of our knowledge of the globe’s viral agents and how they might break out. If we can achieve that, we may retrospectively view the SARS of 2002–2003 as a blessing, that those who perished may not have died in vain. Prof. Joshua Lederberg Raymond and Beverly Sackler Foundation Scholar The Rockefeller University New York
xiv
Foreword
References 1 Institute of Medicine. Microbial Threats to Health: emergence, detection and response. Institute of Medicine, Washington, 2003. 2 Gwaltney JM. The Jeremiah Metzger Lecture. Climatology and the Common Cold. Trans Am Clin Climatological Assoc 1984;96: 159–75. 3 Taylor LH, Latham SM, Woolhouse MEJ. Risk factors for human disease emergence. Phil Trans R Soc Lond 2001;356(B): 983–9.
4 Grillot-Courvalin C, Goussard S, Courvalin P. Wild-type intracellular bacteria deliver DNA into mammalian cells. Cell Microbiol 2002;4: 177–86. 5 Racaniello VR, Baltimore D. Cloned poliovirus complementary-DNA is infectious in mammalian cells. Science 1981;214: 916–19. 6 Cello J, Paul AV, Wimmer E. Chemical synthesis of poliovirus cDNA: Generation of infectious virus in the absence of natural template. Science 2002;297: 1016–18.
Acknowledgements
Many people assisted in the conception, gestation and birth of his book. The project was initiated in discussions with Professor Paul KS Tam, The University of Hong Kong. We acknowledge Dr JM Nicholls for assisting in the editing and proof reading of manuscripts. We thank the team at Blackwell Publishing who helped make this book a
reality, in particular, Claire Bonnett and Maria Khan. We dedicate this book to health care workers the world over who were affected and afflicted by SARS, and in a most special way, to those gave their lives in the pursuit of their profession.
xv
Chapter 1
SARS: A Historical Perspective from Hong Kong Kwok-yung Yuen and Nam-shan Zhong
History is apt to repeat itself. In 1894, an outbreak of plague started in Canton, China. The epidemic soon spread to Hong Kong and was first described by Dr James Lowson, a government medical superintendent.1 It was later carried by ships to California and then on to port cities in South America, Africa and Asia. This epidemic led to the discovery of the plague bacillus,2 Yersinia pestis, after a race between two groups of researchers to identify the cause. The rat flea was subsequently identified as the vector and the main animal reservoir recognized to be Rattus rattus and Rattus norvegicus.3–5 Similarly, this is what happened with an outbreak of atypical pneumonia in late 2002. It is notable that the first intelligence of both epidemics came from journalists, whose pursuit of the truth keeps the public well-informed. Their reporting in the Hong Kong media on 10 February 2003 of an outbreak of a rapidly spreading atypical pneumonia in Guangdong province came earlier than the official announcement. However, not all media reports served to keep the public well-informed of developments. For instance, rumours about the death of many young patients, including doctors and nurses, helped incite panic buying of vinegar and traditional Chinese medicinal herbs in Guangdong province.The public believed that vapour released by boiling vinegar could kill germs circulating in the atmosphere, while herbal extracts could detoxify and dissipate the heat gener-
ated by infection caused by this atypical pneumonia. In the same week, the Hong Kong government formed an expert panel to study the risk of a similar outbreak in the Hong Kong Special Administrative Region (HKSAR) and started the surveillance of severe pneumonia cases in the region. At that time, hospital surveillance did not reveal any increase in the number of cases of severe community-acquired pneumonia. Unfortunately, official exchange of information between the HKSAR and Guangdong province in China was not readily established at that juncture because of an inadvertent ‘firewall’ created by the ‘one country–two systems’ situation. In fact, for 2 months cases of severe atypical pneumonia had been recorded in five cities around Guangzhou.6 But obtaining medical information along official channels from Guangdong province proved bureaucratic and problematic for HKSAR authorities. The first such case of atypical pneumonia was reported in Foshan, a city 24 kilometres away from Guangzhou, on 16 November 2002. A month later, a chef from Heyuan who worked in a Shenzhen restaurant was also infected. He had an occupational history of regular contact with wild game-food animals. His wife, two sisters and seven hospital staff who had contact with him were also affected. All patients had high fever, respiratory symptoms and infiltrate on chest radiograph. In China from 16 November 2002 to 9 February 2003, 305 cases of atypical
1
2
Severe Acute Respiratory Syndrome
pneumonia were reported, 105 of which were in health-care workers. The epidemiology of the disease was described with initial consensus among medical experts in China.7 It was an atypical pneumonia of unknown aetiology but probably viral in origin. The incubation period ranged from 1 to 11 days. There was clustering of cases in families and hospitals, which suggested the necessity of close contact for transmission, probably by respiratory droplet. Thus single-room isolation, environmental decontamination and hospital-staff protection with the use of masks and hand washing were recommended as control measures, in addition to the notification of the disease and contact tracing. With the HKSAR epidemic, the index patient and the key connection to the global epidemic was CCL, a professor of nephrology from a teaching hospital in Guangzhou, China.11 Between 11 and 13 February 2003, he had a history of contact with patients suspected to have this unusual atypical pneumonia. CCL then developed flu-like symptoms, including a fever and cough, which he self-treated with penicillin and ofloxacin before coming to Hong Kong. He checked into the Metropole Hotel on 21 February and within a period of 24 hours had infected 16 people. This was one of those unexplained superspreading events that eventually took on global significance. The professor was admitted to the Kwong Wah Hospital the next day where his condition rapidly deteriorated. No pathogenic bacteria, viruses, fungi, or parasites could be found in his respiratory secretions, blood or other body fluids. The turning point of the whole event was that his brother-in-law, YPC, was soon admitted with a similar condition. After upper respiratory tract specimens and a bronchoalveolar lavage failed to yield an aetiological agent, the decision was made to carry out an open lung biopsy. The sample was sent to the Queen Mary Hospital, the teaching hospital of the University of Hong Kong, for microbiological analysis. This turned out to be the key spec-
imen in identifying the aetiological agent of SARS. As a result of this superspreading event, visitors to the Metropole Hotel unwittingly carried the disease to other hospitals in the HKSAR and by air travel to Vietnam, Canada, Singapore, the Philippines, the United Kingdom, the United States and back again to China12 (Fig. 1.1). Dr Carlo Urbani, a physician stationed at the World Health Organization (WHO) office in Hanoi, answered the call for assistance in investigating a patient who had stayed in the Metropole Hotel while in the HKSAR. Within the next 2 weeks, nosocomial transmission occurred affecting health-care workers, patients and visitors to the hospital in Hanoi. Dr Urbani recognized that he was probably dealing with a hitherto unknown disease and he initiated measures to control this hospital outbreak. He described this unusual disease and informed WHO.13 The WHO subsequently labelled this atypical pneumonia as severe acute respiratory syndrome (SARS). Dr Urbani’s description of the disease, to which he later succumbed, alerted the authorities and resulted in an unprecedented collaboration of 11 research laboratories in 9 countries. This international collaboration involved daily teleconferencing and exchange of specimens and ultimately led to the rapid discovery of the aetiological agent of SARS. Discovery was by educated guess and trial and error. The differentiation between typical and atypical pneumonia is a historical one, and is often difficult.8 Diagnosis rests on the clinical syndromes, the microbiological analysis and the therapeutic response to beta-lactam antibiotics. In general, typical pneumonia has a very acute onset with fever, chills, pleuritic chest pain, tachypnoea and cough with rusty-coloured sputum. It is mainly caused by pyogenic bacteria such as Streptococcus pneumoniae, Haemophilus influenzae and Staphylococcus aureus. Peripheral blood counts usually show an increase in neutrophils. The disease responds very well to beta-lactams. These bacteria are therefore
SARS: A Historical Perspective from Hong Kong
Mainland China (5327)
Hong Kong SAR Hotel Metropole (16)
3
Hong Kong SAR hospitals Cluster in Kwong Wah Hospital (1)
UK (4)
Philippines (14)
3
1
Cluster in St Paul’s Hospital (12)
2
1
Cluster in Prince of Wales Hospital (238)
Canada (251)
2 Cluster in Amoy Gardens (321) USA (29)
3
Singapore (238)
Vietnam (63)
Cluster in United Cluster in Princess Christian Hospital Margaret Hospital
3
1
Index case of the outbreak in Hong Kong SAR, a health care worker with contact history with SARS patients in Guangzhou
Infected contacts at Hotel Metropole
Disseminated via air travel
Dissemination without air travel
Numbers in parentheses refer to the total number of SARS cases in respective locations and hospital-acquired cases institutions Figure 1.1 Superspreading of SARS.
unlikely to be the cause of a major outbreak in the hospital setting where antibiotics are commonly prescribed. Atypical pneumonia usually has a less acute onset with preceding upper respiratory tract symptoms.9 The cough is often nonproductive. The findings on clinical examination are disproportionate to the chest radiographic changes. Historically, cases have been caused mainly by Mycoplasma pneumoniae, followed by Chlamydia pneumoniae, Chlamydia psittaci, Coxiella burnetti, Legionella pneumophila and less commonly various respiratory viruses. Pe-
ripheral blood examination often shows either normal or increased white blood cells. Patients often do not respond to treatment with beta-lactams. The most common cause of atypical pneumonia is Mycoplasma pneumoniae, a bacterium that does not have a cell wall. Moreover, the beta-lactams do not penetrate into infected cells very well and would not be very effective against the other causes of atypical pneumonia, which are predominantly intracellular pathogens. All four bacterial agents are sensitive to treatment with the macrolides, tetracyclines and quinolones. They are often
4
Severe Acute Respiratory Syndrome
associated with a specific antibody response in the convalescent serum of patients. The only remaining differential diagnosis in atypical pneumonia is a viral cause. The influenza virus was therefore high on the list of differential diagnoses because the season was late winter/early spring. However, investigations can easily be misled by reasonable anticipations that turn out to be red herrings. On 11 February 2003, a 33year old man was admitted to the Princess Margaret Hospital in the HKSAR for a severe community-acquired pneumonia and was soon followed by his 9-year-old son. A week earlier, his youngest daughter had died of pneumonia when the family were visiting Fujian in China. The man died on 17 February, while his son eventually recovered. Influenza A subtype H5N1 virus was isolated from the man and his son. This led to the World Health Organization (WHO) issuing a global alert on avian flu on February 19. Previously in 1997, the HKSAR had been struck by an epidemic of influenza A subtype H5N1 involving 18 patients and 6 deaths.10 The outbreak was finally brought under control by the territory-wide slaughter of 1.5 million chickens. Isolated outbreaks of avian flu continued to occur at farms and markets in 2001 and 2002. On 10 December 2002, dead waterfowl found in a recreational park in the HKSAR and subsequently some farm chickens were shown to have influenza A subtype H5N1 virus. These recent events placed a strong bias on the suspected cause of the growing number of atypical pneumonia cases reported in China. In February 2003, the anticipation of an outbreak of influenza A subtype H5N1 turned out to be a red herring. The other possibility was an antigenically drifted human influenza A or B, which happens every few years. This theory was followed by the very ominous possibility of an antigenically shifted influenza A. But these were considered unlikely because cell cultures and chick-embryo inoculation at the University of Hong Kong were optimized for influenza. Moreover, reverse
transcription-polymerase chain reaction (RT-PCR) assays for the matrix gene and antigen tests for nucleoprotein were uniformly negative for influenza. Clinically, patients did not respond to antiviral drugs specific for influenza virus, such as amantadine and oseltamivir. Other possibilities had to be considered in view of the absence of evidence favouring the diagnosis of different types of influenza viruses. Other types of viruses that have an outbreak potential in hospitals include the adenovirus, the parainfluenza viruses, the respiratory syncytial virus, the metapneumovirus, the rhinovirus and the coronaviruses. Only influenza and adenovirus are known to cause severe pneumonia in immunocompetent young adults. The rest are much less likely to cause such a serious illness. Further laboratory investigations with RT-PCR, PCR using consensus primers and antigen testing excluded the first five types of virus. However, these initial investigative findings could not exclude a coronavirus or another novel virus. As was the case in 1894 with the bacillus plague, the initial claims of discovery were conflicting. Though experts in China found electron microscopic evidence of chlamydial infection in the postmortem lung tissue of early cases of SARS,14 this was not found in patients from other countries and most SARS patients did not respond to antichlamydial antibiotics. There was a subsequent report that scientists from the Chinese Academy of Military Medical Science in China might have observed coronavirus-like particles under the electron microscope on 26 February.15 On 19 March, two research groups in Germany and the HKSAR announced the finding of paramyxovirus-like particles in samples from SARS patients. Between 21 and 27 March, groups from the HKSAR, the USA and Germany independently found a novel coronavirus from SARS patients after mounting specific antibody response tests on whole virus immunoassays using im-
SARS: A Historical Perspective from Hong Kong
munofluorescent or ELISA format.16–18 The initial fragments of genes obtained by differential display PCR were used in the development of RT-PCR for rapid diagnosis. On 16 April, Koch’s postulates for the viral causation of SARS by this novel coronavirus were satisfied in a primate model.19 On 12 April, the full genome of the virus was sequenced and available on the internet to all researchers.20,21 This rapid progress in investigating the aetiological agent of SARS is one of the most amazing achievements in medical history. Within 2 months of its discovery, the complete deciphering of the full genome of a pandemic microbe has markedly accelerated the research on the pathogenesis, diagnostics, antivirals and vaccines for SARS. Although no vector like the rat flea that transmitted the plague was involved in SARS, there were other horrifying mechanisms that amplified the outbreak in the HKSAR. On 4 March, a visitor to the Metropole Hotel was admitted to the Prince of Wales Hospital in the HKSAR with acute community-acquired pneumonia. The administration of albuterol by a nebulizer on
5
this patient led to a hospital outbreak that involved 238 persons, mostly health-care workers, patients and visitors.22 One of these affected was a patient on haemodialysis. He was admitted to the Prince of Wales Hospital on 15 March with suspected atypical pneumonia, which was subsequently attributed virologically to influenza A. The patient improved and was discharged on 19 March. He stayed at an apartment in Block E at Amoy Gardens, a private housing estate. What followed was a devastating outbreak involving 321 people in that estate. The amplification mechanism was traced to a faulty sewage system.23 The U-trap of the bathroom floor-drains in most apartments were dry and therefore had lost their function as a seal to the soil stack (Fig. 1.2). The common use of an exhaust fan in a closed bathroom generated a strong negative pressure that drew infected droplets from the soil stack into the bathrooms. On 21 March, a leak occurred in the water pipe to Block E and led to a shutdown of the flushing system. It has since been shown that the SARS coronavirus can survive for 4 days in faecal matter. Therefore, the diarrhoeal stool of
Figure 1.2 Properly maintained U-trap. (Reproduced with permission from Report of the SARS Expert Committee.)
6
Severe Acute Respiratory Syndrome
the haemodialysis patient was passed into the sewage system24 and infected droplets were drawn into the bathroom and followed the current of the exhaust fan through the light well to various floors in Block E. Other mechanisms of amplification of this epidemic at this housing estate by infected cats, rats or passive carriage by cockroaches were suggested but remain unproven despite the positive isolation of the SARS coronavirus from domestic cats with seropositivity towards this virus. This large outbreak at the housing estate was an excellent opportunity for the study of the natural progression of SARS since the patients were exposed at almost the same time. They were home-quarantined and admitted to hospital on the first day of the onset of any symptoms. Their clinical symptoms and signs, blood parameters, radiological changes, microbiological findings and outcome could be prospectively monitored and analysed. A unique pattern of changes of viral load in their respiratory secretions was documented and had implications in the design of treatment strategies for future epidemics.25 The observation that 70% of these patients developed diarrhoea led to the discovery of a large amount of virus being present in the stool. This has strengthened the argument that the major amplification mechanism of the SARS outbreak was due to the faulty architectural design of the bathrooms. Finding the responsible virus and defining the syndrome of SARS are just the beginning of a long and difficult journey. The ultimate goal is to treat and prevent SARS effectively. The necessary tools are rapid and sensitive diagnostic tests, effective and safe antiviral drugs, pragmatic and compliable infection-control measures, and finally a safe and effective vaccine. None of these will come about without painstaking research. The antibody test using the infected cell line as antigen only starts to show a positive response at around day 10 of infection. It takes 28 days before most SARS patients can mount an antibody response. The use of recombinant viral proteins as the antigen
might lead to the production of a more sensitive and specific antibody test.26 Again, the rapid RT-PCR assay is only positive in 25% of SARS patients at the time of admission. The sensitivity of the assay is being improved to over 80% after a careful optimization of the RNA extraction,27 the choice of the primer sequence, the testing conditions, an additional amplification step called a nested reaction, and a final hybridization step. The histopathological findings from the tissues of deceased patients suggest that part of the damage to the lung is mediated by macrophages, which are highly activated as a result of the viral infection.28 Moreover, viral-load study shows that the viruses increase in number during the first 10 days and then decrease irrespective of whether the patient is improving or deteriorating. Thus at least part of the damage might be related to the immune dysregulation of the host, which may be alleviated by steroids. This may decrease the need for intubation and the workload of the intensive care unit. However, such immunosuppressive treatment is also associated with other potentially fatal side-effects that include bacterial or fungal superinfections. The answer to the treatment of SARS must lie in the development of effective antiviral agents that decrease the peak viral load and the associated immune-mediated damage. This is supported by the viral load study at days 10–15 which showed that the manifestations of SARS such as respiratory failure, hepatitis, diarrhoea, abnormal urinalysis and death are closely related to viral genome copies in the relevant clinical specimens.29 At the time of writing, there is still no randomized placebo control trial to show whether any antiviral is effective. Ribavirin by itself did not appear to have sufficient antiviral effect.30 This was used in the early stage of the outbreak as a broad-spectrum antiviral agent when the aetiological agent was still unknown. Later, a relatively low dose of ribavirin combined with a protease inhibitor or an interferon called alfacon-1 were used in clinical trials with only historical
SARS: A Historical Perspective from Hong Kong control.31,32 Ribavirin may also function as an immunomodulator because the drug was reported to be highly effective in the treatment of fulminant hepatitis in mice caused by a mouse coronavirus.33 Because the complete genome sequence of four different SARS coronavirus strains was available within a month of the discovery of the virus by four different research centres in Canada, the United States and the HKSAR, we know that there are a number of enzymatic targets for antiviral therapy. These include the RNA replicase, the helicase, the proteases and mRNA methyltransferase.34 Researchers are now rushing to screen combinatorial chemical libraries to find effective antiviral treatment for SARS. One of the most important surface targets for antiviral treatment is the spike protein, which makes up the surface projections that give the coronavirus the morphology of a crown. It functions like a claw, which allows the virus to attach to human cells and enables the virus envelope to fuse with the host cell membrane. This fusion process allows the virus to enter the cell to start a new cycle of infection. Short chains of amino acids are designed to glue-up the mechanical apparatus of this claw so that the virion can no longer enter the human cell. Short interfering RNA inhibiting viral transcription is another treatment modality that is also undergoing investigation.35 However, all these potentially exciting new drugs will only move from the laboratory and on to the clinical trial stage after at least another year, and they will not be commercially available for a further 2–3 years. A great many in vitro and animal experiments are needed before trials in humans are justified. The same difficulties will apply to the development of a vaccine. The relative importance of systemic or mucosal immunity against surface targets such as spike, membrane, and envelope proteins by neutralizing antibody or by cytotoxic T-lymphocyte response against nucleoprotein and other targets is still unclear. But recent studies suggested that neutralizing antibody against Spike (the spike protein) appeared to be very
7
important. The use of live attenuated coronavirus is out of the question for fear of reversion to virulence or recombination with wild strains to form new wild types. The most readily available vaccine to undergo clinical trials will be an inactivated SARS coronavirus. However, laboratory safety is a major issue when culturing a huge stock of coronavirus. This is particularly worrying in light of a recent reports of a laboratoryacquired cases of SARS in Singapore, Tawain and China. Recombinant proteins expressed in mammalian cells or yeast are alternatives for inducing good neutralizing antibody response. DNA vaccines delivered by non-replicating viral vectors such as modified vaccinia virus Ankara and adenovirus are likely to induce both good antibody and cytotoxic T-cell response. Non-replicating coronavirus particles will be safe and will closely mimic the live coronavirus in terms of immunogenicity. Irrespective of which approach is used for immunization, it is important to note that immune enhancement disease has occurred in feline peritonitis coronavirus,36 inactivated measles and respiratory syncytial virus vaccination.37,38 The economic viability of a SARS vaccine depends on whether and when SARS comes back and whether antigenic variation will affect the effectiveness of this vaccine. Whether SARS will come back depends on the existence of ongoing mildly symptomatic infection in humans, persistent animal reservoirs or unsafe laboratories working with SARS. On 24 May 2003, a collaboration between the University of Hong Kong and the Shenzhen Centre for Disease Control (CDC) identified a precursor of the SARS coronavirus in civet cats and a raccoon dog.39 Sera from people in close contact with these animals had a high prevalence of antibody against the SARS coronavirus. In essence, SARS is probably a zoonosis. The virus may have jumped the human barrier repeatedly without being noticed, but it has not yet mutated sufficiently to cause human-to-human spread. Comparing the animal precursor with the
8
Severe Acute Respiratory Syndrome
human isolates of SARS coronavirus, there is a 29-base-pair deletion between ORF 10 and 11 in the human isolates. The deletion and perhaps some other mutations may have sufficiently changed the property of the virus as to cause a pandemic. The situation is in some way similar to what has happened with influenza viruses. In 1997, genetic reassortment between avian influenza strains of H5N1 and H9N2 or H6N1 produced an epidemic of 18 patients with 6 fatalities in HKSAR.40 Fortunately the genetic reassortment did not enable the new reassortant to spread efficiently among humans. But we are aware that more disastrous events have happened in the past hundred years. In 1918 the Spanish Influenza A pandemic, known as Spanish flu, swept across the world leaving between 20 and 40 million people dead.41 However, the causative agent was not discovered until 1933 by Smith.42 In February 1957, the Asian influenza H2N2 originated in Southern China, overwhelmed the region within a month, spread to Hong Kong and Singapore 2 months later, and became an obvious pandemic by November.43 About a decade later history repeated itself in July 1968 with the emergence of the Hong Kong H3N2 influenza A pandemic, which is often remembered as the Hong Kong flu.44 These experiences strongly suggest that SARS and influenza appeared to have brewed in southern China, where the density and proximity of human and animal populations are unmatched. The genetic change empowering the virus to spread between humans is the key event leading to all of these disasters. The world, and the HKSAR in particular, has a lot to learn from this major epidemic that affected over 1700 people and resulted in almost 300 deaths. The 1894 outbreak of bacillus plague in Hong Kong was largely the result of negligence in personal and environmental hygiene, pest infestation and overcrowded living conditions. The HKSAR is now a modern cosmopolitan region with outspoken mass media, an elected legisla-
ture and groups of physicians and microbiologists working at the cutting edge of research in infectious diseases. Hong Kong has also grown from a small fishing village to become a vital doorway into China and an international centre for trade, finance, business and communications. The HKSAR has the highest throughput container port and the busiest airport in terms of passengers and international cargo in the world. Since China adopted an open economic policy in 1978, increasing numbers of visitors and cargo flow between China and the HKSAR, providing an annual average of 23% in trade value. Since the political transition in 1997, the integration of the HKSAR with the Pearl River Delta of Southern China has accelerated. The open door policy and economic boom in Southern China is naturally associated with an increasing population density and subsequent demand for food animals as a source of dietary protein. This has obviously facilitated the increase in the number of farm and market animals and consequently the flow of emerging pathogenic microbes from animals to humans. Infectious diseases know no international boundaries. Increasing regional and international travel can rapidly import emerging or re-emerging infections into Hong Kong and export them to the rest of the world. The huge population density of China and the HKSAR provide an ideal incubator for brewing and spreading new infectious agents and antimicrobial resistance. Although new agents are often discovered in Hong Kong, their source could well be on the mainland in China. Such detection is made possible because of the HKSAR’s relatively better surveillance and laboratory infrastructure. Known infectious diseases are less likely to pose a major threat in the HKSAR because the behaviour of the diseases is already well established and the government is usually prepared. What led to the recent disaster with SARS was that we faced a previously unknown emerging infectious disease. The first lesson to learn is how to face the un-
SARS: A Historical Perspective from Hong Kong
known. The only way to gain the upper hand is for the HKSAR to be prepared and informed before unknown infectious diseases strike. Therefore, the HKSAR public health division should send field officers to join investigative teams in China (and perhaps other parts of the world) in any unusual infectious disease outbreaks so that they gain experience and first-hand information of relevant procedures. Information can be immediately relayed to the HKSAR so that analysis can be carried out and a preventive strategy devised before an epidemic occurs. Such an approach represents a fundamental change of mindset. Gathering intelligence of emerging infectious diseases is our first priority. The essence of such an approach is to begin the process of interception while the disease is still beyond the borders of the HKSAR. Our city is a densely populated part of Southern China, an area which itself has a huge population of humans and animals living in close proximity. This makes the HKSAR a prime target for both emerging infectious diseases and bio-terrorism. Another important element of this approach is the ability to maintain skill and vigilance. This is analogous to the situation of earthquake rescue-teams that operate internationally. Because major earthquakes are quite rare, earthquake rescue-teams from several countries will join in the rescue effort of any country that experiences a major earthquake. Similar co-operation will allow public health teams to have real-life inservice training to put their skills and limitations to the test. In essence, they can learn on the spot, understand the system and relay any relevant information to the HKSAR. The second lesson to learn is the strategic position of the local ‘wet markets’ and the hospitals as epidemic centres (epi-centres) or amplification premises for SARS and perhaps any other emerging contagion or infectious disease. Data from the Guangdong Centre for Disease Control suggest that the initial SARS patients had a strong history of contact with wild game-food animals. The
9
risk of acquisition of SARS also increased with proximity to wet markets. This brings to mind the 1997 outbreak of avian influenza in the HKSAR, which was probably transmitted via the wet markets and therefore the system of running wet markets in the HKSAR has to be thoroughly reviewed. The HKSAR, as a modern cosmopolitan city, must redevelop and redesign the local wet markets into places that are up to date and hygienically clean. Shoppers at local wet markets should no longer be inadvertently picking up zoonotic microbes from animal excreta that can adhere to their hands, clothes and shoes. Standard infection control practices failed in many HKSAR hospitals and resulted in massive outbreaks amongst health-care workers and patients who subsequently spilled back into the community. A single patient going into a housing estate produced a historically unprecedented form of infectious outbreak. Therefore, we must be aware that the hospital setting is a focal point where any future epidemic may land. A culture of meticulous infection control must be ingrained in all health-care workers. This is the third lesson to learn. The Amoy Garden outbreak could have been prevented if hospital infection control had been tighter and the contact tracing process more speedy. A risk assessment should be conducted for all patients admitted to emergency rooms or clinics and a scoring system used to decide on the need for admission to single-room isolation, a multi-bed cohort, or observation in a general ward. Real-time surveillance of staff illnesses, such as fever, respiratory illness, diarrhoea and exanthematous conditions, should be maintained by the obligatory reporting in all hospitals of suspected infections that may or may not require sick leave. This should be the most sensitive parameter for auditing infection control and the early warning signal for a hospital outbreak of a SARS-like illness. The number of beds should be reduced to allow infection control measures to be realistically effec-
10
Severe Acute Respiratory Syndrome
tive. Lectures and practice drills on infection control are largely ignored in the training of junior doctors and nurses. Doctors also seldom fully comply with good hand hygiene and routine isolation procedures. Thus, common antibiotic resistance, such as methicillin-resistant Staphylococcus aureus (MRSA), is rapidly becoming established in hospitals in the HKSAR. MRSA can also spread by contact and occasionally by respiratory droplets, as was the case with SARS. If hospitals cannot control the spread of MRSA, it is unlikely that SARS can be controlled should it return. Rebuilding a culture of compliance towards infection control measures is a priority: it is a matter of life and death. This message needs to be routinely delivered daily to every healthcare worker at the start of each shift. Healthcare workers cannot know what a patient is carrying. The lack of proven antivirals or immunization against SARS demands an almost perfect compliance to infection control procedures by health-care workers or laboratory workers. The fourth lesson to learn is the need for effective communication with both healthcare workers and the public. Doctors who treat their own colleagues during outbreaks that involve health-care workers can experience great emotional distress. What they have to say to the media may no longer have scientific basis and might be sufficiently combustible to undermine the credibility of the government. Similarly, the public start to panic when they see doctors and nurses succumbing to infection outbreaks. Rapid response experts should be sent to any hospital with an outbreak that involves health-care workers so that an objective assessment, sharing of views, discussion with leading medical authorities and joint press conferences can be conducted. This will allow the release of information to the public in an appropriate manner, without compromising the alertness of the public on protective measures. It will minimize the release of unfounded claims and advice, which are likely to hinder control of the
epidemic, raise emotions unnecessarily and undermine the credibility of those in charge. The fifth lesson to learn is that microbes are far more abundant than humans. Their ability to undergo genetic mutation is much greater than the ability of human wisdom to prevent their emergence. Our defence is diligent surveillance and early control so that the damage is minimized. The warning bell was sounded globally with the 1997 avian influenza outbreak, but it was not heeded well enough. The 2003 SARS outbreak is really the urgent wake-up call for us all. The HKSAR, on the border of China, provides an important sentinel-post in guarding the world against emerging infectious diseases.
References 1 Yule WL. A Scottish doctor’s association with the discovery of the plague bacillus. Scott Med J 1995;40: 184–6. 2 Yersin A. La peste bubonique. Ann Inst Pasteur 1894;8: 666. 3 Simond PL. La propagation de la peste. Ann Inst Pasteur 1898;12: 625–87. 4 Liston WG. Plague rats and fleas. J Bombay Nat Hist Soc 1905;16: 253–73. 5 Pollitzer R, Meyer KF. Studies in Disease Ecology. ed. May JF, New York: Halner, 1961: 433–590. 6 Zhong NS, Zheng BJ, Li YM et al. Epidemiological and aetiological studies of patients with severe acute respiratory syndrome (SARS) from Guangdong in February 2003. Lancet (in press). 7 Chantler C, Griffiths S for the SARS expert committee. SARS in Hong Kong: from Experience to Action, Chapter 3. Oct 2003: 14. 8 Reiman HA. An acute infection of the respiratory tract with atypical pneumonia. JAMA 1938;111: 2377–84. 9 Mandell GL, Douglas RG, and Bennett JE, Acute pneumonia. In Mandell GL, Bennett JE, Dolin R, eds. Principles and Practice of Infectious Disease, 5th edn. Philadelphia: Churchill Livingstone. 2000: 731–2. 10 Yuen KY, Chan PK, Peiris M et al. Clinical features and rapid viral diagnosis of human disease associated with avian influenza A H5N1 virus. Lancet 1998;351: 467–71. 11 Tsang KW, Ho PL, Ooi GC et al. A cluster of
SARS: A Historical Perspective from Hong Kong
12
13
14
15 16
17
18
19
20
21
22
23
24
25
26
cases of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348: 1977–85. Centers for Disease Control. Update: Outbreak of severe acute respiratory syndrome – worldwide, 2003. Morb Mortal Wkly Rep 2003; 52: 241–8. Reilley B, Van Herp M, Sermand D et al. SARS and Carlo Urbani. N Engl J Med 2003;348: 1951–2. Hong T, Wang JW, Sun YL et al. Chlamydialike and coronavirus-like agents found in dead cases of atypical pneumonia by electron microscopy. (Article in Chinese) Zhonghua Yi Xue Za Zhi 2003;83: 632–6. Enserink M. SARS in China: China’s missed chance. Science 2003;301: 294–6. Peiris JS, Lai ST, Poon LL et al. Coronavirus as a possible cause of severe acute respiratory syndrome. Lancet 2003;361: 1319–25. Ksiazek TG, Erdman D, Goldsmith C et al. A novel coronavirus associated with severe acute respiratory syndrome. N Engl J Med 2003;348: 1953–66. Drosten C, Gunther S, Preizer W et al. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. N Engl J Med 2003;348: 1967–76. Fouchier RAM, Kuiken T, Schutten M et al. Koch’s postulates fulfilled for SARS virus. Nature 2003; 423: 240. Marra MA, Jones SJ, Astell CR et al. The genome sequence of the SARS associated coronavirus. Science 2003; 300: 1399–404. Rota PA, Oberste MS, Monroe SS et al. Characterization of a novel coronavirus associated with severe acute respiratory syndrome. Science 2003;300: 1394–9. Lee N, Hui D, Wu A et al. A major outbreak of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348: 1986–94. Department of Health Hong Kong: Investigation Report: an outbreak of severe acute respiratory syndrome at Amoy Garden, Kowloon Bay, Hong Kong. Department of Health; Hong Kong Special Administrative Region, April 2003. World Health Organization: First data on stability and resistance of SARS coronavirus compiled by members of the WHO multicenter laboratory network on SARS etiology and diagnosis. http://www.who.int/csr/sars/ survival_2003_05_04/en/index.html Peiris JS, Chu CM, Cheng VC et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;361: 1767–72. Che XY, Hao W, Qiu LW et al. Antibody response of patients with severe acute respiratory syndrome (SARS) to nucleocapsid anti-
27
28
29
30
31
32
33
34
35
36
37
38
39
40
11
gen of SARS-associated coronavirus. Di Yi Jun Yi Da Xue Xue Bao 2003;23: 637–9. Poon LL, Wong OK, Chan KH et al. Rapid diagnosis of a coronavirus associated with severe acute respiratory syndrome (SARS). Clin Chem 2003;49(6 Pt 1): 953–5. Nicholls JM, Poon LL, Lee KC et al. Lung pathology of fatal severe acute respiratory syndrome. Lancet 2003;361: 1773–8. Chan KH, Poon LL, Cheng VC et al. Detection of SARS coronavirus in patients with suspected SARS. Emerg Infect Dis 2004;10: 294–9. Cinatl J, Morgenstern B, Bauer G et al. Glycyrrhizin, an active component of liquorice roots, and replication of SARS-associated coronavirus. Lancet 2003;361: 2045–6. Chu CM, Cheng VC, Hung IF et al. Role of lopinavir/ritonavir in the treatment of SARS: initial virological and clinical findings. Thorax 2004;59: 252–56. Loutfy MR, Blatt LM, Siminovitch KA et al. Interferon alfacon-1 plus corticosteroids in severe acute respiratory syndrome: a preliminary study. JAMA 2003;290: 3222–8. Ning Q, Brown D, Parodo J et al. Ribavirin inhibits viral-induced macrophage production of TNF, IL-1, the procoagulant fgl2 prothrombinase and preserves Th1 cytokine production but inhibits Th2 cytokine response. J Immunol 1998;160: 3487–93. von Grotthuss M, Wyrwicz LS, Rychlewski L. mRNA cap-1 methyltransferase in the SARS genome. Cell 2003;113: 701–2. Zhang R, Guo Z, Lu J et al. Inhibiting severe acute respiratory syndrome-associated coronavirus by small interfering RNA. Chin Med J [Engl] 2003;116: 1262–4. Olsen CW, Corapi WV, Ngichabe CK et al. Monoclonal antibodies to the spike protein of feline infectious peritonitis virus mediate antibody-dependent enhancement of infection of feline macrophages. J Virol 1992;66: 956–65. Kaplikian AZ, Mitchell RH, Chanock RM et al. An epidemiological study of altered clinical reactivity to respiratory syncytial (RS) virus infection in children previously vaccinated with an inactivated RS virus vaccine. Am J Epidemiol 1969; 89: 405–21. Frey HM, Krugman S. Atypical measles syndrome: unusual hepatic, pulmonary, and immunologic aspects. Am J Med Sci 1981;281: 51–5. Guan Y, Zheng BJ, He YQ. Isolation and characterization of viruses related to the SARS coronavirus from animals in southern China. Science. 2003;302: 276–8. Epub 2003 Sep 04. Guan Y, Shortridge KF, Krauss S et al. Molecular characterization of H9N2 influenza virus-
12
Severe Acute Respiratory Syndrome
es: were they the donors of the ‘internal’ genes of H5N1 viruses in Hong Kong? Proc Natl Acad Sci U S A 1999;96: 9363–7. 41 Crosby AW. Epidemic and Peace, 1918, part IV. Westport, CT: Greenwood Press, 1976. 42 Smith W, Andrewes CH, Laidlaw PP. A virus obtained from influenza patients. Lancet 1933;2: 66–8.
43 Stuart-Harris CH, Schild GC, Oxford JS et al. Influenza. The Viruses and the Disease, 2nd edn. Victoria, Can.: Edward Arnold, 1985: 118–38. 44 Noble GR. Epidemiological and clinical aspects of influenza. In: Beare AS, ed. Basic and Applied Influenza Research. Boca Raton, FL: CRC, 1982: 11–50.
Chapter 2
SARS: A Global Perspective David L Heymann
On 12 March 2003, the World Health Organization (WHO) alerted the world to the appearance of a severe respiratory illness of undetermined cause that had rapidly infected more than 40 staff at hospitals in Vietnam and Hong Kong.1 The alert also referred to two other events that raised the level of alarm: an outbreak of 305 cases, with 5 deaths, of atypical pneumonia reported in mid-February from the southern Chinese province of Guangdong, and an almost simultaneous report from Hong Kong of two confirmed cases of avian influenza A H5N1 in family members with a recent travel history to southern China. The alert described the signs and symptoms of the unidentified illness and recommended that suspected cases be isolated, managed with barrier nursing techniques, and reported simple measures that would provide the cornerstone for containing the outbreak as it spread within, and then outside, Asia. Prior to that alert, several international mechanisms for routine outbreak detection, investigation, and response had already begun to operate with a heightened sense of urgency. A new and potentially pandemic strain of the influenza virus was the first and most greatly feared suspected cause. Laboratories in the WHO Global Influenza Surveillance Network had been on alert since late November 2002, when the Global Public Health Intelligence Network (GPHIN), an electronic intelligence gathering tool, managed by the Ministry of Health of Canada, picked up rumours of
severe ‘flu-like’ outbreaks in Guangdong and Beijing.2 Studies conducted by Chinese scientists and confirmed by a network of influenza laboratories identified strains of influenza B virus as the cause, and concern eased. It mounted to new heights with the mid-February confirmation of avian influenza in Hong Kong, prompting WHO to activate its influenza pandemic preparedness plans.3 To learn more about the outbreak in Guangdong, a team of experts, drawn from the WHO Global Outbreak Alert and Response Network (GOARN), arrived in Beijing on 23 February, but was not granted permission to travel further. A second GOARN team began an emergency investigation in Hanoi on 28 February, 2 days after the first case of atypical pneumonia was admitted to hospital, and established infection control procedures and an isolation ward. Laboratories in the influenza network analysed specimens from this patient and other early cases, and conclusively ruled out influenza viruses as the cause. They also ruled out all other known causes of respiratory illness. With a new disease increasingly suspected, WHO began daily teleconferences linking its country and regional offices and response teams with headquarters operational staff. These mechanisms, too, would prove decisive in tracking the outbreak, gathering the knowledge for recommending effective control measures, and getting support teams to countries requesting assistance. By 15 March, WHO had received reports
13
14
Severe Acute Respiratory Syndrome
of more than 150 new cases of atypical pneumonia of unidentified cause concentrated in the hospitals of six Asian countries and Canada.4 The disease did not respond to antibiotics and antivirals known to be effective against primary atypical pneumonia and other respiratory infections. No patients, including young and previously healthy health workers, had recovered, many were in critical condition, several required mechanical ventilatory support, and four had died. Equally alarming, the disease was spreading rapidly along the routes of international air travel. The potential for further international spread was vividly demonstrated that same day when a medical doctor, who had treated the first cases of atypical pneumonia in Singapore, reported similar symptoms shortly before boarding a flight from New York to Singapore. The airline was alerted and the doctor and his wife disembarked in Frankfurt for immediate hospitalization, becoming the first cases in Europe.5 Faced with these events, WHO issued a second and stronger global alert on 15 March, this time in the form of an emergency travel advisory.4 The alert provided guidance for travellers, airlines and crew, set out a case definition, and gave the new disease its name: severe acute respiratory syndrome (SARS). It also launched a co-ordinated global outbreak response that tested a critical assumption: that rapid and intense public health action could stop a new transmissible disease, of unidentified cause and unknown epidemic potential, from becoming endemic. On 5 July, the last known probable case of SARS completed a 20-day period of isolation, and WHO declared that the international outbreak had been contained.6 While this achievement demonstrates the strength of classical public health measures — case detection, isolation, contact tracing and infection control — it also shows the importance of several new mechanisms, set up at the international level, to improve global capacity to detect and
respond to outbreaks of emerging and epidemic-prone diseases.
International mechanisms put to the test The international response to SARS was the roll-out of a series of mechanisms for outbreak detection and containment that had been under development since 1997.7 These mechanisms were set up to help the public health community deal more effectively with outbreaks of diseases ranging from meningococcal meningitis, cholera, plague and yellow fever to more exotic infections such as Ebola and Marburg. They were the outgrowth of concern about international capacity to detect and contain emerging and epidemic-prone diseases that arose following outbreaks in the early 1990s of cholera in Latin America (1991), pneumonic plague in India (1994), and Ebola haemorrhagic fever in the Democratic Republic of the Congo (1995).8–10 While all these outbreaks caused concern throughout the world, with serious economic consequences and disruptions in travel and trade, it was the highly publicized Ebola outbreak that pointed most urgently to the need for change. That outbreak, which caught the international community by surprise, signalled the need for stronger infectious disease surveillance and control worldwide, for improved international preparedness to provide support when similar outbreaks occur, and for accommodating the needs of the press in providing valid information. A need for more broad-based international health regulations and electronic information systems connecting WHO with its regional and country offices also became evident, as did the realization that timely and adequate outbreak detection and response would need support from a broad coalition of partners.11 Potential partners in global surveillance and outbreak response were first brought together informally by WHO in 1997, and
SARS: A Global Perspective
then formally launched as the GOARN partnership in 2000.12 Electronic communication networks and new computer applications were developed to enhance the network’s power in global surveillance and response.13 The GPHIN intelligencegathering tool, developed and maintained for WHO by Health Canada, was set up as a web-crawling system programmed to search for key words suggesting an outbreak in news sources and electronic discussion groups worldwide. This innovation brought major improvements in the speed of outbreak detection compared to traditional systems, where an alert sounds only after case reports at the local level progressively filter to the national level and are then reported to WHO. WHO also began the task of updating a set of international regulations that had been agreed by its member countries in 1969 to limit the international spread of infectious diseases and provide a mandate for global surveillance and response.14 All of these mechanisms were put to the test during the SARS outbreak, and helped support a co-ordinated and effective global response.
New ways of working internationally The SARS outbreak was the first internationally spreading outbreak during which regularly updated evidence-based recommendations for patient management and outbreak control could be collectively made in real time as events unfolded around the world. As the outbreak evolved, some of the world’s most experienced laboratory experts, clinicians and epidemiologists worked together in virtual networks, taking advantage of up-to-date communication technologies, including the internet, secure websites, and video and telephone conferencing. Laboratories in the existing influenza surveillance network formed the basis for a new virtual network to identify the causative agent, which was achieved within
15
a month, and to develop diagnostic tests.15 Clinicians and field epidemiologists constituted other virtual networks, and by the end of the outbreak more than 150 experts from institutions in 17 countries had demonstrated how close collaboration and sharing of information, despite strong academic pressure to publish information in scientific journals, could serve the public health good. No estimates are available for the number of health staff who risked their lives in caring for patients, though the deaths of many have been documented. The SARS outbreak also marked the first occasion where sufficient information became available rapidly enough to issue evidence-based international travel recommendations as a measure for preventing further international spread, particularly by air travel. As real-time evidence accumulated, further international spread was attributed to persons with SARS who continued to travel internationally by air, in some cases infecting passengers and crew during the flight.16 Daily tracking of cases also revealed that contacts of SARS patients continued to travel, becoming ill upon arrival at their destination. WHO therefore issued recommendations on 27 March that countries with major outbreaks screen departing passengers for fever and other signs of SARS, or known contact with SARS patients.17 The choice of measures for putting this recommendation into effect was left to the discretion of individual countries. Some set up screening measures at international airports and border crossings with a variety of requirements, including a health declaration by each departing passenger, temperature monitoring of each passenger and a stop list of contacts of SARS patients at immigration by which known contacts were asked not to travel. As the outbreak progressed, contacts of probable SARS patients continued to travel and become ill after arrival at their destination, indicating the continuing risk of further international spread. Real-time
16
Severe Acute Respiratory Syndrome
Table 2.1 Summary of probable SARS cases with onset of illness from 1 November 2002 to 31 July 2003 (based on data as of 31 December 2003). Source: http://www.who.int/csr/sars/country/ table2004_04_21/en/ Areas
Australia Canada China China, Hong Kong Special Administrative Region China, Macao Special Administrative Region China, Taiwan France Germany India Indonesia Italy Kuwait Malaysia Mongolia New Zealand Philippines Republic of Ireland Republic of Korea Romania Russian Federation Singapore South Africa Spain Sweden Switzerland Thailand United Kingdom United States Vietnam Total a
Cumulative number of cases Female
Male
Total
4 151 2674 977
2 100 2607 778
6 251 5327b 1755
0
1
1
218 1 4 0 0 1 1 1 8 1 8 0 0 0 0 161 0 0 3 0 5 2 13 39
128 6 5 3 2 3 0 4 1 0 6 1 3 1 1 77 1 1 2 1 4 2 14 24
346c 7 9 3 2 4 1 5 9 1 14 1 3 1 1 238 1 1 5 1 9 4 27 63 8096
Median age (range)
Number of deathsa
15 (1-45) 49 (1-98) Not available 40 (0-100)
0 43 349 299
28
42 (0-93) 49 (26-61) 44 (4-73) 25 (25-30) 56 (47-65) 30.5 (25-54) 50 30 (26-84) 32 (17-63) 67 41 (29-73) 56 40 (20-80) 52 25 35 (1-90) 62 33 43 (33-55) 35 42 (2-79) 59 (28-74) 36 (0-83) 43 (20-76)
0
37 1 0 0 0 0 0 2 0 0 2 0 0 0 0 33 1 0 0 0 2 0 0 5 774
Includes only cases whose death is attributed to SARS. Case classification by sex is unknown for 46 cases. c Since 11 July 2003, 325 cases have been discarded in Taiwan, China. Laboratory information was insufficient or incomplete for 135 discarded cases, of which 101 died. d Includes health-care workers who acquired illness in other areas. e Due to differences in case definitions, the United States has reported probable cases of SARS with onsets of illness after 5 July 2003. f SARS cases arising after the cessation of the initial outbreak (i.e laboratory-associated cases and cases in Guandong in December 2003) are not incuded. b
SARS: A Global Perspective
Case fatality ratio (%)
0 17 7 17
17
Number of imported cases (%)
Number of health-care workers affected (%)
Date onset first probable case
Date onset last probable casef
6 (100) 5 (2) Not applicable Not applicable
0 (0) 109 (43) 1002 (19) 386 (22)
26-Feb-03 23-Feb-03 16-Nov-02 15-Feb-03
1-Apr-03 12-Jun-03 3-Jun-03 31-May-03
5-May-03
5-May-03
25-Feb-03 21-Mar-03 9-Mar-03 25-Apr-03 6-Apr-03 12-Mar-03 9-Apr-03 14-Mar-03 31-Mar-03 20-Apr-03 25-Feb-03 27-Feb-03 25-Apr-03 19-Mar-03 5-May-03 25-Feb-03 3-Apr-03 26-Mar-03 28-Mar-03 9-Mar-03 11-Mar-03 1-Mar-03 24-Feb-03 23-Feb-03
15-Jun-03 3-May-03 6-May-03 6-May-03 17-Apr-03 20-Apr-03 9-Apr-03 22-Apr-03 6-May-03 20-Apr-03 5-May-03 27-Feb-03 10-May-03 19-Mar-03 5-May-03 5-May-03 3-Apr-03 26-Mar-03 23-Apr-03 9-Mar-03 27-May-03 1-Apr-03 13-Jul-03e 14-Apr-03
0
1 (100)
11 14 0 0 0 0 0 40 0 0 14 0 0 0 0 14 100 0 0 0 22 0 0 8 9.6
21 (6) 7 (100) 9 (100) 3 (100) 2 (100) 4 (100) 1 (100) 5 (100) 8 (89) 1 (100) 7 (50) 1 (100) 3 (100) 1 (100) Not available 8 (3) 1 (100) 1 (100) 5 (100) 1 (100) 9 (100) 4 (100) 27 (100) 1 (2) 142
0 (0)
68 (20) 2 (29)d 1 (11) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 4 (29) 0 (0) 0 (0) 0 (0) 0 (0) 97 (41) 0 (0) 0 (0) 0 (0) 0 (0) 1 (11)d 0 (0) 0 (0) 36 (57) 1706
18
Severe Acute Respiratory Syndrome
information demonstrated that contact tracing at some sites did not identify chains of transmission fully and that transmission was occurring outside confined settings, such as the health-care environment, possibly placing the general population at risk. In late March, an outbreak of 329 almost simultaneous probable cases among residents of a housing estate in Hong Kong suggested possible transmission by exposure to some factor in the environment, thus creating further opportunities for exposure in the general population.18 Additional evidence-based guidance was therefore made for the sites where contact tracing could not link all cases, understanding that if the disease were spreading in the wider community it would greatly increase the risk to travellers and the likelihood that cases would be exported to other countries. This guidance was aimed at international travellers, and recommended that they postpone all but essential travel to designated sites in order to minimize their risk of becoming infected.19 The global alerts issued by WHO on 12 and 15 March provided a clear line of demarcation between areas with severe SARS outbreaks and those without. Following the SARS alerts, all areas with imported cases, with the exception of Taiwan, either prevented any further transmission or kept the number of locally transmitted cases very low.20 Likewise, the travel recommendations issued by WHO appear to have been effective in helping to contain international spread of SARS. Of the 40 international flights known to have carried 37 probable SARS cases, current analysis has implicated 5 in transmission to passengers or crew.16, 21 Following the 27 March recommendations for exit screening, no confirmed SARS case associated with in-flight exposure was reported to WHO. This may have been because awareness of screening procedures discouraged persons with fever from attempting to travel.21 Initial information from Hong Kong reveals that two probable SARS cases were
identified by airport screening procedures, immediately hospitalized and prevented from international travel (Hong Kong International Airport, personal communication). Travel recommendations also appear to have provided a benchmark for gauging the safety of international travel; when an area was declared safe from the risk of SARS transmission, traveller confidence was regained. Recommendations concerning travel were ended when epidemiological criteria indicating a low risk to travellers were met. That goal in itself became a motivation for governments and populations to collaborate in bringing the outbreaks under control. Many countries also set a second goal of removal from the list of areas with recent local transmission. The determination to attain this objective may have contributed to the speed with which the cycle of human-to-human transmission was broken globally, and confidence was restored.22 Passenger movement figures provided by the Hong Kong International Airport show a rapid rebound from the lowest number of passengers, 14 670, recorded just before 23 May when the travel recommendations for Hong Kong were removed, to 54 195 on 12 July, a month and a half later (Hong Kong International Airport, personal communication).
The costs of a new disease SARS demonstrated the speed with which a new disease can travel in a highly mobile society. It also demonstrated the farreaching consequences that an emerging disease — especially when severe and readily transmissible — can have in a closely interconnected and interdependent world. The economic impact of the outbreak was considerable: apart from the direct costs of intensive medical care and control interventions, SARS caused widespread social disruption and economic loss. Schools, hospitals, and some borders were closed and thousands of people were placed in quarantine. International travel to affected areas
SARS: A Global Perspective
decreased by 50% to 70% and hotel occupancy dropped by more than 60%.23,24 Businesses, particularly in tourism-related areas, failed, while some large manufacturing facilities were forced to suspend operations when cases appeared among workers. Preliminary estimates have placed the cost of the outbreak at nearly US$ 100 billion, mainly as a result of cancelled travel and decreased investment in Asia alone.25 In evaluating the global significance of the SARS outbreak, these consequences need to be viewed against the much greater consequences the world would now face if containment efforts had failed. Several investigations and assessments of the outbreak have reached a similar conclusion: SARS convinced even the most sceptical government leaders that health problems can have profound social, economic, and political consequences.18,26,27 The global mechanisms for outbreak surveillance and response, severely tested by SARS, performed well, but can never compensate for inadequacies in local surveillance, detection, and reporting capacities. Like the Ebola outbreaks that prompted strengthened international capacity, the SARS outbreak smouldered for two months, undetected and then unreported, in its original emergence zone, gathering the epidemic potential to spread internationally with enormous costs and broad social consequences.18,28,29 If the SARS experience results in greater investment in strengthening local capacities, then public health the world over will benefit. Table 2.1 summarizes the occurrence of SARS cases worldwide.
3
4
5
6
7
8
9 10
11
12
13
14
References 1 WHO issues a global alert about cases of atypical pneumonia: cases of severe respiratory illness may spread to hospital staff. World Health Organization press release 12 March 2003. Available from: URL: http://www.who.int/ csr/sars/archive/2003_03_12/en/, accessed 18 December 2003. 2 SARS — Chronology of events. Ottawa: Health
15
19
Canada, Population and Public Health Branch, 2003. Influenza A(H5N1), Hong Kong Special Administrative Region of China. Wkly Epidemiol Rec 2003;78: 49–50. World Health Organization issues emergency travel advisory. World Health Organization situation update 15 March 2003. Available from: URL: http://www.who.int/csr/sars/archive/ 2003_03_15/en/, accessed 18 December 2003. SARS: lessons from a new disease. In: The World Health Report 2003: Shaping the Future. Geneva: World Health Organization, 2003. Taiwan, China: SARS transmission interrupted in last outbreak area. World Health Organization SARS situation update 96 5 July 2003. Available from: URL: http://www.who.int/csr/ don/2003_07_05/en/, accessed 18 December 2003. WHO Division of Emerging and Other Communicable Diseases Surveillance and Control Annual Report 1997. Geneva: World Health Organization, 1998 (unpublished document WHO/ EMC/98.2). Tauxe RV, Mintz ED, Quick RE. Epidemic cholera in the New World: translating field epidemiology into new prevention strategies. Emerg Infect Dis 1995;1: 141–6. Plague — international team of experts, India. Wkly Epidemiol Rec 1994;69: 321–2. Khan AS, Tshioko FK, Heymann DL et al. The re-emergence of Ebola hemorrhagic fever, Democratic Republic of the Congo, 1995. J Infect Dis 1999;179(Suppl 1): S76–86. Heymann DL, Barakamfitiye D, Szezeniowski M et al. Ebola hemorrhagic fever: lessons from Kikwit, Democratic Republic of the Congo. J Infect Dis 1999;179(Suppl 1): S283–86. A framework for global outbreak alert and response. Geneva: World Health Organization, 2000 (unpublished document WHO/CDS/ CSR/2000.2). Heymann DL, Rodier GR. WHO Operational Support Team to the Global Outbreak Alert and Response Network. Hot spots in a wired world: WHO surveillance of emerging and reemerging infectious diseases. Lancet Infect Dis 2001;1: 345–53. Global crisis — global solutions. Managing public health emergencies of international concern through the revised International Health Regulations. Geneva: World Health Organization, 2002 (unpublished document WHO/CDS/ CSR/AR/2002.4). World Health Organization Multicentre Collaborative Network for Severe Acute Respiratory Syndrome (SARS) Diagnosis. A multicentre collaboration to investigate the cause of severe acute respiratory syndrome. Lancet 2003;361: 1730–3.
20
Severe Acute Respiratory Syndrome
16 Consensus document on the epidemiology of severe acute respiratory syndrome (SARS). Geneva: World Health Organization, 2003 (unpublished document WHO/CDS/CSR/GAR/ 2003.11). Available from: URL: http://www. who.int/csr/sars/en/WHOconsensus.pdf, accessed 18 December 2003. 17 Severe acute respiratory syndrome, update 11–WHO recommends new measures to prevent travel-related spread of SARS. Available from: URL: http://www.who.int/csr/sars/archive/ 2003_03_27/en/print.html, accessed 18 December 2003. 18 SARS in Hong Kong: From experience to action. Hong Kong SAR: SARS Expert Committee, 2003. Available from: URL: http://www.sarsexpertcom.gov.hk/english/reports/reports.ht ml, accessed 18 December 2003. 19 Severe acute respiratory syndrome, update 92–chronology of travel recommendations, areas with local transmission. Available from: URL: http://www.who.int/csr/don/2003_07_01/e n/, accessed 18 December 2003. 20 Summary of probable SARS cases with onset of illness from 1 November 2002 to 31 July 2003 (revised 26 September 2003). Geneva: World Health Organization 2003. Available from: URL: http://www.who.int/csr/sars/country/ table2003_09_23/en/, accessed 18 December 2003. 21 Olsen SJ, Chang H-L, Cheung TY-Y et al. Transmission of the severe acute respiratory syndrome on aircraft. N Engl J Med 2003;349: 25 2416–22. 22 Severe acute respiratory syndrome (SARS): report by the secretariat. Geneva: World Health Organization, 2003 (Executive Board document EB113/33, 27 November 2003). Available from: URL: http://www.who.int/gb/EB_ WHA/PDF/EB113/eeb11333.pdf, accessed 19 December 2003.
23 Bonte-Friedheim R. Sars: Thinking Ahead — Epidemiological and Economic Scenarios. London: Citigroup, 2003. 24 Special SARS analysis: impact on travel and tourism (Hong Kong, China, Singapore, Vietnam). World Travel and Tourism Council, 2003. Available from: URL: http://www.wttc.org/ measure/PDF/Hong%20Kong%20SARS.pdf (Hong Kong); http://www.wttc.org/measure/ PDF/China%20SARS.pdf (China); http:// www.wttc.org/measure/PDF/Singapore%20S ARS.pdf (Singapore); http://www.wttc.org/ measure/PDF/Vietnam%20SARS.pdf (Vietnam) (accessed 18 December 2003). 25 Knight J. How and who does SARS kill? In SARS: what have we learned? Nature 2003; 424 (6945): 121–6. Available from: URL: http:// www.nature.com/nature/focus/sars/sars2.ht ml#who, accessed 18 December 2003. 26 National Intelligence Council. SARS: Down but Still a Threat. Washington DC: National Intelligence Council, 2003. Available from: URL: http://www.fas.org/irp/nic/sars.pdf, accessed on 28 December 2003. 27 Learning from SARS: renewal of public health in Canada. A report of the National Advisory Committee on SARS and public health, October 2003. Ottawa: Health Canada, 2003. 28 Visit of WHO expert team to review the outbreak of atypical pneumonia in Guangdong Province, 24 March–9 April 2003. Final report, 30 April 2003. Geneva: World Health Organization, 2003 (unpublished document). 29 Zhong NS, Zheng BJ, Li YMH et al. Epidemiology and cause of severe acute respiratory syndrome (SARS) in Guangdong, People’s Republic of China, in February, 2003. Lancet 2003;362: 1353–8.
Chapter 3
Clinical Presentation of the Disease in Adults JY Sung and KY Yuen
SARS is the first severe and readily transmissible disease that has emerged in the twenty-first century. SARS has the unusual potential to spread quickly in hospitals and clinics with devastating effects on healthcare workers. The rapid rise of international air travel in recent decades enabled the infection to spread quickly across the continents and become an international threat. At the outset of the SARS epidemics in March 2003, SARS was described as a form of atypical pneumonia. The aetiological agent was unknown and as a diagnostic test was not available, clinical presentation and contact history became the only diagnostic tools at that time. The WHO1 and CDC2 issued their case definitions based purely on clinical features and contact history.
Prodromal symptoms The incubation period of SARS ranges from 1 to 14 days but is on average around 4 days. The illness starts with non-specific systemic symptoms such as fever, myalgia, chills and rigor, non-productive cough, headache and dizziness. The typical presentations of SARS described in case series are summarized in Table 3.1.3–7 Fever >38∞C was included in the initial case definition of SARS and hence most of the patients reported in the early series had high temperature. Yet, not all patients present with a high fever, and in elderly patients, for example, fever may not be a prominent symptom. A high swinging fever is often associated with chills and rigor
and patients often feel very tired with pain and aches of muscles. In some cases, fever subsides spontaneously at around day 4 to day 7 but this does not indicate resolution of symptoms. Resurgence of temperature and deterioration of symptoms often occur during the second week.3 Other non-specific symptoms such as dizziness, headache and malaise are also common among patients suffering from SARS. Severe dizziness has been reported in young and previously healthy subjects and some of them even collapse whilst struggling to get out of bed.4 This is probably related to the hypotension found in these patients. Many patients develop dry cough during the early phase of the disease. Sore throat and coryza are also uncommon in this condition. In this early phase of the disease, patients usually have a clear breath sound on auscultation. The lung fields on radiography are often clear. Depending on their time of presentation, up to 80% of patients had a normal chest radiograph on first encounter. Thus an initially normal chest radiograph does not exclude the diagnosis of SARS and followup imaging is usually necessary.
Respiratory manifestations Pulmonary illness is the predominant manifestation of SARS. Dry cough was the presenting symptom of SARS in 60–85% of cases. Patients usually start to have shortness of breath while coughing. Auscultation of the chest often reveals inspiratory
21
22
Severe Acute Respiratory Syndrome
Table 3.1 Clinical presentations of SARS in large case series Clinical features (%)
Hong Kong Peiris et al.3 (n = 50)
Hong Kong Lee et al.4 (n = 138)
Guangzhou Wu et al.5 (n = 96)
Singapore Hsu et al.6 (n = 20)
Toronto Booth et al.7 (n = 144)
Fever Chills/rigor Myalgia Cough Dyspnoea Headache Dizziness Sputum Diarrhoea Nausea & vomiting Sore throat Malaise
100 74 54 62 20 20 12 – 10 20
100 73.2 60.9 57.3 – 55.8 42.8 29.0 19.6 19.6
100 55.2 21.9 85.4 – 39.6 – 66.7 – –
100 15 45 75 40 20 – – 25 35
99.3 27.8 49.3 69.4 41.7 35.4 4.2 4.9 23.6 19.4
20 50
23.2 –
– 35.4
25 45
12.5 31.2
Figure 3.1 Axial thin section computed tomography of the thorax in a 36-year-old male patient performed 4 days after onset of fever. This image resembles those of BOOP.
crackles at the lung bases but wheezing is usually absent. Towards the end of the first week or at the start of the second week, the pulmonary signs and symptoms begin to deteriorate. Dyspnoea increases and limits physical activity of patients. Oxygen saturation of blood decreases as disease progresses. Airspace consolidation is unilateral and focal at the early phase of the disease but soon becomes multifocal and more extensive in the second week of the illness. Although all lung segments could be involved, there is a predilection towards
lower lobes of the lungs.8 Occasionally, pulmonary infiltrates can be detected shifting from one area to another within one or two days. These shifting radiographic shadows coinciding with reduction in viral load may suggest an immune-related damage, rather than direct viral cytolysis. High resolution CT scan of the thorax shows features resembling bronchiolitis obliterans organizing pneumonia (BOOP)9 (Fig. 3.1), an immunemediated disease which is responsive to corticosteroid therapy. The lesions mainly showed ground glass opacification (Fig. 3.2)
Clinical Presentation of the Disease in Adults
Figure 3.2 Axial thin section computed tomography in a 34-year-old female patient performed 3 days after onset of fever. This magnified image of the right lower lobe demonstrates an area of ground glass opacification, within which the pulmonary vasculature remains visible.
or a mixed ground glass and consolidative opacification. Other findings included thickening of inter-lobular septa and intralobular interstitium. These were only seen superimposed to a ground glass opacification and resulted in a ‘crazy paving’ pattern. Around 20–25% of patients eventually progressed into severe respiratory failure and acute respiratory distress syndrome (ARDS) that necessitated ICU care. Mechanical ventilation will be required when high flow supplementary oxygen cannot maintain their oxygen saturation. Those who required mechanical ventilation have a high mortality. Lower Acute Physiology and Chronic Health Evaluation (APCHE) II scores and higher baseline ratios of PaO2 to fraction of inspired oxygen are associated with earlier recovery. Advanced age, preexisting cardio-pulmonary diseases, bilateral pulmonary infiltrates on presentation, high neutrophil counts, elevated serum creatinine kinase (CPK) and lactate dehydrogenase (LDH) levels are associated with poor clinical outcome. About half of those who required ICU care died in the SARS outbreak. The causes of death include severe ARDS, multiorgan failure, superimposed infection and septicaemia, thromboembolic
23
complications. Postmortem examination showed bronchial epithelial denudation, loss of cilia, squamous metaplasia, giantcell infiltrate and increase in macrophages in the alveoli and interstitium of the lung (see Chapter 9). Alveolar pneumocytes showed cytomegaly with granular amphophilic cytoplasm. Electron microscopy revealed SARS CoV viral particles in the cytoplasm of epithelial cells. Haemophagocytosis can be observed in some patients. Pneumothorax and pneumomediastinum have been frequently reported in severely ill SARS cases.3 These can develop either spontaneously or in association with the use of mechanical ventilation. In one report, 12% of seriously ill SARS patients developed spontaneous pneumomediastinum. Among patients nursed in the ICU, 20% developed pneumothorax or pneumo-mediastinum. The incidence of barotrauma was unusually high despite low-volume low-pressure mechanical ventilation. The reason for this phenomenon is unclear. In early onset SARS, there is pulmonary oedema with hyaline membrane formation and cellular exudates in the airspaces, suggesting that reduction in lung compliance may be partially responsible for the high incidence of barotrauma.
Enteric manifestations Apart from respiratory symptoms, diarrhoea is the most common and important manifestation of SARS. Twenty percent of patients had diarrhoea on presentation,4 and up to 70% patients had diarrhoeal symptoms during the course of illness.3 This diarrhoea is usually a high volume watery stool with no mucus or blood. The profound water and electrolyte loss can lead to volume depletion and electrolyte disturbance in severe cases. In some patients, diarrhoea and fever are the only initial manifestations of SARS in the absence of pneumonia on chest radiograph. In others, diarrhoea starts in the second week of the illness and coincides with
24
Severe Acute Respiratory Syndrome
recurrence of fever and progression of pulmonary disease. Fortunately, the diarrhoea is usually self-limiting. There is, so far, no documented case of mortality related to diarrhoea. Intestinal biopsies obtained by colonoscopy or during autopsy showed minimal inflammation or architectural disruption.10 However, ultrastructural studies showed the presence of viral particles (60–90 nm in size) within both small and large intestinal cells. Viral particles are confined to the epithelial cells, primarily in the apical surface enterocytes and rarely in the glandular epithelial cells. Intracellularly, viral particles are contained within dilated cytoplasmic vesicles consistent with dilated endoplasmic reticulum. The vesicles containing the viral particles are often seen towards the apical cytoplasm. Clusters of coronavirus can also be found on the surface microvilli which suggests virus leaving from the luminal surface of enterocytes. There was no evidence of villous atrophy despite viral adhesion and colonization (Fig. 3.3). These findings support the intestinal tropism of SARS coronavirus. With viral replication in the intestinal epithelium, it is not difficult to envisage the passage of virus into patients’ faeces. During the second week of the illness, virus can be found by RT-PCR in stool of almost all patients.3 The virus can be detected by PCR in stool for up to 2 months from symptom onset in a patient.10 The long carriage of virus in patients’ faeces has substantial implications on transmission of infection and infection control. Though diarrhoea is a common manifestation of SARS, many questions remain unanswered. It remains elusive whether the route of transmission (respiratory versus faecal–oral) is important in determining the gastrointestinal manifestation of SARS. Moreover, further studies are necessary to characterize the mechanism underlying the diarrhoea related to SARS-coronavirus infection.
Haematological manifestations Haematological features of SARS offer important clues to the diagnosis of the disease. Lymphopenia (absolute lymphocyte count <1000/mm3) is found in almost all patients with SARS.11 In most cases, the lymphocyte count reaches its lowest point in the second week of the illness. In fact, the absence of progressive lymphopenia would render a diagnosis of SARS quite unlikely. In the majority of cases, the lymphocyte count starts to climb in the third week of illness. The recovery from lymphopenia coincides with clinical improvement. However, 30% of patients remain lymphopenic at the fifth week of SARS. Differential counting shows that B lymphocytes are unaffected. On the contrary, both CD4 and CD8 T-lymphocyte counts drop early in the course of illness. Since both cell lineages are affected, the CD4/CD8 ratio remains normal. The lower counts of CD4 and CD8 cells actually correlate with adverse clinical outcomes of ICU admission and/or death.11 Leucocytosis, largely related to neutrophilia, is also commonly found in patients with SARS.11 Neutrophilia might be related to treatment with corticosteroids but some patients have neutrophilia on presentation. In patients with neutrophilia, full sepsis work-up needs to be performed and empirical broad-spectrum antibiotics should be considered. The other possible reason for neutrophilia is tissue necrosis in those with extensive pulmonary injury. A high neutrophil count on presentation has been associated with poor clinical outcome of the patients. Thrombocytopenia is found in more than half of patients with SARS and may be profoundly low.11 Unlike haemorrhagic fevers, bleeding diathesis due to low platelet count is extremely unusual. In the majority of patients, platelet count recovered in the recovery phase of the illness, followed by reactive thrombocytosis in some patients. Thrombocytopenia may be caused by immune mechanisms or by direct effect of
Clinical Presentation of the Disease in Adults
25
Figure 3.3 Biopsy of the colon in a SARS patients with diarrhoea. (Reproduced from Gastroenterology 2003; Leung WK, To KF, Chan PKS, Chan HLY, Wu A, Lee N, Yuen KY, Sung JJY, Enteric involvement of Severe Acute Respiratory Syndrome Coronavirus Infection; pages 1011–17, with permission from the American Gastroenterological Association.)
viruses on megakaryocytes or platelets. Megakaryocytes are capable of harbouring a variety of viruses, although the mechanism of viral entry into megakaryocytes is not well understood. Dysmorphic megakaryocytes containing inclusion bodies, vacuoles, degenerating nuclei or naked nuclei may be seen in infected marrow. Occasionally, thrombocytopenia occurs as a result of disseminated intravascular coagulopathy
(DIC). Petechiae and ecchymosis can be found.12 This complication is, however, the exception rather than the rule. Paradoxically, venous thrombosis has been reported quite frequently in some patient series. This could be manifested as deep venous thrombosis of the legs which was found in around a third of the cases reported in Singapore.6 In their experience, thromboembolism was found in a substan-
26
Severe Acute Respiratory Syndrome
tial proportion of postmortem cases. The clinical significance of pulmonary thromboembolism and its contribution to respiratory failure are not clear. One should be aware of this condition especially in those who had rapidly developing hypoxaemia in the absence of pulmonary infiltrates on the chest radiograph.
Hepatic manifestations Approximately a quarter of patients had elevated ALT on admission, and over 70% of patients had ALT elevation during the course of illness.13 In the majority of patients, ALT levels start to rise toward the end of first week and peak at the end of second week. High peak ALT level appears to be an independent predictor of more severe illness and worse clinical outcome. Most patients, however, had transient elevation of ALT, which normalize spontaneously with the recovery of SARS. The underlying cause for ALT elevation is uncertain. In the model of mouse hepatitis virus, a coronavirus related to SARS CoV, disturbance in microcirculation associated with sinusoidal microthrombi appeared to play a role in the pathogenesis causing granulomatous hepatitis and hepatocellular necrosis. However, no coronavirus inclusion bodies or necroinflammatory changes typical of viral hepatitis could be identified in the liver at autopsy, implying that SARS CoV-induced viral hepatitis is unlikely to be the cause of liver dysfunction in these patients. Although SARS Co-V can be identified in the liver of one patient on autopsy,13 there was no accompanying evidence of hepatitis. It therefore appears that the elevation of liver enzyme is a reactive response towards SARS CoV infection. The hepatic acute phase response involving cytokine release from inflammatory cells is a defence reaction of the body against the causative agent to protect the vital functions of the liver. This is usually a transient reaction and therefore the majority of patients have ALT levels returned to normal after recovery.
In one study, it has been suggested that co-infection with hepatitis B virus is associated with more severe respiratory disease.3 Subsequent studies failed to confirm any difference in clinical outcomes among chronic hepatitis B patients as compared to the HBsAg-negative patients.13 It is possible that those patients who suffered from severe flare-up of hepatitis or decompensated liver cirrhosis might have a higher risk of mortality. The absence of specific hepatic lesions in hepatitis B virus and SARS CoV co-infected patients on autopsy further supported the absence of influence of SARS on chronic hepatitis B infection.13,14 Yet, HBV infected patients with elevated transaminases should be prescribed a nucleoside analogue such as lamivudine to prevent development of hepatic decompensation.
Cardiovascular manifestations Symptoms related to the cardiovascular system are uncommon. In a series from Hong Kong, some 50% of patients experienced significant hypotension (defined by systolic blood pressure <100 mmHg ± diastolic blood pressure <50 mmHg) during hospitalization. The low blood pressure may account for the dizziness in many patients. Persistent tachycardia has been reported in 40% of patients even in the absence of fever. These abnormal cardiac rhythms are transient and do not warrant therapy. Patients were mostly asymptomatic. Quite a substantial proportion of patients have elevated creatine phosphokinase (CPK) levels.3,4 Yet, these are non-cardiac in origin and myocardial ischaemia has not been seen as a feature of SARS. Myocardial injury is unlikely to be the cause of hypotension despite the elevation of CPK in these patients because neither the MB isoenzyme nor troponin T is usually elevated.
Neurological manifestations Neurological manifestations of SARS appear
Clinical Presentation of the Disease in Adults
to be rare. Cases of epileptic fits, mental confusion and disorientation have been reported as isolated cases. Focal neurological deficit has not been found and CT and MRI of the brain have not revealed any structural abnormality. Lumbar puncture and analysis of cerebrospinal fluid (CSF) has been normal in most cases. It is in general difficult to determine whether the neurological features observed were directly related to the infection of SARS CoV or the treatment given (e.g. steroid, antibiotic). There was, however, at least one instance of SARS CoV detection in the CSF in a patient who developed status epilepticus during his illness.15 SARS CoV RNA was present in both the CSF and serum at substantial levels and hence may not be dismissed as coincidence. It is interesting to note that coronaviruses have been implicated in demyelinating brain pathology. Arbour et al. have documented the presence of the seemingly harmless human respiratory coronavirus OC43 in the brain parenchyma of patients with multiple sclerosis.16 Murine hepatitis virus, another coronavirus, has been linked to chronic inflammation and demyelination of the central nervous system. Therefore, SARS CoV infection of the brain is a distinct possibility. Further studies will be needed to address the potential neuropathological sequelae of SARS CoV infection of the central nervous system.
Atypical manifestations In common with many other diseases, SARS can present in an atypical manner. This is especially the case in the elderly patients and in those who are immunocompromised, making the clinical diagnosis much more difficult, if not impossible. Although high swinging fever is the most common presenting symptom of SARS, elderly patients may have no documented fever even in the presence of progressive pneumonia. Furthermore, patients may present not with respiratory but gastrointestinal symptoms such as diarrhoea, nausea or vomiting,
27
which, in frail elderly persons, may mimic faecal incontinence and poor feeding. Elders with SARS often have a longer incubation period of 14 to 21 days because of delayed detection of symptom onset. This has important clinical implications for diagnosis, contact tracing, duration of surveillance, transmission and infection control measures in nursing homes as well as in hospitals. The frequent occurrence of multiple pathologies (e.g. chronic obstructive airway diseases, old tuberculosis, congestive heart failure, diabetes mellitus) in the elderly may also mask the diagnosis of SARS. A retrospective study showed that two-thirds (100 out of 150) of elderly patients (above the age of 65) referred for suspected SARS had alternative diagnoses, compared with only onethird of younger patients. Because the presentations of SARS in the elderly can be non-specific, a positive contact history may be the first important clue. Diagnosis of SARS in old age requires a high index of suspicion, knowledge of the geriatric presentations of infections, awareness of the age-assessment changes in physical and functional state, alertness to any contact history of SARS, and an updated knowledge of the current prevalence of SARS in the locality. Patients with SARS could present with totally unrelated medical or even surgical conditions prior to the ultimate manifestations of SARS. There are reports of SARS patients presenting as acute pulmonary oedema, exacerbation of chronic obstructive airway disease, influenza, bacteraemia, acute abdomen and even hip fracture. These so-called ‘hidden’ SARS patients are the most challenging cases to infection control in hospitals and clinics. These unrecognized cases of SARS have been implicated in hospital-associated outbreaks in Hong Kong, Singapore, Taiwan, and Toronto. A high index of suspicion, the availability of a sensitive and specific virological testing and the judicious use of high resolution CT scan (HRCT) of thorax may be useful in the early diagnosis of SARS. The need of a rapid
28
Severe Acute Respiratory Syndrome
and reliable diagnostic test cannot be overemphasized.
response cytokines are produced in large quantities.17 While these cytokines are meant to eradicate the virus they may stimulate cellular proliferation in the lung. Patients start to develop respiratory symptoms such as shortness of breath (especially in the standing or sitting position); cough and chest radiographs show multiple areas of consolidation involving both lungs. This is a critical period in the illness. If the ‘cytokine storm’ does not resolve spontaneously or under therapeutic control, the lung may be permanently damaged. In the third week onward, the patient will go into the pulmonary destruction phase. Groundglass appearance of the lung will be seen on chest radiographs, arterial blood oxygenation will decline, and supplementary oxygen requirement will increase. Eventually, a reasonable oxygenation saturation level cannot be maintained with oxygen therapy, patients will require positive pressure ventilation. It is also during this phase of
SARS: triphasic disease From our experience in Hong Kong looking after over 1700 patients, it appears that SARS can be represented as a triphasic disease. In the first week of the illness, most patients presented with fever, myalgia and chills. There are usually minimal or no respiratory symptoms. This is likely to be a viral replicative phase. The virus multiplies in the host cells and is released into the bloodstream. Damage in the lung by the virus itself is limited at this stage but immune mediators (e.g. cytokines) are being produced. The disease progresses relatively slowly in the first week. Some patients may recover spontaneously in this phase of the disease. At least 70% of patients then proceed to the second phase which is an immune hyperactive phase. At this stage, early
Immune hyperactive phase
Max. daily body temperature (° C)
Viral replicative phase
Pulmonary destruction phase
4
3
3
3
3 1
2
3
4
5
6
7
8
9
10
11
12
13
Days after onset of disease Figure 3.4 Triphasic disease of SARS.
14
15
16
17
18
Clinical Presentation of the Disease in Adults
the disease that the immune system ‘burns out’, partly as a result of lymphopenia. The immune-paralysis state of the patient may lead to superimposed bacterial, viral or even fungal infections. The three phases of the disease are illustrated in Fig. 3.4.
References 1 WHO. Severe Acute Respiratory Syndrome (SARS). http://www.who.int/csr/media/sars 2 CDC. Updated Interim Surveillance Case Definition for Severe Acute Respiratory Syndrome (SARS). http://www.cdc.gov/ncidod/sars/ casedefinition. 3 Peiris JS, Chu CM, Cheng VC et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;361: 1767–72. 4 Lee N, Hui D, Wu A et al. A major outbreak of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348: 1986–94. 5 Wu W, Wang J, Liu P et al. A hospital outbreak of severe acute respiratory syndrome in GuangZhou, China. Chin Med J (Engl) 2003;116: 811–18. 6 Hsu LY, Lee CC, Green JA et al. Severe acute respiratory syndrome (SARS) in Singapore: clinical features of index patient and initial contacts. Emerg Infect Dis 2003;9: 713–17. 7 Booth CM, Matukas LM, Tomlinson GA et al. Clinical features and short-term outcomes of 144 patients with SARS in the greater Toronto area. JAMA 2003;289: 2801–9.
29
8 Wong KT, Antonio GE, Hui DS et al. Severe acute respiratory syndrome: radiographic appearances and pattern of progression in 138 patients. Radiology 2003;228: 401–6. 9 Antonio GE, Wong KT, Hui DS et al. Thinsection CT in patients with severe acute respiratory syndrome follow-up of hospital discharge: preliminary experience. Radiology 2003;228: 810–15. 10 Leung WK, To KF, Chan PKS et al. Enteric involvement of severe acute respiratory syndrome coronavirus infection. Gastroenterology 2003;125: 1011–17. 11 Wong RSM, Wu A, To KF et al. Hematological changes in patients with severe acute respiratory syndrome. Br Med J 2003;326: 1358–62. 12 Wu EB, Sung JJY. Hemorrhagic fever-like changes and normal chest radiograph in a doctor with SARS. Lancet 2003;361: 1520. 13 Chan HL, Leung WK, To KF et al. Retrospective analysis of liver function derangement in severe acute respiratory syndrome. Am J Med 2004;116: 566–7. 14 Hui AY, Chan HL, Liew CT et al. Fatal outcome of SARS in a patient with reactivation of chronic hepatitis B. Am J Med 2003;115: 334–6. 15 Hung EC, Chim SS, Chan PK et al. Detection of SARS coronavirus RNA in the cerebrospinal fluid of a patient with severe acute respiratory syndrome. Clin Chem 2003;49: 2108–9. 16 Arbour N, Day R, Newcome J et al. Neuroinvasion by human respiratory coronaviruses. J Virol 2000;74: 8913–21. 17 Wong CK, Lam CW, Wu AK et al. Plasma inflammatory cytokines and chemokines in severe acute respiratory syndrome. Clin Exp Immunol 2004;136: 95–103.
Chapter 4
SARS in Children CW Leung
Epidemiology Globally, it is estimated that children <18 years of age only accounted for about 5% of the 8096 patients with probable SARS to date.1 There was no mortality reported in children.2 In Hong Kong, a total of 121 children <18 years of age were clinically diagnosed with SARS, accounting for 6.9% of all patients notified.3,4 The crude attack rate was 8.9 per 100 000 population <18 years old. Complete serological workup was available for 111 of these 121 children clinically diagnosed with SARS in Hong Kong and 89 (46 boys and 43 girls) had evidence of seroconversion to SARS CoV. The corrected attack rate for children, based on serologic confirmation, was thus 6.6 per 100 000 population <18 years. Children <18 years of age account for about 20% of the general population in Hong Kong. The apparent discrepancy in age distribution of patients as compared to the general population is probably a result of a lower risk of exposure to infection. Most SARS transmission occurred in hospitals rather than in the home. In addition, the closure of schools during the epidemic probably further reduced transmission among children. Whether children with SARS are less infectious than adult patients remains unknown.2 Transmission from children has been documented,5 but this seems to be a rare occurrence. While the extent of asymptomatic SARS CoV infection in at-risk adult contacts appears low (see Chapters
30
13 and 14) there are fewer data about children. Of the 89 children with evidence of seroconversion to SARS CoV, 64 (72%) were hospitalized in two public hospitals in Hong Kong, the Princess Margaret Hospital (PMH) (n = 44)3 and United Christian Hospital (UCH) (n = 21)4 respectively. This review is largely based on data from these two cohorts (note: one adolescent who was transferred from UCH to PMH for critical care was included in both published reports). The median age was 12 years in both cohorts. The youngest patient was a premature infant aged 50 days.6 An epidemiologic link was established in all except two (97%) of these children. Fifty-one children (80%) were directly related to the Amoy Gardens community outbreak.7 There was no evidence of perinatal transmission to infants born to pregnant women with SARS in Hong Kong or elsewhere. 8–11
Clinical manifestation SARS in children is characterized by an abrupt onset of fever. Constitutional upsets of malaise, lethargy, chills, rigor, dizziness, headache, myalgia and anorexia are variable and more commonly reported by older children and adolescents. Coryza is found in 41% of patients and appears to be more commonly seen in children than in adults (see Chapter 3). Symptoms of cough, shortness of breath, difficulty in breathing and sputum production are not always present,
SARS in Children
even in children with radiographic evidence of moderately severe pneumonia. Gastrointestinal symptoms of nausea, vomiting, abdominal pain and diarrhoea may be present. Although the lungs are the site of major pathology in SARS, enteric involvement is clearly part of the disease process.12 Fever, malaise, cough, coryza, chills or rigor, sputum production, headache and myalgia are the most common presenting features (Table 4.1). Adolescents >12 years of age generally present with symptoms similar to those seen in adults. Malaise (77% vs 38%, p = 0.002), chills or rigor (50% vs 18%, p = 0.008), headache (47% vs 12%, p = 0.002), dizziness (33% vs 6%, p = 0.008) and myalgia (50% vs 9%, p = 0.003) are more common in teenagers, while children ⭐12 years are more likely to present with coryza (56% vs 23%, p = 0.01). In the PMH cohort, 91% of children presented with two or more clinical features in addition to fever, and 50% presented with four or more.3 The median duration of fever before hospitalization was 3 days (range 0–12 days). Twenty per cent of the children in PMH
developed respiratory distress and required oxygen supplementation at a mean of 6.2 ± 2.7 days (range 3–10 days) from onset of fever3 while only 10% of children admitted to UCH needed oxygen.4 Since many children received specific therapy besides supportive care, the natural history of untreated infection in children remains unclear. Nevertheless, three of the children with mild disease recovered on supportive therapy alone, apart from empiric antibiotic therapy for community-acquired pneumonia.13
Diagnosis An epidemiologic link and the failure of clinical response to antibiotics for the presumptive treatment of communityacquired pneumonia appeared to be the most important clues to clinical diagnosis during the outbreak of SARS.14 The original WHO case definitions for surveillance of SARS require that lower respiratory symptoms of cough, shortness of breath or difficulty in breathing be present.15 Since many
Table 4.1 Presenting clinical features of SARS in children Features
No. of children (n = 64)*
(%)
Fever Malaise Cough Coryza Chills or rigor Sputum production Headache Myalgia Anorexia Nausea and/or vomiting Dizziness Diarrhoea Sore throat Dyspnoea Abdominal pain Lethargy Chest pain Cyanotic attack
62 36 36 26 21 19 18 18 15 13 12 11 7 6 4 3 1 1
(97) (56) (56) (41) (33) (30) (28) (28) (23) (20) (19) (17) (11) (9) (6) (5) (2) (2)
*Data based on references 3 and 4.
31
32
Severe Acute Respiratory Syndrome
children do not present with these symptoms, applying this case definition to children would have resulted in many children being missed. The current revised WHO clinical case definitions for the postoutbreak period suffer from the same limitation.16 Furthermore, the WHO case definitions are not specific enough to differentiate SARS from other causes of community-acquired pneumonia. Paediatricians in Hong Kong have adopted a more clinically oriented case definition to aid diagnosis (Table 4.2).17 The performance characteristics of this clinical case definition were validated in PMH, with sensitivity, specificity, positive and negative predictive values of 97.8%, 92.7%, 88% and 98.7%, respectively, using seroconversion to SARS CoV as the gold-standard.3 However, it should be emphasized that the predictive values of any clinical case definition depend on the prevalence of the disease. Physical examination, radiological, haematological and biochemical investigations are contributory to the differential diagnosis of SARS, but may not confirm the diagnosis in many instances. Despite prominent radiographic evidence of pulmonary infiltrates, definite physical signs of consolidation are rarely evident. Chest radiographic findings are also not diagnostic, although some characteristic changes have been described. (See Chapter 6 and refs 3, 4, 14 and 18–20.) The primary abnormality is airspace disease, either ground-glass opa-
city or focal consolidation. Findings at presentation are often peripheral or mixed central and peripheral in distribution. Progression to unilateral multifocal or bilateral involvement with reduction in lung volumes is typical in patients who subsequently develop respiratory distress. HRCT of the thorax should be considered when the suspicion of SARS is high, but the initial chest radiograph is normal or equivocal.3,4,14 Small areas of focal subpleural pneumonic consolidation can be demonstrated earlier. The commonest laboratory findings in children with SARS at presentation include lymphopenia, leukopenia, thrombocytopenia, elevated lactate dehydrogenase level and mildly prolonged activated partial thromboplastin time. In comparison with adults, elevated hepatic transaminases, creatine kinase and d-dimer levels are less commonly seen in children. Microbiological investigations to exclude pathogens associated with communityacquired pneumonia, including those causing atypical pneumonia are essential, namely influenza A and B, parainfluenza types 1 to 4, respiratory syncytial virus, adenovirus, human metapneumovirus, Mycoplasma pneumoniae, Chlamydia pneumoniae, Chlamydia psittaci and Legionella pneumophila. The first-generation reverse transcription polymerase chain reaction (RT-PCR) assay for SARS CoV had a disappointingly low
Table 4.2 Clinical case definition of SARS used for children in Hong Kong Fever (rectal temperature ≥ 38.5°C or oral temperature ≥ 38°C) AND Chest X-ray findings of pneumonia or acute respiratory distress syndrome (ARDS) AND Suspected or probable contact with a person under investigation for or diagnosed with SARS, OR exposure to a locality with suspected or documented community transmission of SARS, either through travel or residence, within 10 days of onset of symptoms AND one or more of the following: Chills, malaise, myalgia, muscle fatigue, cough, dyspnoea, tachypnoea, hypoxia, lymphopenia, falling lymphocyte count, or failure to respond to antibiotics covering the usual pathogens of communityacquired pneumonia (e.g. a broad spectrum beta-lactam plus a macrolide) after 2 days of therapy in terms of fever and general well-being
SARS in Children
sensitivity with only 48% of children in PMH having a positive RT-PCR result in their nasopharyngeal aspirates (NPA) obtained soon after admission, generally within 1 week from onset of fever (median 5 days). Second-generation molecular tests now are believed to have higher sensitivity in the early stage of the illness (Chapter 8). Only 16% of children in the PMH cohort had successful isolation of SARS CoV from their NPA by cell culture.3 Seroconversion beyond 28 days from onset remained the gold standard for diagnosis.
Treatment During the outbreak of SARS in Hong Kong, the treatment of SARS in children was modelled on initial reports of success in treating adult patients with a combination of ribavirin and corticosteroids (Table 4.3)17 and it appears that children tolerate this regimen better with fewer side-effects than adults who may also suffer from comorbidities or advanced age. However, without data from multicentre randomized control trials of antiviral agents in adults to demonstrate efficacy (Chapter 20), the use of antiviral
33
agents cannot be recommended in children at present. Given the milder clinical course of SARS in young children, teenagers would probably be the at-risk group for consideration of any future antiviral regimen.
Outcome and prognosis Of the 64 children hospitalized in PMH and UCH, 11 (17%) developed respiratory distress necessitating oxygen supplementation. Five of these received intensive care and assisted ventilation was subsequently required in three. Of those who required intensive care, four were teenagers and the remaining patient was the youngest premature infant.6 It appears that teenagers are at higher risk of developing severe illness.21 Univariate analysis performed on the PMH cohort suggests that age >12 years (excluding the only infant who is considered an outlyer), sore throat at presentation, lowest absolute lymphocyte count, peak neutrophilia (>10 ¥ 109/l) and peak alanine aminotranserase (ALT) level (>80 IU/l) are associated with severe illness as reflected by oxygen requirement.3 Multivariate analysis by stepwise logistic regression shows that
Table 4.3 Treatment protocol used for children with SARS in Hong Kong Third generation cephalosporin (e.g. cefotaxime) plus macrolide (e.g. erythromycin or clarithromycin) for coverage of the usual pathogens of community-acquired pneumonia. Commence ribavirin 40–60 mg/kg/day po div Q8H if epidemiologic link forthcoming and febrile on admission. If fever persists, and no improvement in general well-being after 48 hours despite above regimen, commence corticosteroid: prednisolone 1–2 mg/kg/day po div BD or hydrocortisone 1–2 mg/kg iv Q6h if cannot tolerate oral medication. If the child develops hypoxaemia, and progressive clinical and/or radiographic deterioration, administer methylprednisolone 10 mg/kg/dose iv Q24H for up to 3 doses, depending on clinical response plus ribavirin 20–30 mg/kg/day iv div Q8H to replace oral ribavirin. Prednisolone 1–2 mg/kg/day or hydrocortisone 1–2 mg/kg Q6h is administered for a total of 2 weeks, and then tapered over 1–2 weeks. Ribavirin will be given for a total of 10–14 days. Antibiotics may be discontinued if afebrile for 5 days. The antibiotic regimen can be modified on clinical grounds if secondary or hospital-acquired infection is suspected.
34
Severe Acute Respiratory Syndrome
sore throat at presentation and peak neutrophilia are probable independent risk factors for severe illness. Neutrophilia may reflect the degree of inflammation or may be secondary to administration of steroids. However, it is worth noting that in two of nine children who required oxygen therapy in the PMH cohort, peak neutrophilia was already documented on admission, prior to administration of specific therapy.
Long-term sequelae In the PMH cohort, 27% of children who recovered from SARS have reported a mild decrease in exercise tolerance lasting from 1 week to 3 months after discharge.3 Subtle exercise impairment was detected in some children at 6 months from disease onset.22 However, even for the most severely affected teenagers, physical examination and pulmonary function tests have been normal at 6 months follow-up. Although residual minor subpleural ground-glass opacification and air trapping on expiration were demonstrated in some of them with the use of HRCT, no evidence of significant pulmonary fibrosis, bronchial wall thickening, bronchiectasis or lung volume loss was present. The need for oxygen supplementation and lymphopenia are risk factors in predicting abnormal HRCT findings at follow-up.23 About 40% of children reported increased shedding or diffuse thinning of hair, generally 2–3 months after disease onset. Hair shedding was self-limiting and all recovered spontaneously within 1–3 months.3 The phenomenon is most likely to be acute telogen effluvium which is manifested by excessive loss of normal club hairs. As ribavirin is not known to cause alopecia and glucocorticoids cause hypertrichosis or hirsutism instead, the condition is most probably secondary to the disease process itself. Telogen effluvium secondary to febrile systemic illness and severe psychological stress under life-threatening situations is well described and the timing of hair loss in the
SARS affected children is compatible with this. Avascular necrosis of the femoral head and, to a lesser extent, multifocal osteonecrosis were detected by screening magnetic resonance imaging (MRI) in both adults (Chapter 5) and children24 recovering from SARS, even in those without bone pain. At 6 months follow-up, 9% of children (<18 years) in Hong Kong had MRI evidence of osteonecrosis of varying severity (e-SARS database, Hospital Authority, Hong Kong Special Administrative Region, data on file). The long-term effects of SARS on children have not been fully realized. In summary, children, especially those under 12 years of age, generally have milder disease when compared with adults. Teenagers are at higher risk of developing severe illness and should be closely monitored for clinical and radiological deterioration. Diagnosis relies on a high index of suspicion, diligent search for an epidemiologic link, and the development of a reliable rapid diagnostic test. No evidence-based therapeutic approach exists to date. The importance of providing good supportive care, including oxygen and assisted ventilation, cannot be overemphasized. Children who recovered warrant longer-term followup.
References 1 World Health Organization. Summary of probable SARS cases with onset of illness from 1 November 2002 to 31 July 2003. Available from: URL: http://www.who.int/csr/sars/ country/table 2004_04_2 (accessed on 9 November 2004). 2 World Health Organization. Consensus document on the epidemiology of severe acute respiratory syndrome (SARS). Available from: URL: http://www.who.int/csr/sars/en/WHO consensus.pdf/(accessed on 26 June 2004). 3 Leung CW, Kwan YW, Ko PW et al. Severe acute respiratory syndrome among children. Pediatrics 2004;113: e535–e543. Available from: URL: http://pediatrics.aappublications. org/cgi/content/full/113/6/e535 (accessed on 26 June 2004). 4 Chiu WK, Cheung PCH, Ng KL et al. Severe acute respiratory syndrome in children:
SARS in Children
5
6
7
8
9
10
11
12
13
14
experience in a regional hospital in Hong Kong. Pediatr Crit Care Med 2003;4: 279–83. Chan WM, Kwan YW, Wan HS et al. Epidemiologic linkage and public health implication of a cluster of severe acute respiratory syndrome in an extended family. Pediatr Infect Dis J 2004; 23:1156–9. Sit SC, Yau EKC, Lam YY et al. A young infant with severe acute respiratory syndrome. Pediatrics 2003;112: e257–e260. Available from: URL: http://pediatrics.aappublications.org/ cgi/content/full/112/4/e257 (accessed on 26 June 2004). Department of Health, Hong Kong Special Administrative Region. Outbreak of severe acute respiratory syndrome (SARS) at Amoy Gardens, Kowloon Bay, Hong Kong: main findings of the investigation. Available from: URL: http:// www.info.gov.hk/info/ap/pdf/amoy_e.pdf (accessed on 26 June 2004). Shek CC, Ng PC, Fung GPG et al. Infants born to mothers with severe acute respiratory syndrome. Pediatrics 2003;112: e254–e256. Available from: URL: http://pediatrics. aappublications.org/cgi/content/full/112/4/ e254 (accessed on 26 June 2004). Wong SF, Chow KM, Leung TN et al. Pregnancy and perinatal outcomes of women with severe acute respiratory syndrome. Am J Obstet Gynecol 2004;191: 292–7. Shang JP, Wang YH, Chan LN et al. Clinical analysis of pregnancy in second and third trimester complicated severe acute respiratory syndrome. Zhonghua Fu Chan Ke Za Chi 2003;38: 516–20. Robertson CA, Lowther SA, Birch T et al. SARS and pregnancy: A case report. Emerg Infect Dis 2004;10: 345–8. Leung WK, To KF, Chan PK et al. Enteric involvement of severe acute respiratory syndrome-associated coronavirus infection. Gastroenterology 2003;125: 1011–17. Kwan MY, Chan WM, Ko PW et al. Severe acute respiratory syndrome can be mild in children. Pediatr Infect Dis J 2004;23:1172–3. Leung CW, Chiu WK. Clinical picture, diagnosis, treatment and outcome of severe acute
15
16
17
18
19
20
21
22
23
24
35
respiratory syndrome (SARS) in children. Paediatr Respir Rev 2004;5: 275–88. World Health Organization. Case definitions for surveillance of severe acute respiratory syndrome (SARS). Available from: URL: http://www.who.int/csr/sars/casedefinition/en/(accessed on 26 June 2004). World Health Organization. Alert, verification and public health management of SARS in the post-outbreak period. Available from: URL: http://www.who.int/csr/sars/postoutbreak/en/(accessed on 26 June 2004). Leung CW, LI CK. PMH/PWH interim guideline on the management of children with SARS. HK J Paediatr (New Series) 2003;8: 168–9. Babyn PS, Chu WCW, Tsou IYY et al. Severe acute respiratory syndrome (SARS): chest radiographic features in children. Pediatr Radiol 2004;34: 47–58. Bitnun A, Allen U, Heurter H et al. Children hospitalized with severe acute respiratory syndrome-related illness in Toronto. Pediatrics 2003;112: e261–e268. Available from: URL: http://pediatrics.aappublications.org/ cgi/content/full/112/4/e261 (accessed on 26 June 2004). Puthucheary J, Lim D, Chan I. Severe acute respiratory syndrome in Singapore. Arch Dis Child 2004;89: 551–6. Fong NC, Kwan YW, Hui YW et al. Adolescent twin sisters with severe acute respiratory syndrome (SARS). Pediatrics 2004;113: e146–e149. Available from: URL: http:// pediatrics.aappublications.org/cgi/content/ full/113/2/e146 (accessed on 26 June 2004). Li AM, Chan CH, Chan DF. Long-term sequelae of SARS in children. Paediatr Resp Rev 2004;5: 296–9. Li AM, So HK, Chu W et al. Radiological and pulmonary function outcomes of children with SARS. Pediatr Pulmonol 2004;38: 427–33. Chan CW, Chiu WK, Chan CC et al. Osteonecrosis in children with severe acute respiratory syndrome. Pediatr Infect Dis J 2004;23: 888–90.
Chapter 5
SARS: Sequelae and Implications for Rehabilitation David S Hui and Kenneth W Tsang
Introduction Although there are many descriptions on the clinical features and other aspects of SARS (Chapters 3 and 4) there are few studies reporting on the sequelae of SARS.1 The short duration of follow-up available on SARS patients so far and the reluctance of patients and physicians alike to screen systematically for pulmonary and extrapulmonary sequelae are key hurdles in our understanding of this subject.
Pulmonary sequelae Acute respiratory failure is the first major sequela encountered during the early phase of the disease. The clinical course of SARS is characterized initially by fever, myalgia and other systemic symptoms that generally improve after a few days, followed by a second phase with recurrence of fever, oxygen desaturation, and radiological progression of pneumonia. Approximately 40–50% of patients may need supplemental oxygen during clinical progression. A retrospective study has shown that oxygen requirement was significantly higher among patients not given pulse steroid therapy.2,3 While the majority of patients improve with pulse steroid treatment, 20–36% of patients may require intensive care admission whereas 13–26% may progress into acute respiratory distress syndrome (ARDS) (Fig. 5.1) necessitating invasive mechanical ventilatory support.4–7 Seventeen out of 37 adults admitted
36
Figure 5.1 Chest radiograph of a 70-year-old male with SARS who developed radiographic changes of acute respiratory distress syndrome. The patient required invasive mechanical ventilation and finally succumbed.
to an ICU were noted to develop nosocomial bacterial infections.8 In six patients, the diagnosis of nosocomial infection was made before high dose methylprednisolone was started. Infections included pneumonia in ten patients (methicillin-resistant Staphylococcus aureus in six, Stenotrophomonas maltophilia in two, Candida in one and polymicrobial in one), urinary tract infection in two and bacteraemia with no clearly identified primary site of infection in five.8
SARS: Sequelae and Implications for Rehabilitation
37
Figure 5.2 Chest radiograph of a 33-year-old male with SARS who developed spontaneous pneumomediatinum with surgical emphysema.
Compared with adults and teenagers, SARS seems to run a less aggressive clinical course in younger children, and in one case series, none of the children aged below 13 years required supplemental oxygen (see Chapter 4). Despite the use of low volume and low pressure during mechanical ventilation for SARS patients with severe respiratory failure, the incidence of barotrauma is unexpectedly high. Spontaneous pneumothorax has been observed in 3% of cases9 whereas 12% of patients developed spontaneous pneumomediastinum over a period of 3 weeks in another case series (Fig. 5.2).4 In addition, 7 of 27 (25.9%) ventilated patients developed evidence of barotraumas.10 The CT features of late-stage ARDS are similar to those seen in late-stage ARDS of other causes but severe SARS-induced ARDS may result in cyst formation independently.11 Lung pathology is summarized in Chapter 10.
In a major outbreak at the Prince of Wales Hospital, Hong Kong, over half of those infected were health-care workers, most of whom were young and fit before contracting SARS.5 The chest radiographs of some of these patients have shown persistent, bilateral alveolar opacities at 3 weeks from onset of illness. High resolution computer tomography (HRCT), performed selectively on 24 patients with residual opacities has revealed multiple patchy ground-glass appearance and interstitial thickening (n = 9, 38%) whereas CT evidence of fibrotic changes was noted in 15 (62%) patients (Figs 5.3 and 5.4).12 It is possible that active alveolitis, probably as a result of the inflammatory response triggered by the virus may lead to pulmonary fibrosis in some patients (Chapter 10). Preliminary experience suggests, however, that some of these fibroticlooking lesions on CT scan appear to resolve, in contrast to the other pulmonary fibrotic conditions.
38
Severe Acute Respiratory Syndrome
Figure 5.3 HRCT of a 39-year-old patient with SARS who still had residual bilateral ground-glass parenchymal opacities, and fibrotic changes at the right upper lobe 3 months after disease onset.
Figure 5.4 HRCT of a 54-year-old female who developed right-sided pneumothorax at 60 days from symptom onset. Honeycomb changes in keeping with fibrosis are noted bilaterally. Large cystic changes are noted at the right lung and may be related to barotraumas.
SARS: Sequelae and Implications for Rehabilitation
Full lung function tests performed on a large cohort of SARS survivors5 at 6 months from symptom onset of illness have shown a restrictive spirometric pattern (FEV1/FVC >80%) in over 80% of patients. About 15% of patients demonstrated a reduction of diffusing capacity (corrected for haemoglobin and reflecting alveolar surface area for gas exchange) of <80% of the predicted values with well-preserved diffusing capacity per unit volume. This suggests that there is an increase of the intra-alveolar diffusion pathway, which may be due to atelectasis, alveolitis and pulmonary fibrosis. There was evidence of respiratory muscle weakness on testing with maximum inspiratory (Pimax) and expiratory (Pemax) pressures, measured with mouth manometers, of <80 cm H20 in 14% and 22% of patients, respectively, with a concomitant increase in residual volume.13 The muscle weakness might be related to muscle deconditioning, prolonged bed rest, catabolic effects related to sepsis and steroid myopathy. In addition, the 6-minute walk distance, an integrated test of cardiorespiratory fitness, was also reduced significantly among all age groups compared to normative data. Not surprisingly, evaluation of health-related quality of life among these survivors has shown impairment in all domains.13
Extra-pulmonary sequelae Adverse reactions of ribavirin The initial speculation that SARS was most likely to be a viral illness and thus should be treated with a broad-spectrum antiviral agent led to the use of ribavirin in combination with corticosteroid in the treatment of initial cases of SARS in Hong Kong.2,4,13 The favourable outcome for some of these cases probably led to the wide adoption of ribavirin as an antiviral agent in the treatment regime for SARS in Hong Kong and elsewhere, largely anecdotally and historically.6–9 With the dosage of ribavirin used in Hong Kong (1.2 g tds orally), we
39
have observed a decrement in haemoglobin by 2 g/dL in 59% of patients but none required blood transfusion or developed symptomatic bradycardia.8 A recent retrospective analysis by the Hong Kong Hospital Authority has shown that SARS patients who were treated with (n = 1611) and without ribavirin (n = 84) dropped their haemoglobin by (mean ± SD) 3.0 ± 1.9 and 1.7 ± 1.4 g/dL respectively, p<0.001 (unpublished data). A much higher dose of ribavirin, based on dosage for treatment of haemorrhagic fever viruses, has been reported by Booth et al.9 to be associated with more significant toxicity, including haemolysis (in 76%) with a decrease in haemoglobin of 2 g/dL (in 49%), elevated transaminases (in 40%) and bradycardia (in 14% of SARS patients). These side effects have been shortlived and generally resolved within 1 month after cessation of ribavirin. The attribution to ribavirin for liver dysfunction also appears inappropriate; it now seems that this is likely to be a facet of the multi-system involvement of SARS-CoV infection.14 On follow-up, folic acid 5 mg daily for 1 month should be prescribed to those with significant haemolysis and haemoglobin <10 g/dL.
Neurological and psychiatric aspects At the early post-discharge phase, over 40% of 204 SARS patients had at least a diagnosable psychiatric disorder. A spectrum of psychiatric morbidities was encountered at the initial recovery stage of SARS. The commonest diagnosis was organic mood disorder related to SARS illness and possible steroid effect. This was followed by major depression, adjustment disorder, post traumatic stress disorder and anxiety disorder. In the recovery period, there was late onset psychiatric morbidity, especially depressive illness related to resumption of work and continuing psychosocial stress.15 The prolonged isolation and extreme uncertainty imposed on SARS patients, coupled with intense
40
Severe Acute Respiratory Syndrome
media attention must have scarred some, if not all, patients psychologically. This is compounded by the understandable phobia expressed towards SARS survivors from some members of the general public, particularly in the initial phase of the outbreak because of fear of acquiring the infection.16 Many patients also reported difficulties in mental concentration, although neither the aetiology nor the natural history of this defect is known. It was estimated that 1% of SARS patients developed psychotic disorders. Steroid-related mania was the commonest clinical diagnosis. Compared to nonpsychotic controls, SARS-related psychosis was more common among patients with family psychiatric history.17 A survey on SARS patients in the rehabilitation phase has shown that general stress was high, and negative psychological effects were common among SARS patients, indicating risk of mood and stress-related disorders. Functional impairment was present in the post-recovery phase (Dr Siew Chau, University of Hong Kong, personal communication). Although SARS CoV has been detected in the cerebrospinal fluid and serum samples of two patients who developed status epilepticus18,19 the contribution of direct viral invasion of the CNS to the neuropsychiatric impairment is uncertain.
Musculoskeletal aspects Avascular necrosis (AVN) of large joints has been detected on MRI screening in 135 (12.1%) of 1117 SARS patients in Hong Kong at 6 months from disease onset (Fig. 5.5).20 While therapy is likely to be a contributing factor in the pathogenesis of AVN, it is possible that other factors such as active inflammation, hyperviscosity, and the presence of thrombosis in systemic small vessels in SARS patients (Chapter 10) may all play a role. It is important to note that SARS patients with AVN confirmed on MRI scan are largely symptomatic with joint pain where-
Figure 5.5 MRI T1 weighted coronal image of a middle-aged male patient with osteonecrosis involving the left femoral head (courtesy of Dr James Griffith and Dr Greg Antonio, Chinese University of Hong Kong).
as 79% of cases of arthralgia among the SARS survivors are not associated with MRI abnormalities. Bone densitometry study is in progress to detect other potential complications such as osteoporosis.
Summary SARS is an emerging infectious disease with a significant morbidity and mortality. Acute respiratory failure is the major complication in the acute stage, and about one in four and one in eight patients require intensive care admission and invasive ventilatory support, respectively. Other possible short-term pulmonary sequelae include pneumothorax and pneumomediastinum. CT evidence of alveolitis and respiratory muscle weakness are frequently observed after discharge but few patients develop significant pulmonary fibrosis after the acute illness. SARS may be complicated by a wide range of extra-pulmonary sequelae and a comprehensive rehabilitation programme and assessment with a multi-disciplinary approach are needed.
SARS: Sequelae and Implications for Rehabilitation
References 1 Chan KS, Zheng JP, Mok YW et al. SARS — Prognosis, outcome and sequelae of SARS. Respirology 2003;8: S36–40. 2 Ho JC, Ooi GC, Mok TY et al. High dose pulse versus non-pulse corticosteroid regimens in Severe Acute Respiratory Syndrome. Am J Respir Crit Care Med 2003;168: 1449–56. 3 Tsang KW, Zhong NS. Severe acute respiratory syndrome — pharmacotherapy. Respirology 2003;8: S25–30. 4 Peiris JSM, Chu CM, Cheng VCC et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;361: 1767–72. 5 Lee N, Hui DS, Wu A et al. A major outbreak of Severe Acute Respiratory Syndrome in Hong Kong. N Engl J Med 2003;348: 1986–94. 6 Chan JW, Ng CK, Chan YH et al. Short term outcome and risk factors for adverse clinical outcomes in adults with Severe Acute Respiratory Syndrome (SARS). Thorax 2003;58: 686–9. 7 Tsui PT, Kwok ML, Yuen H et al. Severe Acute Respiratory Syndrome: Clinical outcome and prognostic correlates. Emerg Infect Dis 2003;9: 1064–9. 8 Sung JJ, Wu A, Joynt GM et al. Severe Acute Respiratory Syndrome: report of treatment and outcome after a major outbreak. Thorax 2004;59: 414–20. 9 Booth CM, Matukas LM, Tomlinson GA et al. Clinical features and short-term outcomes of 144 patients with SARS in the greater Toronto area. JAMA 2003;289: 2801–9. 10 Gomersall CD, Joynt GM, Lam P et al. Shortterm outcome of critically ill patients with Severe Acute Respiratory Syndrome. Intensive Care Med 2004;30: 381–7.
41
11 Joynt GM, Antonio GE, Lam P et al. Late-stage adult respiratory distress syndrome caused by severe acute respiratory syndrome: abnormal findings at thin-section CT. Radiology 2004; 230: 339–46. 12 Antonio GE, Wong KT, Hui DS et al. Thinsection CT in patients with severe acute respiratory syndrome following hospital discharge: preliminary experience. Radiology 2003;228: 810–15. 13 Hui DS. The impact of SARS on pulmonary function, exercise capacity, and quality of life in a cohort of survivors. Hong Kong SARS Forum and Hospital Authority Convention, 8–11 May 2004: (Abstract) 72. 14 Wong WM, Ho JC, Ooi GC et al. Temporal patterns of hepatic dysfunction and disease severity in SARS. JAMA 2003;290: 2663–5. 15 Wing YK, Leung CM, Kam IW et al. Hidden morbidity: psychosocial perspective. Hong Kong SARS Forum and Hospital Authority Convention, 8–11 May 2004: (Abstract) 73. 16 Wong GW, Hui DS. Severe acute respiratory syndrome: epidemiology, diagnosis and treatment. Thorax 2003;58: 558–60. 17 Lee DT. SARS-related psychosis: a case-control study. Hong Kong SARS Forum and Hospital Authority Convention, 8–11 May 2004: (Abstract) 49. 18 Hung EC, Chim SS, Chan PK et al. Detection of SARS coronavirus RNA in the cerebrospinal fluid of a patient with severe acute respiratory syndrome. Clin Chem 2003;49: 2108–9. 19 Lau KK, Yu WC, Chu CM et al. Possible central nervous system infection by SARS coronavirus. Emerg Infect Dis 2004;10: 342–4. 20 Leung L, Antonio GE, Yuen MK et al. Magnetic resonance screening for skeletal abnormalities in post-SARS patients. Hong Kong SARS Forum and Hospital Authority Convention, 8–11 May 2004: (Abstract) 45.
Chapter 6
Radiology of SARS Clara GC Ooi
Introduction
A
Radiological assessment of SARS plays a vital role in the management of the disease. 1–21 This chapter deals with the utility of imaging in the diagnosis and follow-up of SARS, the relationship between radiological and clinical parameters and the exclusion of other differential diagnoses.
Radiological diagnosis Radiological diagnosis of SARS depends on the presence of pneumonia, shown either on a chest radiograph or by high resolution computed tomography (HRCT). The radiological hallmark of SARS, as with most other pneumonic illness, is airspace shadowing,4–16,19,20 although one caveat to be noted is that a patient with suspected SARS can have a normal initial chest radiograph. In published series on SARS, 10–40% of symptomatic patients had normal chest radiographs on initial evaluation.4,9,12,13
B
Radiographical features The earliest radiographical manifestation of SARS is ground-glass opacity (Fig. 6.1A),4,6,8 that typically and rapidly progresses in the majority of patients to focal, multifocal or diffuse consolidation (Fig. 6.1B).4,6,8,9,11,12,19,20 In general, the consolidations are usually ill-defined and multifocal (Fig. 6.2), although well-defined unifocal consolidations that enlarge with
42
Figure 6.1 A 46-year old female with a 3-day history of fever, proven to have SARS. (A) Admission chest radiograph showed ground-glass opacification in the right lower lobe, (B) which progressed to diffuse consolidation (ref 5).
Radiology of SARS
43
A
Figure 6.2 A 29-year-old female with SARS. Multifocal ill-defined areas of consolidation are noted in the lower zones of both lungs (ref 5).
time (Fig. 6.3) can also be found. An interstitial pattern, common in other atypical pneumonias such as mycoplasma pneumonia, is almost never seen at presentation although nodular opacities are encountered in a small minority of patients.1,6–8 On the chest radiograph the consolidation is commonly unilateral at presentation4,7,9 becoming bilateral at maximal lung abnormality (Fig. 6.1),4,7 although an HRCT examination of the thorax performed in the acute stages will often reveal bilateral lung involvement in a large proportion of cases. The distribution of consolidation has been described to be peripheral with lower zone predominance,1,2,6,8,9,11 while a pure upper or mid-zone (Fig. 6.4) distribution is found in only a smaller proportion of patients.6,9,7 When diffuse, the consolidation can be indistinguishable from acute respiratory distress syndrome (ARDS) (Fig. 6.1B). The pattern of disease progression with respect to severity assessed on the chest radiograph has been described in three series.4,9,11 In one series of 40 patients with serologically proven SARS radiographical opacities continued to deteriorate with radiographical scores peaking at 9.35 ± 4.09
B
Figure 6.3 A 35-year-old female with SARS. (A) Focal ground-glass opacification is noted on the right lower zone of her chest on her admission radiograph. (B) Four days later the focal opacity had become larger and more opaque (ref 5).
(median 9) from onset of illness and 5 days after treatment before declining.4 In a larger cohort of 138 probable cases of SARS, Wong et al. described 4 different patterns of progression.9 The most common pattern was found in 70% of patients in which radiographical scores peaked at 8.6 ± 3.1 days after onset of fever before declining. A smaller proportion of patients (17.4%) experienced two radiographical peaks,
44
Severe Acute Respiratory Syndrome
Figure 6.4 39-year-old female with SARS. Chest radiograph show left upper zone focal consolidation.
which occurred at 6.3 ± 3 and 13.5 ± 3.7 days respectively after onset of fever before improvement in scores. A minority (10) of patients (7.2%) showed minimal radiographical progression throughout their illness, while relentless progression of radiographical opacities culminating in death occurred in 6 of 7.9 In a third review of probable SARS cases, two patterns of serial radiographical changes that reflected clinical outcome were reported.11 The consolidations evaluated on the chest radiograph either improved rapidly or, in the majority of patients (72%), progressed to bilateral extensive pneumonia.11 Other features such as pleural effusions and mediastinal lymphadenopathy are not typical radiographical features of SARS, although small pleural effusions can be detected on HRCT scans.4–14,17 When effusions are present, they are usually associated with fluid overload or heart failure. Atelectasis can be a feature in the later stages of the illness, in the convalescent period or with disease progression when there is marked volume loss noted on the chest radiograph. Cavitating consolidations have never been reported in the acute period but can develop in patients with a protracted
Figure 6.5 A 44-year old man with SARS. Pneumomediastinum (arrowheads) is present with opacities affecting bilateral lower zones.
course particularly in those with mechanical ventilation, who are susceptible to superadded infections. Other complications that occur, particularly in association with extensive disease and/or mechanical ventilation, are pneumomediastinum and pneumothorax (Fig. 6.5).
Correlation between radiological and clinical parameters The clinical utility of radiographical evaluation particularly using radiographical scores for correlation with clinical parameters in SARS has been substantiated in several studies.4,5,9,15,16 The severity of lung abnormalities in SARS on chest radiographs reflected temporal changes in clinical and laboratory parameters such as heart rate, temperature, oxygen saturation (SaO2) and liver transaminases.4,15,16 Deteriorating radiographical scores corresponded linearly to decreasing SaO2 and increasing liver aminotransaminases.4,16 These findings suggest that the severity of the disease on a chest radiograph can serve as an indirect indicator of impaired oxygen exchange and liver derangement.
Radiology of SARS
The radiographical pattern of consolidation in SARS has been shown to have consequences for patient prognosis and condition.5,21 Liu et al. compared the radiographical features of 36 patients with severe SARS (11 of whom died) and 224 patients with a milder form of the disease.21 Patients with severe SARS had significantly more cases with large areas of consolidation or lung infiltration, diffuse consolidation and bilateral lung involvement on a chest radiograph compared with their counterparts.21 In another study, diffuse consolidation at maximal radiographical abnormality was associated with an increased likelihood of requiring O2 supplementation.5 Such a relationship was not found with either focal or multifocal (patchy) consolidation. In the same study it was shown that radiographical parameters that were associated with treatment response included delay in treatment after onset of symptoms, time to develop maximal lung abnormalities after treatment, maximal and treatment radiographical scores, and SaO2 at maximal radiographical score. Poor responders had significantly higher radiographical scores, lower SaO2 at maximal lung changes, and took a longer time from treatment to develop maximal lung change compared with either good or fair responders. Admission radiographical scores of poor responders were not significantly different from those of good responders and hence initial radiographical severity was not a factor that differentiated a good from a poor responder. However, severity at maximal lung abnormality and the time it took to reach that stage after initiation of treatment were.5
High resolution computed tomography (HRCT) Although chest radiography is the routine investigation for any pneumonical illness, HRCT was extremely useful in confirming airspace opacities during the SARS epidemic when initial chest radiographs were normal
45
in patients who had strong contact history and signs and symptoms that were highly suspicious of SARS. Confirmation of pneumonia thus allowed appropriate management, which included prompt isolation of these patients thereby preventing further spread of the virus.6–9,18 HRCT is superior to chest radiography in documenting the extent of the disease particularly in regions of the lungs not easily visualized on the chest radiograph such as the retrocardiac area and the lung bases and therefore allows more detailed analysis and quantification of lung abnormalities. The characteristic HRCT feature in the acute phase of SARS, as with chest radiography was ground-glass opacities, sometimes associated with consolidation in a subpleural location, which progressed rapidly to involve other areas of the lungs.6–8,10,17,18 The ground-glass opacities were characteristically associated with smooth interlobular and intralobular septal thickening (Fig. 6.6). The temporal pattern of lung abnormalities on HRCT was evaluated in 30 patients who had serial HRCT scans in one study.18 Ground-glass opacification with or without subpleural consolidation (Fig. 6.7A) was found in the first week of illness. By the second week after onset of fever, irregular reticular opacities superimposed on ground-glass opacities (Fig. 6.7B) were noted; the number of scans with this feature peaked during the fourth week of illness.18 Purely reticular (linear) opacities were observed from the second week of illness, reaching their highest proportions by the fourth week. After the fourth week, 55% of patients had features suggestive of early fibrosis, i.e reticular opacities with or without residual ground-glass opacities that were associated with architectural distortion and bronchial dilatation (Fig. 6.8). In a separate review of 24 cases discharged from hospital, HRCT evidence of fibrosis was present in 15 (62%) cases, while 8 of the remaining 9 cases had HRCT evidence of residual disease that did not resemble fibrosis.17 The group with HRCT evidence of fibrosis were older,
46
Severe Acute Respiratory Syndrome
Figure 6.6 HRCT showing extensive ground-glass opacification in both lower lobes with superimposed smooth interlobular septal thickening (arrowheads) Note nodular subpleural consolidation in both lower lobes (arrows).
A
B
Figure 6.7 A 45-year-old female with SARS. (A) Axial HRCT scan performed 10 days after onset of symptoms shows diffuse ground-glass opacities affecting the upper lobes. (B) Irregular linear opacities have developed in the areas of ground-glass opacities by day 20 of illness (ref 18).
Radiology of SARS
47
Figure 6.8 36-year-old man with SARS. HRCT performed 3 months after discharge shows features consistent with fibrosis (arrows) in the lower lobes.
had higher rates of intensive care treatment, more requirement for pulsed steroid therapy, higher peak lactate dehydrogenase levels and higher peak opacification on chest radiographs than their counterparts. The authors concluded that SARS patients who were older and had more severe disease were more likely to develop lung fibrosis.17
SARS versus non-SARS pneumonia When contact history or exposure is easily established such as during an outbreak or epidemic, the radiographical appearance and progression including HRCT characterization are very helpful aids to diagnosis. However, at the end of an outbreak or when contact history is tenuous or difficult to establish, it may be difficult to diagnose SARS based on radiographical appearances alone, as these features are non-specific. It has to be reiterated that the airspace abnormalities described above can also be found in other atypical pneumonias, particularly viral and bacterial infections. The muchreported peripheral distribution of airspace opacities in SARS is shared with other conditions such as bronchiolitis obliterans with organizing pneumonia (BOOP) and eosinophilic pneumonitis. Differential diagnoses for HRCT lung appearances in SARS, particularly those with ground-glass opacities and smooth interlobular septal
thickening, include ARDS, acute interstitial pneumonitis, BOOP, alveolar proteinosis, pulmonary oedema, haemorrhage and also other viral pneumonias such as herpes simplex and cytomegalovirus infections.1,22–27 However, the predilection for extensive ground-glass opacities rather than consolidation on HRCT may help differentiate SARS from bacterial pneumonias, which characteristically present as segmental or lobular consolidation, with perifocal ground-glass opacities.18,25,28 In addition, the absence of peribronchiolar consolidation, bronchopneumonia and nodules, particularly centrilobular and tree-in-bud nodules, are further helpful pointers against viral and mycoplasma pneumonias as well as infectious bronchiolitis.24–28 Compared with non-SARS pneumonias, the lung abnormalities in SARS on chest radiographs tend to progress at a faster rate and also frequently involve other lung zones and contralateral lung that were not initially involved (GC Ooi, personal communication). Non-SARS pneumonias tend to remain unilateral and within the lung zone of involvement despite an increase in size of the opacity. Radiographical patterns at presentation and at maximal lung abnormality may also be helpful in differentiating SARS from non-SARS pneumonia. In a review of chest radiographs of SARS and non-SARS pneumonia patients, the
48
Severe Acute Respiratory Syndrome
proportion of chest radiographs with either normal appearance or ground-glass opacities at presentation were higher in patients with SARS than in those with non-SARS pneumonia (GC Ooi, personal communication). Conversely, focal or multifocal (patchy) consolidation was more common at presentation in non-SARS pneumonia than in SARS (unpublished data). At maximal lung abnormality, focal consolidation was found in over half of patients with nonSARS pneumonia whilst multifocal or diffuse involvement was more frequently found in the SARS group. Although the lung abnormalities in SARS could not be accurately described as ‘flitting from one lung area to another’, as one might expect to see in immune-mediated lung disorders, its hallmark of progressive deterioration that involves other lung areas within days appears to set SARS aside from other pneumonias. Whether this is due to immunologic lung injury or to the SARS CoV itself remains unclear.
Follow-up of lung changes At discharge, a significant proportion of patients will have some residual lung abnormalities on the chest radiograph. The question of whether these residual abnormalities are irreversible changes remains unanswered as preliminary HRCT studies of patients in the convalescent or early postdischarge period have raised the possibility that these may represent early fibrosis.7,10,18 Longer-term follow-up of discharged SARS patients, particularly those with residual opacities on the chest radiograph with HRCT as well as lung function assessment, would be required to establish the longterm sequelae of these patients with respect to irreversible fibrosis.
Control of cross-infection Satellite radiography centres with dedicated mobile radiography machines, chest stands and lead screens should be set up in
the vicinity of SARS wards and clinics. For non-ambulatory patients from SARS wards and intensive care units, mobile radiography machines from the satellite radiography centres should be used. All attending radiology personnel should adhere to infection control measures including wearing of gowns, caps, masks and gloves, even when handling film cassettes from these high risk areas.29 After radiography or CT examination of patients with SARS or suspected SARS the gantry table and floors should be cleansed and bed linen changed. In addition, control panels, computed key boards, radiography mobile units and telephones and desk surfaces of the satellite areas and CT suite should be regularly disinfected.
Conclusion The imaging of SARS in the first instance is largely based on radiographical evaluation, with the application of HRCT as an additional imaging modality for further characterization and quantification of lung abnormalities and exclusion of other differential diagnoses. The plain chest radiograph is extremely useful in providing a rapid snapshot view of disease severity in the lungs of SARS patients on a daily basis. The visual quantification of lung abnormalities can be used to correlate with disease activity reflected by clinical and laboratory parameters, an exercise which may be useful in the clinical management of these patients. The role of HRCT lies in identification of airspace opacities in normal or questionable radiographs allowing early treatment and isolation, as well as in the later stages of the disease to detect complications and irreversible changes.
References 1 Tsang KW, Ho PL, Ooi GC et al. A cluster of cases of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348: 1977–85. 2 Lee N, Hui D, Wu A et al. A major outbreak of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348: 1986–94.
Radiology of SARS
3 World Health Organization. Case Definitions for Surveillance of Severe Acute Respiratory Syndrome (SARS). Revised 1 May 2003. Available from: URL: http://www.who.int/ csr/sars/casedefinition/en/ 4 Ooi GC, Khong PL, Lam B et al. Severe acute respiratory syndrome: relationship between radiologic and clinical parameters. Radiology 2003;229: 492–9. 5 Ooi GC, Khong PL, Ho JC et al. Severe acute respiratory syndrome: radiographic evaluation and clinical outcome measures. Radiology 2003;229: 500–6. 6 Muller NL, Ooi GC, Khong PL et al. Severe acute respiratory syndrome: radiographic and CT findings. Am J Roentgenol 2003;181: 3–8. 7 Muller NL, Ooi GC, Khong PL et al. Highresolution CT findings of severe acute respiratory syndrome at presentation and after admission. Am J Roentgenol 2004;182: 39–44. 8 Wang R, Sun H, Song L et al. Plain radiograph and CT features of 112 patients with SARS in acute stage. Beijing Da Xue Xue Bao 2003; 35(S): 29–33. 9 Wong KT, Antonio GE, Hui DS et al. Severe acute respiratory syndrome: radiographic appearances and pattern of progression in 138 patients. Radiology 2003;228: 401–6. 10 Antonio GE, Wong KT, Hui DS et al. Thin-section CT in patients with severe acute respiratory syndrome following hospital discharge: preliminary experience. Radiology 2003;228: 810–15. 11 Grinblat L, Shulman H, Glickman A et al. Severe acute respiratory syndrome: radiographic review of 40 probable cases in Toronto, Canada. Radiology 2003;228: 802–9. 12 Nicolaou S, Al-Nakshabandi NA, Muller NL. SARS: Imaging of severe acute respiratory syndrome. Am J Roentgenol 2003;180: 1247–9. 13 Hon KL, Leung CW, Cheng WT et al. Clinical presentations and outcome of severe acute respiratory syndrome in children. Lancet 2003;361: 1701–3. 14 Tsou IY, Loh LE, Kaw GJ et al. Severe acute respiratory syndrome (SARS) in a paediatric cluster in Singapore. Pediatr Radiol 2004;34: 43–6. 15 Ho JC, Ooi GC, Mok TY et al. High-dose pulse versus non-pulse corticosteroid regimens in severe acute respiratory syndrome. Am J Respir Crit Care Med 2003;168: 1449–56. 16 Wong WM, Ho JC, Ooi GC et al. Temporal patterns of hepatic dysfunction and disease severity in patients with SARS. JAMA 2003;290: 2663–5.
49
17 Wong KT, Antonio GE, Hui DS et al. Thinsection CT of severe acute respiratory syndrome: evaluation of 73 patients exposed to or with the disease. Radiology 2003;228: 395–400. 18 Ooi GC, Khong KL, Muller NL et al. Severe acute respiratory syndrome: temporal lung changes at thin-section CT in 30 patients. Radiology 2004;230: 836–44. 19 Lu P, Zhou B, Chen X et al. Chest X-ray imaging of patients with SARS. Chin Med J 2003;116: 972–5. 20 Du X, Zhu Q, Hong N et al. The investigation of clinical images of 88 hospital staff with SARS. Beijing Da Xue Xue Bao 2003;35S: 34–7. 21 Liu J, Tang X, Jiang S et al. The chest X-ray image features of patients with severe SARS: a preliminary study. Chin Med J 2003;116: 968–71. 22 Bouchardy LM, Kuhlman JE, Ball WC et al. CT findings in bronchiolitis obliterans organizing pneumonia (BOOP) with radiographic, clinical and histologic correlation. J Comput Assist Tomogr 1993;17: 352–7. 23 Müller NL, Guerry-Force ML, Staples CA et al. Differential diagnosis of bronchiolitis obliterans with organizing pneumonia and usual interstitial pneumonia: clinical, functional and radiologic findings. Radiology 1987;162: 151–6. 24 Johkoh T, Itoh H, Müller NL et al. Crazypaving appearance at thin-section CT: spectrum of disease and pathologic findings. Radiology 1999;211: 155–60. 25 Tanaka N, Matsumoto T, Kuramitsu T et al. High resolution CT findings in communityacquired pneumonia. J Comput Assist Tomogr 1996;20: 600–8. 26 Kim EA, Lee KS, Primack SL et al. Viral pneumonias in adults: radiologic and pathologic findings. Radiographics 2002;22: S137–49. 27 McGuinness G, Scholes JV, Garay SM et al. Cytomegalovirus pneumonitis: spectrum of parenchymal CT findings with pathologic correlation in 21 AIDS patients. Radiology 1994;192: 451–9. 28 Reittner P, Ward S, Heyneman L et al. Pneumonia: high-resolution CT findings in 114 patients. Eur Radiol 2003;13: 515–21. 29 Ho SS, Chan PL, Wong, PK et al. Eye of the storm: The roles of a radiology department in the outbreak of severe acute respiratory syndrome. Am J Roentgenol 2003;181: 19–24.
Chapter 7
Aetiology of SARS Malik Peiris and Albert DME Osterhaus
The initial outbreak Severe acute respiratory syndrome (SARS) emerged in southern China in late 2002. Historical aspects of its emergence are summarized in Chapters 1 and 2. The disease came to international attention when outbreaks affecting health-care workers occurred in the provincial capital of Guangdong province, Guangzhou in January 2003. Clinicians in Guangzhou had recognized it as a rapidly progressive ‘atypical’ pneumonia that failed to respond to conventional antibiotic therapy, associated with clusters of cases in the family or in health-care workers and termed it ‘infectious atypical pneumonia’.1 By 11 February 2003, WHO had been informed of 305 cases of this disease with 5 deaths in Guangdong province, over 100 of which were those of health-care workers.2 In response, health authorities in Hong Kong set up enhanced epidemiological and microbiological surveillance of cases of atypical pneumonia, especially for those with a history of recent travel to mainland China. However, ‘atypical pneumonia’, even where the disease is severe enough to warrant intensive care and intubation, is a common condition. One possibility considered was the emergence of an avian influenza A subtype H5N1 virus that had adapted to transmit efficiently from human to human, giving rise to a potentially pandemic situation. The World Health Organization Global Influenza Program was in a
50
state of heightened alert. One of the first unusual findings arising out of the enhanced surveillance of severe pneumonia in Hong Kong in early February was the detection of avian influenza A virus subtype H5N1 in two members of a family who had just returned from travelling to Fujian in mainland China. This was the first time since the ‘bird-flu incident’ in Hong Kong 1997 that human disease associated with influenza A H5N1 had been documented. A global pandemic alert was declared and preparations for vaccine development initiated.3 It subsequently became apparent that avian influenza was not the aetiological agent of the novel atypical pneumonia outbreak in Guangdong. But this global pandemic alert was not altogether unfounded — just months later, H5N1 avian influenza exploded across Asia with transmission to humans.4
Identification of the agent Efforts to identify the aetiological agent of this novel atypical pneumonia continued in Guangdong. Chlamydia had been detected in autopsy tissue of some patients dying of atypical pneumonia.5 However, the lack of a response to macrolide antibiotics argued against a role for chlamydia in this new disease. Other viruses such as influenza A (subtype H3N2) and adenovirus had been identified in some patients with suspected SARS.6 These influenza A virus isolates were genetically sequenced to ex-
Aetiology of SARS
clude the possibility that it may have acquired novel internal genes that may account for the unusual virulence for humans. However, they were found to be typical human H3N2 viruses with no basis for unusual virulence6 and therefore unlikely to be the cause for this novel disease presentation in Guangdong. In Hong Kong, individual patients with severe pneumonia were being investigated and a range of aetiological agents were identified, including adenovirus, Chlamydia psittaci, Mycoplasma pneumoniae and human influenza A and B viruses. These pathogens are to be expected from patients with atypical pneumonia and were unlikely to explain the unusual disease presentation being reported in Guangdong. By the end of February, there were as yet no significant clusters of disease in Hong Kong; the ‘cluster’ of patients arising from the transmission at Hotel M on 21 February 2003 were largely unrecognized in late February. This was because the secondary cases, with a few exceptions, had dispersed across the globe. However, the physician from Guangdong who was the index case of the Hotel M outbreak and his brother-in-law who had been in close contact with the index case were also hospitalized and investigated together with a number of other patients with pneumonia who had a history of recent travel to Guangdong. Initial microbiological investigations for known respiratory pathogens on these patients were unproductive. More invasive clinical specimens including a broncho-alveolar lavage and an open lung biopsy were obtained from the brother-inlaw for further investigation. The microbiological investigation moved on from the search for the known to a hunt for the unknown. Cell lines not commonly used to culture respiratory viruses were inoculated with clinical specimen extracts in the hope of growing an unconventional agent. PCR and RT-PCR methods for detecting known viruses as well as primers directed to conserved parts of the genome of virus genera of families to identify potentially novel
51
pathogens were tried. Electron microscopy on respiratory specimens, broncho-alveolar lavages and lung biopsy tissue was used. Serological tests were done on paired serum specimens with known viral antigens in the hope of detecting cross-reacting antibody responses to an antigenically related agent. By early March, an outbreak of nosocomial pneumonia was being reported in the French Hospital in Hanoi. Dr Carlo Urbani, a WHO epidemiologist who described and reported this outbreak and named the disease severe acute respiratory syndrome (SARS) himself succumbed to the disease (Chapter 1). Around 10 March 2003, a cluster of patients with pneumonia was reported from the Prince of Wales Hospital in Hong Kong. These two clusters in Hanoi and Hong Kong were the first definite clusters of this new disease to be apparent outside Guangdong province. In response, the WHO issued a global alert and provided a preliminary case definition of the disease, naming it severe acute respiratory syndrome (SARS). Soon, further cases were reported from Toronto and Singapore and WHO issued a travel advisory warning against non-essential travel to affected regions.7 Attempts to identify the aetiological agent of this unusual disease was a priority. On 18 March, WHO initiated a global network of laboratories investigating patients with suspected SARS.7 The specimens of patients from Vietnam were being investigated in reference laboratories in the USA (Centers for Disease Control, Atlanta), Japan and Paris. Patients in Singapore, Toronto and Hong Kong were being investigated by laboratories in each of these cities. This virtual laboratory network was linked by daily teleconferences and a secure website that was used to share findings of ongoing investigations on a ‘real-time’ basis.7 Collectively, laboratories within the WHO network were able to exclude a range of conventional respiratory pathogens including influenza, respiratory syncytial virus, parainfluenza, adenovirus, Mycoplasma
52
Severe Acute Respiratory Syndrome
pneumoniae, Nipah, Hendra and Hanta viruses, among others, as causes of this novel disease syndrome. Laboratories within the network were then using multiple strategies in the search for the aetiologic agent. Virus culture using a range of cell lines — HEp-2, MRC-5, MDCK, LLC-Mk2, HeLa, RDE, NCI-H292, HUT-292, B95-8, A549 among others — proved unproductive. However, a number of pathogens were detected from individual patients with suspected SARS by members of the WHO network. On 18 and 19 March, participating laboratories in Germany and Singapore reported particles with pleomorphic morphology compatible with paramyxoviruses detected by direct electron microscopy.7,8 Detection of human metapneumovirus (hMPV) in respiratory tract specimens of patients with SARS at the Prince of Wales Hospital in Hong Kong was reported on 18 March7,9 and subsequently (21 March) in Toronto.10 Detection of Chlamydia was reported from a clinical specimen from a patient in Germany.8 Rhinovirus had been detected by RT-PCR and virus culture in another patient with SARS.11 Between 21 and 24 March, three laboratories within the WHO network independently reported isolating a virus in FRhK-412 or Vero-E68,11 cells. At the University of Hong Kong, two virus isolates were initially obtained in FRhK-4 cells, from the lung biopsy of a 53-year-old male, the brother-in law of the index case of the Hotel M. cluster; and from the nasopharyngeal aspirate of a 42-year-old woman with pneumonia who had recently returned from travel to Guangzhou.12 The cytopathic effect on primary isolation was subtle, with a few refractory and rounded cells. The cytopathic effect did not appear to progress. However, the agent grew rapidly on passage onto fresh cell monolayers to produce full cytopathic effect within 24–48 hours. Mycoplasma contamination of the cell cultures (one cause of the cytopathic effect on cell cultures) was excluded. When acetone-fixed virus-infected cells were used
as antigen in an indirect immunofluorescent test, seroconversion was found in all eight suspected SARS patients but not in controls with atypical pneumonia diagnosed prior to the emergence of SARS. Thinsection electron microscopy of virus infected FRhK-4 cells revealed virus particles within the smooth-walled vesicles of the golgi-ER and on the cell surface. Virus particles of similar size were seen by electron microscopy within the lung biopsy of the 53-year-old male patient.12 At the Centers for Disease Control, USA, oropharyngeal specimens from a patient from Thailand who had acquired SARS infection in Hanoi, Vietnam were inoculated onto Vero E6 cells and NCI-H292 cells. A cytopathic agent was isolated in Vero-E6 cells with cytopathic effect appearing on day 5 post-inoculation and subsequently spreading to affect the whole cell sheet.11 A rhinovirus was also isolated in NCI-H292 cells from the same clinical specimen. In Frankfurt, a virus was isolated in Vero-E6 cells from the sputum collected on the day 7 of illness from a 32-year-old physician from Singapore who had treated one of the first patients with suspected SARS in Singapore.8 The cytopathic effect-causing agent failed to react with standard antisera used to identify influenza A, B, RSV, parainfluenza and adenoviruses. The spectrum of cells that supported the isolation and the type of cytopathic effect produced was not compatible with known respiratory viruses. Electron microscopy of negatively stained cell culture supernatants revealed coronavirus-like particles.8,11,12 Random primer RT-PCR8,12 and coronavirus group consensus primer11 strategies were used to amplify fragments of viral replicase RNA that had homology (based on the deduced amino acid sequence) with other known coronaviruses. However, the coronavirus associated with SARS was clearly genetically distinct from all other coronaviruses known hitherto. In an alternative approach to identify the virus isolate associated with SARS, total nucleic acid from virus-infected
Aetiology of SARS
Vero cells was hybridized with a pan-viral microarray.13 Eight oligonucleotides designed to represent coronavirus and astrovirus genetic sequences cross-hybridized to the nucleic acid of the SARS virus. The probes used for coronaviruses and astroviruses had fortuitous cross-homology and this explains the hybridization result to probes designed for both viruses.13 However, microarray could only be applied to virus already isolated in cell culture but not directly to the clinical specimen. Thus, by the end of March 2003, just a few weeks after the SARS virus had spread beyond Guangdong province, a novel coronavirus had been isolated by three separate laboratories from patients acquiring SARS in Hong Kong, Vietnam and Singapore. However, human metapneumovirus (hMPV) continued to be reported from two cohorts of patients, those from the Prince of Wales Hospital in Hong Kong and from Toronto. The hMPV was not consistently detected in other cohorts of patients with SARS.11,12 The genetic sequence of the hMPV virus detected seemed essentially similar to previously known strains of this virus and it was difficult to explain why an endemic human pathogen that had been in the human population for at least 50 (probably many more) years would suddenly give rise to a novel and unusual disease syndrome such as SARS.
The SARS CoV In order to establish conclusively the aetiologic role of the novel CoV with SARS, the virus had to be consistently found in the relevant clinical specimens of patients with the disease and not in healthy controls. The isolation of the virus in cell culture, the detection of viral RNA by RT-PCR based on initial genome sequences of the viral replicase gene and the demonstration of antibody responses by indirect immunofluorescence and enzyme immunoassays provided the tools to investigate for evidence of viral infection in patients with suspected
53
SARS.8,11,12 Using these approaches, infection with the novel CoV, subsequently to be named SARS CoV, was detectable in 45 of 50 consecutive patients with clinically suspected SARS in Hong Kong. All 32 patients from whom paired sera were available seroconverted to SARS CoV. Using RT-PCR, there was no evidence of this novel coronavirus infection in healthy controls or in patients with respiratory infection. Nor was antibody to the virus detectable in sera from 200 healthy blood donors in Hong Kong.12 In a separate investigation of 19 patients with SARS from Vietnam, Hong Kong and Singapore, evidence of SARS coronavirus was detected by RT-PCR or virus culture in all 19 patients investigated. All nine patients from whom serum was available had rising antibody titres to the virus.11 Further, 384 randomly selected US blood donor sera had no reactivity with SARS CoV in EIA assays.11 In a third study, 5 out of 5 patients with probable SARS and 3 out of 13 suspected SARS originating in Singapore and Vietnam were found to have evidence of SARS CoV RNA by RT-PCR.8 Collating all the results being reported to the WHO network from participating laboratories indicated that in most sites, SARS CoV was detectable in a proportion of patients with suspected SARS. However, hMPV was only consistently detected in two cohorts of patients, those at the Prince of Wales Hospital in Hong Kong and from Toronto. But it was subsequently shown that even these patients were co-infected with SARS CoV.14
Koch’s criteria The criteria postulated by Robert Koch need to be fulfilled to incriminate a pathogen as the cause of an infectious disease. These are: a) the demonstration of the pathogen consistently in patients with the disease and not in appropriate controls, and ideally, the demonstration of the microbe at the site of pathology, b) isolation of the pathogen in pure culture, c) the reproduction of the disease in an appropriate experimental model,
54
Severe Acute Respiratory Syndrome
and d) re-isolation of the pathogen from the site of pathology from diseased animals and the demonstration of a serological response to it. The findings above had demonstrated that a novel virus had been isolated from patients with SARS in pure culture and the virus had been consistently detected in patients with SARS but not in controls. The SARS CoV was detected in the lung biopsy and bronchoalveolar lavage of patients with SARS by virus culture, RT-PCR and electron microscopy.8,11,12 The lung pathology of patients with fatal SARS revealed bronchial epithelial denudation, loss of cilia, type 2 pneumocyte hyperplasia, and in those patients dying later in the course of the disease, multinucleate giant cells in the alveoli.11,15,16 SARS CoV RNA was demonstrated in the lungs by RT-PCR .15–17 Later, virus RNA and protein were also demonstrated in the lung tissue by in situ hybridization18–20 and immunohistology ( JM Nicholls and JSM Peiris, unpublished observations), respectively. No evidence of hMPV was detected in the lungs by RT-PCR, by electron microscopy or by a serological response in paired sera.15 The final requirements for the fulfilment of Koch’s postulates to establish the aetiology of an infectious disease were to reproduce the disease in a relevant animal model and to re-isolate the virus from the site of pathology and demonstrate an antibody response to the virus in these infected animals. Infection of cynomolgous macaques with an isolate of SARS CoV led to disease pathologically similar to that seen in human patients with SARS-epithelial necrosis, serosanguinaous alveolar exudates, hyaline membranes, type 2 pneumocyte hyperplasia and the presence of syncytia (see Chapter 9).14,21 The virus was successfully re-isolated from these lesions and an antibody response to the virus was demonstrated in the infected animals.22 In contrast, macaques infected with hMPV developed a mild suppurative rhinitis with minimal erosion in the infected airways.23
The possibility of co-infection with hMPV exacerbating disease caused by SARS CoV remained. However, the lack of evidence of hMPV infection by RT-PCR or serological methods in many patients with severe and even fatal SARS in whom SARS CoV was readily demonstrated suggested that hMPV was not necessary for the development of severe or fatal SARS.15 Infection of macaques with SARS CoV followed by hMPV did not produce enhanced disease.21 Thus, it was concluded that SARS CoV alone was both necessary and sufficient to cause the full syndrome of SARS.22 It still remains possible that co-infection with other pathogens enhance disease severity or transmissibility; however, it should be noted that hMPV was not the only co-infecting agent detected in patients with SARS.
SARS CoV in other species Coronaviruses are a genus in the order Nidovirales and are enveloped viruses with a single, positive-stranded RNA genome. Based on antigenic and genetic studies, previously known CoV are classified into three groups. These are: group 1 (including human CoV 229E and porcine transmissible gastroenteritis virus), group 2 (e.g. human CoV OC43; bovine coronavirus, mouse hepatitis virus) and group 3 (e.g. infectious bronchitis virus). SARS CoV showed some serological cross-reactivity with antisera raised to human CoV 229E (group 1 CoV). The partial genetic sequence information available initially indicated that the SARS coronavirus was distinct from previously known human and animal coronaviruses.8,11,12 By 12 April, the complete 29.7 kb long genome of the SARS CoV had been sequenced.24,25 Analysis of the genome demonstrated that it retained its distinctiveness from all other known coronviruses across the whole of its genome (Chapter 8) and hence was not derived through recombination between previously known animal or human coronaviruses. This also confirmed that SARS CoV was not a virus that
Aetiology of SARS
was artificially generated in the laboratory, as a bioterrorist agent. Given the lack of serological reactivity in the human population, it appeared that SARS CoV was probably recently introduced to the human population and not an endemic human virus that had recently mutated to unusual transmissibility and virulence. While extraterrestrial origins were adduced by some to explain its origin,26 the most probable explanation was that it was a hitherto unrecognized animal virus that had recently acquired the ability for human-to-human transmission. The earliest patients with presumed SARS that occurred in Foshan and Heyuan in Guangdong province, China in late 2002 had epidemiological links to the live game animal market trade,27,28 suggesting a possible origin for the virus that infected humans (see Chapter 11). In addition to causing disease in macaques, SARS CoV infects and replicates in cats, ferrets, mice, hamsters and African green monkeys.29–32 Virus-infected ferrets and cats can transmit infection to uninfected animals.29 These provide alternative animal models for investigation of SARS. Of these, only the macaque, hamster and ferret animal models demonstrate significant pathological disease.21,29
Characteristics of SARS CoV Identification of the aetiological agent of SARS allowed diagnostic tests to be developed and the characteristics of the virus to be investigated. SARS coronavirus was detected by culture and RT-PCR in the respiratory tract of patients with SARS. While viral RNA is detectable for over 3 weeks in many patients, the virus usually cannot be cultured after the third week of illness.33–35 SARS CoV can also be detected in the faeces and urine of infected patients by RT-PCR and culture and in the peripheral blood by RT-PCR methods,33–36 suggesting that SARS CoV results in a disseminated infection. Evidence of viral replication in the colonic mucosa was also demonstrated by electron
55
microscopic examination of biopsy material.37 The viral load in the respiratory tract and in the faeces progressively increased to peak in the second week of the illness, providing a possible explanation for the observation that transmission was less common in the first few days of illness.33,34,38 This pattern of viral-load kinetics in the course of the disease was in marked contrast to that of influenza39 where transmission of infection occurs early in the course of the disease. High viral load in the respiratory tract or in the serum is predictive of disease severity.40 SARS CoV is unusually stable in the environment. When dried on surfaces, the virus remains viable for many days at room temperature. It is also retains viability for days in faeces.41–43 The stability of the virus suggests that transmission may occur through fomites and indirect contact in health-care and community settings. It has been suggested that faecal contamination played a role in the large community outbreak at Amoy Gardens housing estate in Hong Kong.44
Conclusion The speed at which the aetiological agent of SARS was identified, its genome sequenced and diagnostic tests developed was remarkable in comparison with progress on other novel emerging infectious diseases, such as AIDS. The rapid progress of understanding was in part due to technological advances, although it should be noted that conventional virology as well as new technology proved indispensable in this enterprise. In addition, the WHO collaborative network of SARS laboratories within which findings were shared allowed emerging evidence to be evaluated quickly and provides a model for confronting emerging infectious diseases in future.
References 1 Zhong NS, Zeng GQ. Our strategies for fighting severe acute respiratory syndrome (SARS). Am J Respir Crit Care Med 2003;168: 7–9.
56
Severe Acute Respiratory Syndrome
2 World Health Organization. Acute respiratory syndrome, China. Wkly Epidemiol Rec 2003; 78: 41. 3 World Health Organization. Influenza A (H5N1), Hong Kong, Special Administrative Region of China. Wkly Epidemiol Rec 2003;78: 49–50. 4 World Health Organization. Avian influenza A (H5N1). Wkly Epidemiol Rec 2004;79: 65–70. 5 Hong T, Wang JW, Sun YL et al. Chlamydialike and coronavirus-like agents found in dead cases of atypical pneumonia by electron microscopy. Zhonghua Yi Xue Za Zhi 2003;83: 632–6. 6 Zhong NS, Zheng BJ, Li YM et al. Epidemiological and aetiological studies of patients with severe acute respiratory syndrome (SARS) from Guangdong in February 2003. Lancet 2003;362: 1353–8. 7 World Health Organization. A multicentre collaboration to investigate the cause of severe acute respiratory syndrome. Lancet 2003; 361: 1730–3. 8 Drosten C, Gunther S, Preiser W et al. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. N Engl J Med 2003;348: 1967–76. 9 Chan PKS, Tam JS, Lam CW et al. Human metapneumovirus detection in patients with severe acute respiratory syndrome. Emerg Infect Dis 2003;9: 1058–63. 10 Poutanen SM, Low DE, Henry B et al. Identification of severe acute respiratory syndrome in Canada. N Engl J Med 2003;348: 1995–2005. 11 Ksiazek TG, Erdman D, Goldsmith C et al. A novel coronavirus associated with severe acute respiratory syndrome. N Engl J Med 2003;348: 1953–66. 12 Peiris JS, Lai ST, Poon LL et al. Coronavirus as a possible cause of severe acute respiratory syndrome. Lancet 2003;361: 1319–25. 13 Wang D, Coscoy L, Zylberg M et al. Microarray based detection and genotyping of viral pathogens. Proc Natl Acad Sci USA 2002;99: 15687–92. 14 Kuiken T, Fouchier RAM, Schutten M et al. Newly discovered coronavirus as the primary cause of severe acute respiratory syndrome. Lancet 2003;362: 263–70. 15 Nicholls JM, Poon LL, Lee KC et al. Lung pathology of fatal severe acute respiratory syndrome. Lancet 2003;361: 1773–8. 16 Franks TJ, Chong PY, Chui P et al. Lung pathology of severe acute respiratory syndrome (SARS): a study of 8 autopsy cases from Singapore. Hum Pathol 2003;34: 43–8. 17 Mazzulli T, Farcas GA, Poutanen SM et al. Severe acute respiratory syndrome–associated coronavirus in lung tissue. Emerg Infect Dis 2003;10: 20–4.
18 Nakajima N, Asahi-Ozaki Y, Nagata N et al. SARS Coronavirus-infected cells in lung detected by new in situ hybridization technique. Jpn J Infect Dis 2003;56: 139–41. 19 Chong PY, Chui P, Ling AE et al. Analysis of deaths during the severe acute respiratory syndrome (SARS) epidemic in Singapore. Arch Pathol Lab Med 2004;128: 195–204. 20 To KF, Tong JH, Chan PK et al. Tissue and cellular tropism of the coronavirus associated with severe acute respiratory syndrome: an in-situ hybridization study of fatal cases. J Pathol 2004;202: 157–63. 21 Fouchier RAM, Kuiken T, Schutten M et al. Koch’s postulates fulfilled for SARS virus. Nature 2003;423: 240. 22 Osterhaus AD, Fouchier RA, Kuiken T. The aetiology of SARS: Koch’s postulates fulfilled. Philos Trans R Soc Lond B Biol Sci 2004;359: 1081–2. 23 Kuiken T, van den Hoogen BG, van Riel DA et al. Experimental human metapneumovirus infection of cynomolgus macaques (Macaca fascicularis) results in virus replication in ciliated epithelial cells and pneumocytes with associated lesions throughout the respiratory tract. Am J Pathol. 2004;164: 1893–900. 24 Marra MA, Jones SJ, Astell CR et al. The genome sequence of the SARS associated coronavirus. Science 2003;300: 1399–404. 25 Rota PA, Oberste MS, Monroe SS et al. Characterization of a novel coronavirus associated with severe acute respiratory syndrome. Science 2003;300: 1394–9. 26 Wickramasinghe C, Wainwright M, Narlikar J. SARS — a clue to its origins? Lancet 2003;361: 1832. 27 Breiman RF, Evans MR, Priser W et al. Role of China in the quest to define and control severe acute respiratory syndrome. Emerg Infect Dis 2003;9: 1037–41. 28 Xu RH, He JF, Evans MR et al. Epidemiologic clues to SARS origin in China. Emerg Infect Dis 2004;10: 1030–7. 29 Martina BE, Haagmans BL, Kuiken T et al. Virology: SARS virus infection of cats and ferrets. Nature 2003;425: 915. 30 Subbarao K, McAuliffe J, Vogel L et al. Prior infection and passive transfer of neutralizing antibody prevent replication of severe acute respiratory syndrome coronavirus in the respiratory tract of mice. J Virol 2004;78: 3572–7. 31 Buchholtz UJ, Bukreyev A, Yang L et al. Contributions of the structural proteins of severe acute respiratory syndrome coronavirus to protective immunity. Proc Natl Acad Sci USA, 2004;101: 9804–9. 32 Bukreyev A, Lamirande EW, Buchholz UJ et al. Mucosal immunization of African green
Aetiology of SARS
33
34
35
36
37
38
39
monkeys (Cercopithicus aethiops) with an attenuated parainfluenza virus expressing the SARS coronavirus spike protein for the prevention of SARS. Lancet 2004;363: 2122–7. Peiris JS, Chu CM, Cheng VC et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;361: 1767–72. Chan KH, Poon LLM, Cheng VCC et al. Detection of SARS coronavirus (SCoV) by RT-PCR, culture, and serology in patients with acute respiratory syndrome (SARS). Emerg Infect Dis 2004;10: 294–9. Chan PKS, To WK, Ng KC et al. Laboratory diagnosis of SARS. Emerg Infect Dis 2004;10: 825–31. Ng EK, Ng PC, Hon KL et al. Serial analysis of the plasma concentration of SARS coronavirus RNA in pediatric patients with severe acute respiratory syndrome. Clin Chem 2003;49: 2085–8. Leung WK, To KF, Chan PK et al. Enteric involvement of severe acute respiratory syndrome-associated coronavirus infection. Gastroenterology 2003;125: 1011–17. Cheng PK, Wong DA, Tong LK et al. Viral shedding patterns of coronavirus in patients with probable severe acute respiratory syndrome. Lancet 2004;363: 1699–700. Kaiser L, Briones MS, Hayden FG. Performance of virus isolation and Directigen FluA
40
41
42
43
44
57
to detect influenza A virus in experimental infection. J Clin Virol 1999;14: 191–7. Tsang OT, Chau TN, Choi KW et al. Coronavirus-positive nasopharyngeal aspirate as predictor for severe acute respiratory syndrome mortality. Emerg Infect Dis 2003;9: 1381–7. World Health Organization. First data on stability and resistance of SARS coronavirus compiled by members of the WHO laboratory network. Geneva: World Health Organization., 2003. Accessed 16 August 2004 at http://www.who.int/csr/sars/survival_ 2003_05_04/en/index.html Rabenau HF, Cinatl J, Morgenstern B et al. Stability and inactivation of SARS coronavirus. Med Microbiol Immunol (Berl) 2005;194: 1–6. Duan SM, Zhao XS, Wen RF et al. SARS Research Team. Stability of SARS coronavirus in human specimens and environment and its sensitivity to heating and UV irradiation. Biomed Environ Sci 2003;16: 246–55. Department of Health, Hong Kong. An outbreak of Severe Acute Respiratory Syndrome at Amoy Gardens Kowloon Bay, Hong Kong, Department of Health; Hong Kong Special Administrative Region. (Published in April 2003, available from: http://www.info.gov. hk/info/ap/pdf/amoy_e.pdf)
Chapter 8
Structure of the Genome of SARS CoV Paul A Rota, Xin Liu, Byron T Cook and Suxiang Tong
Genome structure and organization A novel coronavirus, SARS CoV, was identified as the aetiological agent responsible for severe acute respiratory syndrome (SARS).1 Within a month of the initial isolation, the complete genomic sequences of two strains of SARS CoV were reported.2,3 At the time of writing, more than 100 complete genome sequences representing isolates obtained from different phases of the outbreak and from several countries are available on the public databases. The structure of the genome of SARS CoV is that of a typical coronavirus.4 The genome of SARS CoV is approximately 29.7 kb in length and consists of a single-stranded, positive sense RNA that has a 5’ cap and 3’ poly-A tail. The 265 nucleotide untranslated region at the 5’ terminus of the genome contains the conserved coronavirus core leader sequence and the 342 nucleotides at the 3’ terminus include a 32base pair region corresponding to the conserved s2m motif (stem-loop II-like motif).5 This has been identified as a common feature in astroviruses.2,6 The gene order of SARS CoV is: 5’-replicase-spikeenvelope-membrane-nucleocapsid-3’6 (Fig. 8.1) and is similar to the other coronaviruses. The genome of the Urbani strain contains a total of 14 open reading frames (ORFs). Ten of the open reading frames are predicted to encode 24 putative nonstructural and accessory proteins after post-
58
translational processing (Table 8.1).6,7 In addition, four open reading frames encoding the four structural proteins S (spike), E (envelope), M (membrane) and nucleocapsid (N) are located downstream of the replicase gene in the 3’ proximal part of the genome (Fig. 8.1). A haemagglutinin esterase gene, a common feature of the group 2 coronaviruses, was not found in the SARS CoV genome. The genomes of coronaviruses contain a number of non-structural genes of unknown function. Besides the replicase gene and the genes coding for the four structural proteins, the genome of the Urbani strain of SARS CoV contains eight additional open reading frames. Although the functions of these proposed non-structural proteins are not known, recent studies have shown that at least two of these non-structural proteins are expressed in infected cells.8,9 SARS CoV replication employs a sophisticated transcriptional, translational and post-translational regulatory mechanism and the genomic termini are part of the regulatory elements for virus replication. After the genome of SARS CoV enters the cytoplasm of the cell, replication begins with translation of two large overlapping open reading frames (ORF1a and ORF1b) that encode the replicative enzymes. A-1 ribosomal frameshift that depends on a 7nucleotide slippery sequence (5’UUUAAAC3’) and a secondary RNA pseudoknot structure that are located near the 3’ terminus of ORF1a are necessary for the
Structure of the Genome of SARS CoV
5’ cap
PLP2
3CL
HEL
S
E
M
N
59
3’ RNA
1 29Kb
Genomic RNA
ORF 1a
nsp 1- 11
polyA
nsp 12-16 ORF 1b
3’
Minus strand RNA
5’ (-) 2 3 4 5 6 7 8 9
8.3 kb 4.6 kb 3.8 kb 3.5 kb 3.0 kb 2.6 kb 2.1 kb 1.8 kb
Subgenomic RNAs 5’ Leader
3’ TRS
Figure 8.1 Genomic organization of SARS CoV and generation of (-) strand RNA and (+) subgenomic mRNAs. The 29 727 genomic RNA is predicted to encode non-structural ORFs 1a (nsp1–11) and 1b (nsp 12–16) and S, M, E, N structural proteins. The genomic RNA is 29 727 bp in length and is transcribed into minus strand RNA first. The (-) strand RNA is predicted to generate eight subgenomic mRNAs which all have the same 5’ leader and 3’ untranslated regions as that of the genomic RNA.
translation of ORF1b.4,10 ORF1a and ORF1b comprise two-thirds of the virus genome and code for large polyproteins that undergo extensive post-translational processing to generate the RNA-dependent RNA polymerase, viral proteases and RNA helicase. The replicase genes of most other coronaviruses code for two papain-like cysteine proteinases (PLP1 and PLP2) which cleave the NH2-proximal part of the polyprotein along with a 3C-like cysteine proteinase which cleaves the COOHproximal part of the replicase polyprotein. The replicase gene of SARS CoV contains an orthologue of PLP2, but not PLP1. Alignments of the predicted amino acid sequences of other coronaviruses have been used to infer the cleavage sites on the 1a and 1b polyproteins and experimental data suggest these inferences are correct.6 In total, these two polyproteins can be cleaved into 16 non-structural proteins that form a viral replication complex. There are nine virally encoded RNA species in cells infected with SARS CoV including the genomic RNA. Northern blot-
ting, RT-PCR amplification, and sequence analysis of the SARS CoV mRNA indicated that there are eight, 3’ co-terminal subgenomic mRNAs (Fig. 8.1). A conserved transcription-regulating sequence (TRS), 5’ACGAAC-3’, is located immediately upstream of the open reading frames on each viral RNA. ORF1a and ORF1b are translated directly from genomic RNA, while the remaining proteins are translated from subgenomic RNAs. A 72 nucleotide leader sequence is found at the 5’ terminus of both subgenomic and genomic RNAs. The favoured model for synthesis of subgenomic mRNAs proposes that discontinuous transcription occurs during synthesis of the negative sense strands. Negative sense copies of the subgenomic RNAs containing a complementary copy of the leader sequence at their 3’ termini then serve as templates for synthesis of positive sense, subgenomic mRNAs .4,10 Comparison of the nucleotide and predicted amino acid sequences of SARS CoV to the sequences of other coronaviruses indicated that SARS CoV was a truly novel
60
Severe Acute Respiratory Syndrome
Table 8.1 Predicted open reading frames and products of SARS CoV Proteins
Protein size (aa)
Predicted product
Open reading framea
Corresponding RNAc
nsp1 nsp2 nsp3 nsp4 nsp5 nsp6 nsp7 nsp8 nsp9 nsp10 nsp11 nsp12 nsp13 nsp14 nsp15 nsp16 S sars3a sars3b E M sars6 sars7a sars7b sars8a sars8b N Sars9b
180 638 1922 500 306 290 83 198 113 139 13 932 601 527 346 298 1256 275 155 77 222 64 123 45 40 85 423 99
Putative leader protein Putative counterpart of MHV p65 Putative coronavirus nsp1 — PL2 TM2 3CL TM3 Putative coronavirus nsp4 Putative coronavirus nsp5 Putative coronavirus nsp6 GFL ?d RdRp ZD, NTPase,HEL1 Exonuclease NTD, endoRNase 2’-O-MT Spike protein precursor ? (¥1)b ? (¥2) b Envelope protein Membrane protein ? (¥3) b ? (¥4) b ? ? ? (¥5) b Nucleocapsid ?
ORF1a ORF1a ORF1a ORF1a ORF1a ORF1a ORF1a ORF1a ORF1a ORF1a ORF1a ORF1b ORF1b ORF1b ORF1b ORF1b ORF2 ORF3a ORF3b ORF4 ORF5 ORF6 ORF7a ORF7b ORF8a ORF8b ORF9a ORF9b
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 3 3 4 5 6 7 7 8 8 9 9
a
ORF1a was referred to as Replicase 1A and ORF1b was referred to as Replicase 1B by Marra et al.14 ¥ was used by Rota et al.16 c RNA number 1 is the genomic RNA and numbers 2–9 refer to subgenomic RNAs. d ? Function unknown.
b
coronavius and that no part of its genome was derived by recombination with another known virus. Sequence analysis did not provide any obvious clues for the origin of SARS CoV. Pairwise amino acid homologies with other coronaviruses were very low for the predicted structural genes (18–39%), but slightly higher for the predicted polymerase and helicase domains of the replicase gene (58–68%). Initial phylogenetic analyses suggested that SARS CoV was not closely related to any of the other coronaviruses and possibly represented a fourth group.2,3 Sub-
sequent analysis using a tree rooted in equine toroviruses suggested that SARS CoV is distantly related to group 2 coronaviruses and may have been an early split-off from the main group 2 branch.6 The genome contains features found in both mammalian and avian coronaviruses.11
Genetic variation Although the origin of SARS CoV is still unknown, molecular epidemiologic studies have provided a great deal of information
Structure of the Genome of SARS CoV
about the course of the outbreak.12–18 A recent analysis of 63 complete genomic sequences showed that the viruses could be divided into three groups based on the temporal phase of the outbreak.12 Early phase sequences were obtained from animal isolates or isolates from sporadic cases representing recent animal to human infection in Guangdong Province, China. Middle phase sequences represented isolates associated with human to human transmission in Guangdong Province prior to its first spread to Hong Kong, China. Late phase sequences primarily came from isolates linked to events after the Metropole Hotel cluster in Hong Kong.12 Though 299 single nucleotide variations were noted among the 63 sequences, nucleotide substitutions at 5 positions in the genome were considered to be characteristic of viruses circulating during each phase.12 These five substitutions occurred at positions 17 564, 21 721, 22 222, 23 823 and 27 827 relative to the GZ02 reference sequence. Three of these positions are in the S gene, one falls within the ORF1b polyprotein and the other is within ORF8a. Only one signature variation distinguishes early from middle phase sequences, while four signature variations distinguish middle from late phase sequences (Plate 1). Exceptions do exist, but those sequences are considered transitional or divergent from the main chain of transmission and do not define other groupings of sequences. Overall, there was relatively more nucleotide variation observed in early than in the middle and late phase sequences.12–14,19 Most studies of sequence variation on SARS CoV have focused on substitutions present in more than one isolate to correct for substitutions that were present in a minority of the quasi-species population, were detected by chance, were acquired during passage in tissue culture or represented PCR artifacts.7,12,13,15,19,20 It was estimated that SARS CoV RNA-dependent polymerase misincorporates 8.26 bases per million which is similar to the rates obtained for other RNA viruses.12,19
61
However, the rate of nucleotide change observed during the SARS outbreak was much lower. The ratio of non-synonymous to synonymous substitutions was higher during the early phase of the outbreak compared to the later phases suggesting that pressure to select for amino acid changes decreased during the outbreak.12,14,19 It is possible that early isolates were being selected for increased ability to replicate and transmit in humans. Statistical analysis of nonsynonymous and synonymous substitution rates showed that the ratio dropped from being greater than one to less than one, further indicating the overall stabilization of the virus over the course of the outbreak.12,14,19 It should be noted that some of the homogeneity of later SARS sequences may be related to the preponderance of cases linked to spread from a single patient in the Metropole hotel,12,13,21 although sequence information suggests that there was more than one introduction of SARS CoV into Hong Kong from Guangdong.15,22 The rate of nucleotide substitutions within the SARS CoV genome varied considerably. The nucleocapsid gene had the lowest rate of change among the structural proteins as multiple isolates differed by only 0.079%.14 The non-structural coding sequence, ORF1ab, was also relatively stable with a substitution rate of 0.09%. Regions of ORF1ab containing predicted enzymatic activities show greater sequence homology to other coronaviruses than do the genes encoding the structural or nonstructural proteins.2,3 In fact, no variation was detected in Orf1ab between positions 14 807 and 16 325. The spike glycoprotein has two, predicted functional domains, S1 and S2. In other coronaviruses, S1, which contains the majority of the surface epitopes as well as the receptor-binding domain, can be variable, containing varying numbers of deletions and substitutions.16,23,24 The SARS S1 region had a substitution rate of only 0.219%, possibly because of the lack of prior immunity in most humans and, therefore, lack of
62
Severe Acute Respiratory Syndrome
immunological pressure that would select for antigenic changes.12,14 The genes for the other two structural proteins, E and M, had substitution rates of 0.433% and 0.333% respectively. The highest rate of nucleotide change, 3.333%, was seen in ORF8 and was due in part to the presence of major deletions including 82-nt and 29-nt deletions from isolates of early phase, and 415-nt deletion from isolates from the late phase.12 The predicted antiparallel reverse symmetrical sequences might contribute to the observed high deletion rate in ORF8. The 29-nt deletion has been a source of speculation relative to the origin of SARS CoV and its adaptation to humans. It appears that one likely source of introduction to humans from animals was associated with wild-animal markets in southern China17,25 where SARS coronaviruses exhibiting 99.8% homology to human viruses were isolated from Himalayan palm civets (Paguma larvata) and a raccoon dog (Nyctereutes procyonoides). This is more evident when one sequence from the later small cluster of SARS (n = 4) in Guangdong, December 2003, in which direct or indirect contact with caged animals was implicated, shows more homology (99.9%) to animal isolates (unpublished data). One striking feature of the animal viruses is a string of 29 nucleotides found in ORF8, near the 3’ terminus of the genome. With the exception of both early human cases from late 2002 and late 2003, all isolates from human cases lacked this 29 nucleotide sequence. This suggested that a deletion event, which occurred during human transmission of the virus, resulted in the loss of these 29 nucleotides and truncation of the predicted protein. The functional significance of the truncation, or lack thereof, remains to be explained.
References 1 Ksiazek TG, Erdman D, Goldsmith CS et al. A novel coronavirus associated with severe acute respiratory syndrome. N Engl J Med 2003;348: 1953–66.
2 Marra MA, Jones SJ, Astell CR et al. The genome sequence of the SARS-associated coronavirus. Science 2003;300: 1399–404. 3 Rota PA, Oberste MS, Monroe SS et al. Characterization of a novel coronavirus associated with severe acute respiratory syndrome. Science 2003;300: 1394–9. 4 Holmes KV, Lai MC. Coronaviridae: the viruses and their replication. In: Fundamental Virology, 3rd edn (Fiends BN, Knipe DM and Howley PM, eds). Lippincott-Raven, Philadelphia, 1996: 541–54. 5 Jonassen CM, Jonassen TO, Grinde B. A common RNA motif in the 3’ end of the genomes of astroviruses, avian infectious bronchitis virus and an equine rhinovirus. J Gen Virol 1998;79: 715–18. 6 Snijder EJ, Bredenbeek PJ, Dobbe JC et al. Unique and conserved features of genome and proteome of SARS-coronavirus, an early split-off from the coronavirus group 2 lineage. J Gen Virol 2003;84: 2305–15. 7 Brown EG, Tetro JA. Comparative analysis of the SARS coronavirus genome: a good start to a long journey. Lancet 2003;361: 1756–7. 8 Fielding BC, Tan YJ, Shuo S et al. Characterization of a unique group-specific protein (U122) of the severe acute respiratory syndrome coronavirus. J Virol 2004;78: 7311–18. 9 Tan YJ, Teng E, Shen S et al. A novel severe acute respiratory syndrome coronavirus protein, U274, is transported to the cell surface and undergoes endocytosis. J Virol 2004;78: 6723–34. 10 Thiel V, Ivanov KA, Putics A et al. Mechanisms and enzymes involved in SARS coronavirus genome expression. J Mol Biol 2003;331: 991–1004. 11 Stavrinides J, Guttman DS. Mosaic evolution of the severe acute respiratory syndrome coronavirus. J Virol 2004;78: 76–82. 12 Chinese SARS Molecular Epidemiology Consortium. Molecular evolution of the SARS coronavirus during the course of the SARS epidemic in China. Science 2004;303: 1666–9. 13 Ruan YJ, Wei CL, Ee AL et al. Comparative fulllength genome sequence analysis of 14 SARS coronavirus isolates and common mutations associated with putative origins of infection. Lancet 2003;361: 1779–85. 14 Wang ZG, Li LJ, Luo Y et al. Molecular biological analysis of genotyping and phylogeny of severe acute respiratory syndrome associated coronavirus. Chin Med J 2004;117: 42–8. 15 Guan Y, Peiris, JSM, Zheng B et al. Molecular epidemiology of the novel coronavirus that causes severe acute respiratory syndrome. Lancet 2004;363: 99–104. 16 Fazakerley JK, Parker SE, Bloom F et al. The V5A13.1 envelope glycoprotein deletion mutant of mouse hepatitis virus type-4 is
Structure of the Genome of SARS CoV
17
18
19
20
neuroattenuated by its reduced rate of spread in the central nervous system. Virology 1992;187: 178–88. Zhong NS, Zheng BJ, Li YM et al. Epidemiology and cause of severe acute respiratory syndrome (SARS) in Guangdong, People’s Republic of China, in February. Lancet 2003;362: 1353–8. Xu D, Zhang Z, Chu F et al. Genetic variation of SARS coronavirus in Beijing hospital. Emerg Infect Dis 2004;10: 789–94. Yeh SH, Wang HY, Tsai CY et al. National Taiwan University SARS Research Team. Characterization of severe acute respiratory syndrome coronavirus genomes in Taiwan: molecular epidemiology and genome evolution. Proc Natl Acad Sci USA 2004;101: 2542–7. Tong S, Lingappa JR, Chen Q et al. Direct sequencing of SARS–coronavirus S and N genes from clinical specimens shows limited variation. J Infect Dis 2004;190: 1127–31.
63
21 Centers for Disease Control and Prevention. Severe acute respiratory syndrome — Singapore, 2003. MMWR Morb Mortal Weekly Rep 2003;52: 405–11. 22 Chim SS, Tsui SK, Chan KC et al. Genomic characterization of the severe acute respiratory syndrome coronavirus of Amoy Gardens outbreak in Hong Kong. Lancet 2003;362: 1807–8. 23 Hingley ST, Gombold JL, Lavi E et al. MHV-A59 fusion mutants are attenuated and display altered hepatotropism. Virology 1994;200: 1–10. 24 Pratelli A, Martella V, Decaro N et al. Genetic diversity of a canine coronavirus detected in pups with diarrhoea in Italy. J Virol Methods 2003;110: 9–17. 25 Guan Y, Zheng BJ, He QY et al. Isolation and characterization of viruses related to the SARS coronavirus from animals in southern China. Science 2003;302: 276–8.
Chapter 9
Viral Diagnosis of SARS C Drosten, KH Chan and LLM Poon
The necessity for diagnostic testing The recognition that a novel coronavirus was the cause of SARS provided the basis for introducing specific laboratory diagnostic tests.1–3 Such tests are necessary for two reasons. Firstly, early identification and isolation of patients with SARS are important for appropriate clinical management, interruption of virus transmission and for minimizing hospital cross-infection. Secondly, laboratory tests are required to define the epidemiology, ecology and pathogenesis of a novel disease. In the absence of epidemiological linkage, SARS cannot be differentiated from other causes of atypical pneumonia on purely clinical grounds and thus diagnostic tests that can identify patients with SARS as early as possible are an advantage. As with any virus infection, the options for laboratory diagnosis include detection of the virus or its components and detection of a serological response in the host. In principle, virus detection can be accomplished by isolation in culture, virus antigen detection or molecular detection of viral nucleic acid.
Methods of virus detection RT-PCR The first methods for detecting SARS CoV in clinical samples were described along with
64
the identification of the agent.1–4 The firstgeneration tests that were used in these studies relied on classical RT-PCR methodology, i.e. RNA extraction, reverse transcription and PCR amplification, followed by detection of PCR products in agarose gel electrophoresis. The replicase gene of SARS CoV was the first part of the genome to be genetically sequenced and these first-generation RT-PCR tests used it as the target gene. In a study in Hong Kong, first-generation RT-PCR was positive in nasopharyngeal aspirate or throat swab specimens in around 30% of patients within the first 4 days of disease onset. The positive rate progressively increased to peak at around 80% for nasopharyngeal aspirates by days 11–12 after disease onset and fell thereafter. Faeces became positive slightly later in the disease (i.e. by day 6) but subsequently yielded higher rates positive with over 90% of specimens positive by day 11–12 post-onset (Fig. 9.1). Viral RNA was still detectable in some patients for longer than a month. Viral RNA was detected in the urine but rates of positivity were much lower. Deep respiratory specimens such as sputum and endotracheal aspirates yielded higher rates of RT-PCR positives in the early stages of illness.5,6 However, many patients with SARS usually do not have a productive cough in the early stage of the illness. Testing two samples per patient led to a small increase in detection rate. A similar improvement occurred if the same specimen was tested twice, increasing the chance of detection of
Viral Diagnosis of SARS
65
100% 90% 80%
% Positive
70% 60%
N PA
50%
N S/T S Stool U r ine
40% 30% 20% 10% 0% 0- 2
3- 4
5- 6
7- 8
9- 10
11- 12
13- 14
15- 16
17- 18
19- 22
23- 26
27- 30
31- 40
41- 50
> 50
7 ND 56 59
0 ND 31 42
3 ND 17 35
D ays after onset
NPA NS/TS Stool Uri ne
30 9 5 5
47 19 10 2
48 31 10 8
31 27 15 18
36 16 19 22
20 9 30 18
No. of specimens tested 10 23 29 43 40 7 63 39 22 57 58 26
11 17 47 62
10 7 37 49
8 7 32 52
Figure 9.1 RT-PCR % positive in nasopharyngeal aspirates (NPA), nose and throat swabs (NS/TS), stool and urine at different days post-onset of disease in patients with serologically confirmed SARS. (Re-printed from KH Chan et al 2004; with permission of Emerging Infectious Diseases.)
randomly distributed RNA in low concentration samples.7 Real-time RT-PCR is potentially more sensitive and has the advantage of being a closed system with less potential for giving rise to problems with PCR contamination. It also permits viral load quantitation.2 Quantitative assays of viral RNA in respiratory and faecal specimens indicate that viral load is low in the first few days of illness and progressively increases to peak early in the second week of the disease.6,8 This explains why conventional RT-PCR methods were poor for detection of virus in the first few days of illness, when laboratory diagnosis is most critical. The sensitivity of the RT-PCR assays was improved by using a larger volume of the clinical specimen for RNA extraction leading to sensitivity increasing from 22% to 44% in 50 nasopharyngeal aspirate samples collected in the first 3 days of illness.9 However, this approach may not be suitable with all RT-PCR assays and all specimen types because of the presence of
background nucleic acids and PCR inhibitors.10 Evaluation for RT-PCR inhibitors is important and may be achieved by parallel testing of spiked patient samples or by use of internal inhibitor controls in the assays.11 By using the enhanced RNA extraction method together with real-time RT-PCR techniques, the sensitivity of detection in nasopharyngeal aspirates in the first 3 days of illness was increased to around 80%.9 In these same RNA extracts, conventional first-generation RT-PCR only had a sensitivity of 44%. Another option for increasing sensitivity of molecular detection was to use an alternative target in the virus genome. During replication, coronaviruses generate an excess of subgenomic mRNA species that all contain the nucleocapsid (N) gene sequence.12 If virus-replicating cells are present in clinical samples, one would therefore expect a higher sensitivity in N gene-based RT-PCR. Alhough this hypothesis was initially supported by animal
66
Severe Acute Respiratory Syndrome
experiments,13 it could not be confirmed in three independent studies on SARS patients. Two studies comparing different sets of RT-PCR assays for detection of viral replicase and nucleocapsid RNA in nasopharyngeal aspirate samples and serum from patients with SARS failed to demonstrate a significant difference in sensitivity between the two viral gene targets.14,15 In a third study where 66 acute-phase (mainly respiratory and stool) samples from confirmed German, Singaporean, and Hong Kong patients (Table 9.1) were tested, there was no difference in sensitivity achieved with two optimized commercial replicase-based tests and an optimized N-gene assay (p = 0.68 in ANOVA analysis) (Table 9.1).16 These observations suggested that the subgenomic mRNA exists in low abundance in routine
clinical specimens. Indeed, a recent in vitro study demonstrated that most of subgenomic mRNA molecules in infected cells are rapidly degraded.16 Blood is a less commonly used specimen for diagnosis of respiratory virus infections using RT-PCR. Using first generation SARS CoV RT-PCR methods, none of 86 acute phase sera from confirmed SARS patients were positive.17 Quantitative real-time RTPCR assays showed that the viral load (190 copies per ml) in plasma may be below the detection limit of first-generation tests.2 However, using sensitive real-time RT-PCR assays, tests of serial plasma specimens from 12 SARS patients yielded detection rates of 50% on the day of admission to hospital, 50% on day 7, and 25% on day 14 of fever, respectively.15 In a different patient cohort,
Table 9.1 Clinical sensitivities of three different SARS-CoV RT-PCR assays (from Drosten et al.16) Sensitivity in confirmed patients Assay target gene
Artus kit Replicase
Roche kit Replicase
In-house Nucleocapsid
Samples (n = 66a) Upper respiratoryb
11/19
10/18e
11/19
Lower respiratoryc
12/12
12/12
12/12
Stool
18/22e
18/23
20/23
Plasma
1/7
2/7
3/7
Otherd
4/5
1/4e
3/5
Total 95% CI aSamples
70.8% (59–82%)
67.1% (55–79%)
74.2% (63–85%)
Target concentrationf Replicaseg Median (range)
Nucleocapsid
5.5E2 (1.1E0–7.8E3) 1.2E6 (2.1E4–1.2E10) 4.3E4 (3.1E1–1.0E8) 3.9E3 – 1.5E2 (3.1E2–4.7E4) 2.0E4 (1.1E0–1.2E10)
5.2E2 (2.5E1–1.15E4) 2.8E6 (1.5E4–4.7E9) 5.5E4 (2.5E1–2.1E7) 5.9E2 (7.2E1–5.6E3) 4.1E2 (8.1E1–2.6E2) 2.5E4 (2.5E1–4.7E9)
from 29 laboratory-confirmed SARS patients (25 patients with seroconversion in immunofluorescence assay, 4 patients with positive RT-PCR in three independent assays [equivalent WHO criteria for laboratory confirmation of SARS). The majority of these samples have been contributed by Evelyn Koay, National University of Singapore. b Three saliva samples, 16 nasopharyngeal swabs. cEight sputum samples, two endotracheal aspirate specimens, two brocho-alveolar lavage samples. dOne jejunectomy sample, one dialysis fluid specimen, three urine samples. eNo result for one sample due to failed internal control. Sample omitted from sensitivity evalution for this assay. fCopies per ml or per swab specimen, median (range). gDetermined by Artus kit.
Viral Diagnosis of SARS
the same investigators found up to 78–87% of serum samples from patients collected a mean of 2.6 days post-fever onset tested positive. Similar findings were also reported by others.18 RT-PCR has been found to be more sensitive in plasma than in serum for other viruses.19,20 With SARS, this remains to be further defined by comparing serum and plasma in the same patient cohort. Viral load as determined by quantitative real-time RT-PCR in blood15 and nasopharyngeal aspirates21 have been shown to be predictive of disease severity. It should be noted that virus RNA in the serum can only be detected for a limited period of time during the course of the illness.22 Commercial RT-PCR kits are now available, assuring more reliable results by standardization and incorporation of internal controls (Table 9.1).23 Alternative molecular amplification methods such as LAMP (loop mediated isothermal amplification assay) can provide an option for detection that does not require expensive equipment (e.g. thermal cyclers) and may be more suited for the developing world. SARS CoV can be detected using LAMP with a sensitivity comparable to the first-generation RTPCR assays but less sensitive than current real-time RT-PCR assays.24,25
The appropriate choice of specimen for diagnosis of SARS by RT-PCR Early in the illness, the best specimens for molecular diagnosis are respiratory specimens (nasopharyngeal aspirate or throat swab as second best), faecal specimen and serum or plasma. Nasopharyngeal aspirates and throat swabs are specimens that have been extensively used for diagnosis of patients with SARS. However, there has been concern that collection of nasopharyngeal aspirates poses an infection risk for healthcare workers collecting such specimens. On the other hand, in periods of low SARS activity, most patients will have alternative diagnoses of viruses such as influenza, adenovirus or parainfluenza. The best op-
67
tion for a rapid diagnosis of these agents is by using nasopharyngeal aspirates or nasopharyngeal swabs for immunofluorescent diagnosis. Thus these specimens remain critically important. Deep respiratory specimens such as endotracheal aspirates or bronchoalveolar lavage specimens appear to have higher viral load and provide higher sensitivity (Table 9.1).16 However, sputum is not often available in the early stage of SARS because most patients do not have a productive cough. Saliva is a less satisfactory specimen.5 Serum or plasma is a less hazardous specimen with regard to hospital infection control than either a nasopharyngeal aspirate or throat swab. It is also an easier specimen to standardize for quantitative studies. However, since neither serum nor plasma presently allow detection of other respiratory viruses that are important differential diagnoses in a patient with suspected SARS (see above), a respiratory specimen has to be obtained in any event.
Antigen detection in the serum Given the low viral load in the upper respiratory tract, it may be expected that antigen detection tests may be insufficiently sensitive early in the disease.26,27 This has indeed been found to be the case in early prototype tests where antigen detection tests had sensitivities of around 71% at days 6–10 of illness but only 50% at days 3–5 of disease.26
Virus culture SARS CoV can be isolated from clinical specimens by culture on VeroE6 or FRhK-4 cells, both continuous primate kidney cell lines.1–3 The use of trypsin in the cell culture medium does not improve virus isolation rates. Primary virus isolation is usually difficult and takes too long to be clinically relevant. However, once adapted to cell culture, SARS CoV grows readily to high titre. While viral RNA can be detectable for 4 weeks or longer, especially in the faeces, virus
68
Severe Acute Respiratory Syndrome
isolation becomes less likely after the third week of illness.5 In this study, virus isolation rates from respiratory specimens at the first, second and later stages of illness were 12%, 7% and 1%, respectively. In comparison, only 5% yielded a virus isolate in the first week from faeces or rectal swabs, and there were no virus isolates later in the illness.28 This is in marked contrast to the detection rates by RT-PCR which is highest in faeces in the second week of illness. It is possible that the appearance of antibodies leads to neutralized virus–antibody complexes that may be still detectable by RT-PCR.
Antibody detection Immunofluorescent tests (IFA) and virus neutralization tests on acute and convalescent phase sera were the first diagnostic tests to be available and remain the gold standard for confirmation of a diagnosis of SARS. Other options such as EIA assays using viral lysate or recombinant protein antigens have been described, but it is best if these results are confirmed by IFA or neutralization tests.
Indirect immunofluorescence assay (IFA) IFA was one of the first methods used for diagnosis of SARS CoV infection and remains a gold standard method for confirmation or exclusion of SARS. While patients with SARS invariably develop antibodies by both IFA and neutralization assays, healthy community controls rarely have antibodies.5,28 However, some groups such as workers in live game animal markets in southern China may become seropositive through exposure to the precursor animal SARS CoV virus without developing SARS.29 SARS CoV grown in Vero-4 or FRhK-4 cells to 60–70% infected cells are spotted out on Teflon-coated slides, fixed in -20oC cold acetone and can be used for IFA diagnosis.5 While IFA slides fixed in acetone at -20oC have been found to be non-infectious, fixa-
tion of cells in acetone at room temperature does not inactivate the virus reliably and poses an infection risk (KH Chan, unpublished data). Other protocols for inactivating SARS CoV slides use gamma-irradiation of cells after fixation or treatment with 5% formalin. It is important that serum samples are tested with infected and uninfected cells in parallel to exclude any non-specific reactivity against uninfected cells. Antibody responses detected by IFA and neutralization tests follow similar kinetics, with seroconversion occurring on average around day 10 post-infection.30 (See also KH Chan, unpublished observations.) IgM antibody does not provide earlier diagnosis. Some patients, especially if treated with corticosteroids, may not seroconvert until the fourth week of illness.8 Therefore, tests on a convalescent serum sample collected at least 28 days after onset of disease are required to exclude the diagnosis of SARS.
Enzyme immunoassay (EIA) To increase the throughput of serological testing, several laboratories have developed EIA assays. Infected cell culture extracts after appropriate inactivation of infectivity3 or recombinant viral proteins31,32 have been used to detect antibodies in serum. Antibody kinetics in EIA tests are similar to those described above for IFA tests. As with other virus infections, IgM antibody responses in SARS decline and disappear within 3 months of disease onset.33 However, indirect IgM tests do pose substantial risk of serological crossreaction, a problem that can be minimized with IgM capture EIA methods. Commercial EIA test kits are now available, but there are few published data on their performance. An early study using SARS CoV EIA based on cell culture extract antigen showed no cross-reaction occurred between human coronavirus (hCo-) OC43 and hCo-229E positive serum samples.3 However, depending on the antigen preparation and opti-
Viral Diagnosis of SARS
mization of the EIA test, serological crossreactions may pose problems in some assays. Thus, especially in periods of low SARS activity in a community, positive EIA tests need to be confirmed by IFA or neutralization tests. The use of control antigen in parallel with the test antigen is critical in order to detect false positive reactions in both EIA and IFA tests.
Virus neutralization test (NT) NT is generally the most specific in terms of cross-reactivity between related viruses and remains a gold standard method for confirming a serological diagnosis of SARS. Either FRhK-4 cells or Vero-E6 cells can be used for NT assays and the result is usually readable as neutralization of viral cytopathic effect by 96 hours. The neutralization titres appear with similar kinetics to that of IFA IgG antibody. About 75% of patients continue to have virus neutralization titres ≥1:160 at 6 months after the onset of disease (K. H. Chan, unpublished results).
Confirmation of SARS CoV infection In periods of low SARS activity, the majority of patients with initial positive screening tests for SARS are likely to be false positives and this is true even with the use of tests of high specificity. It is therefore essential that initial positive screening tests are confirmed in an appropriate manner and guidelines are available from WHO (http://www. who.int/csr/resources/publications/en/SA RSReferenceLab.pdf). Sources of false positive results could arise from contamination of the original specimen or RNA extract with PCR amplicons (PCR cross-contamination) or with the positive control used in the RT-PCR assay. Care to avoid cross-contamination starts with appropriate handling of the specimen. Possible sources of contamination leading to false positive results in laboratory tests include laboratory culture of the infectious virus, cloned DNA or PCR products. Clinical
69
specimens should be handled in areas where none of these risks exist. To exclude risk of PCR cross-contamination, RNA from the clinical specimen should be reextracted and the test repeated, ideally using an RT-PCR for a different viral gene target (e.g. polymerase gene RT-PCR confirmed with a RT-PCR targeting the nucleocapsid gene or vice versa). Obtaining a separate clinical specimen for retesting is very desirable. In addition, independent serological confirmation by demonstration of seroconversion or rising antibody titres is reassuring. However, this may not be possible until 10 days or so after disease onset (see above). To avoid possible contamination, each suspected clinical specimen should be divided into three separate aliquots, so that unopened aliquots can be sent to the national or international reference laboratory if necessary. Since the first diagnosis of SARS in a low endemic period has such major global public health implications, the World Health Organization (WHO) suggests that a diagnosis of SARS be confirmed in an independent reference laboratory. The WHO has set up a three-tier laboratory network for SARS diagnosis (i.e. a system composed of local laboratories, national reference laboratories and international reference laboratories). Positive results from national reference laboratories would be regarded as probable cases and samples will be sent to international reference laboratories for verification. Those cases that are confirmed positive in international reference laboratories would be classified as confirmed SARS cases.
Biosafety considerations for SARS diagnosis The recent laboratory-acquired SARS CoV infections highlight the importance of biosafety in handling infectious agents in laboratories. Biosafety level (BSL) 3 work practices and BSL 3 containment should be
70
Severe Acute Respiratory Syndrome
adopted for work involved in handling of viral cultures (e.g. virus isolation, preparation of viral antigens for IF or EIA) or handling of infectious SARS CoV at high concentration (e.g. concentrating virus by ultracentrifugation) (see guidelines for handling SARS specimens from WHO at http://www.who.int/csr/sars/biosafety 2003_12_18/en/). Routine serology from patients with suspected SARS can be done in a BSL-2 laboratory with respiratory protective gear (FFP 3 or N 95 grade breathing masks). Since SARS CoV has been detected in plasma, heat inactivation at 56oC for 30 minutes is recommended. Such pretreatment does not affect antibody detection, but reduces SARS CoV infectivity in serum by 104-fold (WHO multi-centre collaborative network on SARS diagnosis, 2003). Diagnostic tests which do not require propagation of SARS CoV but require manipulation of potentially infectious clinical samples (e.g. extracting viral RNA from clinical samples for RT-PCR, for performing IF tests on serum and aliquoting clinical specimens) can be performed in a BSL2 containment with BSL3 work practices. Diagnostic tests which do not require infectious SARS CoV (e.g. handling of inactivated SARS CoV) can be performed in a BSL2 containment with BSL2 practices. For details of BSL containment and BSL practices, refer to relevant laboratory biosafety manuals e.g. that issued by WHO (http://www.who.int/ csr/resources/publications/biosafety/ Labbiosafety.pdf). It is important to note that biosafety is achieved by a combination of well-designed facilities, well-defined and appropriate standard operation procedures, training and medical surveillance. Each link in the chain is equally critical.
Summary The sensitivity of molecular tests for diagnosis of SARS has improved markedly but no one test is sufficiently sensitive to exclude conclusively a diagnosis of SARS in the first few days of illness. Serology by
immunofluorescence or neutralization remains the gold standard test for confirmation of a diagnosis of SARS.
Acknowledgement This work is supported by Public Health Research Grant A195357 from the National Institute of Allergy and Infectious Diseases, USA, the Research Grant Council of Hong Kong (HKU 7543/03M), the German Ministry of Health (grant No. 325–4539–85/3).
References 1 Peiris JSM, Lai ST, Poon LLM et al. Coronavirus as a possible cause of severe acute respiratory syndrome. Lancet 2003;361: 1319–25. 2 Drosten C, Günther S, Preiser W et al. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. N Engl J Med 2003;348: 1967–76. 3 Ksiazek TG, Erdman D, Goldsmith CS et al. A novel coronavirus associated with severe acute respiratory syndrome. N Engl J Med 2003;348: 1953–66. 4 Poon LLM, Wong OK, Chan KH et al. Rapid diagnosis of a coronavirus associated with severe acute respiratory syndrome (SARS). Clin Chem 2003;49: 953–5. 5 Chan KH, Poon LLM, Cheng VCC et al. Detection of SARS coronavirus in patients with suspected severe acute respiratory syndrome. Emerg Infect Dis 2004;10: 294–9. 6 Cheng PKC, Wong DA, Tong LKL et al. Viral shedding patterns of coronavirus in patients with probable severe acute respiratory syndrome. Lancet 2004;363: 1699–700. 7 Drosten C, Gottig S, Schilling S et al. Rapid detection and quantification of RNA of Ebola and Marburg viruses, Lassa virus, CrimeanCongo hemorrhagic fever virus, Rift Valley fever virus, dengue virus and yellow fever virus by real-time reverse transcription-PCR. J Clin Microbiol. 2002;40: 2323–30. 8 Peiris JSM, Chu CM, Cheng VCC et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;61:1767–72. 9 Poon LLM, Chan KH, Wong OK et al. Early diagnosis of SARS Coronavirus infection. J Clin Virol 2003;28: 233–8. 10 Drosten C, Weber M, Seifried E et al. Evaluation of a new PCR assay with competitive internal control sequence for blood donor screening. Transfusion 2000;40: 718–24.
Viral Diagnosis of SARS
11 Poon LL, Wong BW, Chan KH et al. A one step quantitative RT-PCR for detection of SARS coronavirus with an internal control for PCR inhibitors. J Clin Virol 2004;30: 214–17. 12 Rota PA, Oberste MS, Monroe SS et al. Characterization of a novel coronavirus associated with severe acute respiratory syndrome. Science 2003;300: 1394–9. 13 Kuiken T, Fouchier RA, Schutten M et al. Newly discovered coronavirus as the primary cause of severe acute respiratory syndrome. Lancet 2003;362: 263–70. 14 Poon LLM, Chan KH, Wong OK et al. Detection of SARS Coronavirus in SARS patients by conventional and real-time quantitative RTPCR assays. Clin Chem 2004;50: 67–72. 15 Ng EK, Hui DS, Chan KC et al. Quantitative analysis and prognostic implications of SARS coronavirus RNA in the plasma and serum of patients with severe acute respiratory syndrome. Clin Chem 2003;49: 1976–80. 16 Drosten C, Chiu LL, Panning M et al. Evaluation of advanced reverse transcription-PCR assays and an alternative PCR target region for detection of severe acute respiratory syndrome-associated coronavirus. J Clin Microbiol 2004;42: 2043–7. 17 Yam WC, Chan KH, Poon LL et al. Evaluation of reverse transcription-PCR assays for rapid diagnosis of severe acute respiratory syndrome associated with a novel coronavirus. J Clin Microbiol 2003;41: 4521–4. 18 Grant PR, Garson JA, Tedder RS et al. Detection of SARS coronavirus in plasma by real-time RT-PCR. N Engl J Med 2003;349: 468–9. 19 Griffith BP, Rigsby MO, Garner RB et al. Comparison of the Amplicor HIV-1 monitor test and the nucleic acid sequence-based amplification assay for quantitation of human immunodeficiency virus RNA in plasma, serum, and plasma subjected to freeze-thaw cycles. J Clin Microbiol 1997;35: 3288–91. 20 Manzin A, Bagnarelli P, Menzo S et al. Quantitation of hepatitis C virus genome molecules in plasma samples. J Clin Microbiol 1994;32: 1939–44. 21 Tsang OT, Chau T, Choi K et al. Coronaviruspositive nasopharyngeal aspirate as predictor for severe acute respiratory syndrome mortality. Emerg Infect Dis 2003;9: 1381–7. 22 Ng LF, Wong M, Koh S et al. Detection of severe acute respiratory syndrome coronavirus in blood of infected patients. J Clin Microbiol 2004;42: 347–50. 23 Mahony JB, Petrich A, Louie L et al. Performance and cost evaluation of one commercial
24
25
26
27
28
29
30
31
32
33
71
and six in-house conventional and real time reverse transcriptase PCR assays for detection of severe acute respiratory syndrome. J Clin Microbiol 2004;42: 1471–6. Poon LL, Leung CS, Tashiro M et al. Rapid detection of the severe acute respiratory syndrome (SARS) coronavirus by a loop-mediated isothermal amplification assay. Clin Chem 2004;50: 1050–2. Hong TC, Mai QL, Cuong DV et al. Development and evaluation of a novel loopmediated isothermal amplification method for rapid detection of severe acute respiratory syndrome coronavirus. J Clin Microbiol 2004;42: 1956–61. Che XY, Qiu LW, Pan YX et al. Sensitive and specific monoclonal antibody based capture enzyme immunoassay for detection of nucleocapsid antigen in sera from patients with severe acute respiratory syndrome. J Clin Microbiol 2004;42: 2629–35. Lau KP, Woo PCY, Wong BHL et al. Detection of severe acute respiratory syndrome (SARS) coronavirus nucelocapsid protein in SARS patients by enzyme-linked immunosorbent assay. J Clin Microbiol 2004;42: 2884–9. Chan PKS, To WK, Ng KC et al. Laboratory diagnosis of SARS. Emerg Infect Dis 2004;10: 825–31. Guan Y, Zheng BJ, He YQ et al. Isolation and characterization of viruses related to the SARS coronavirus from animals in southern China. Science 2003;302: 276–8. Hsueh PR, Hsiao CH, Yeh SH et al. SARS Research Group of National Taiwan University College of Medicine and National Taiwan University Hospital. Microbiologic characteristics, serologic responses, and clinical manifestations in severe acute respiratory syndrome, Taiwan. Emerg Infect Dis 2003;9: 1163–7. Woo PCY, Lau SKP, Tsoi HW et al. Relative rates of non-pneumonic SARS coronavirus infection and SARS coronavirus pneumonia. Lancet 2004;363: 841–5. Woo PCY, Lau SKP, Wong BHL et al. Longitudinal profile of immunoglobulin G (IgG) IgM, amd IgA antibodies against the severe acute respiratory syndrome (SARS coronavirus nucelocapsid protein in patients with pneumonia due to the SARS coronavirus.) Clin Diagn Lab Immunol 2004;11: 665–8. Li G, Chen X, Xu A. Profile of specific antibodies to the SARS-associated coronavirus. N Engl J Med 2003;349: 508–9.
Chapter 10
Pathology and Pathogenesis JM Nicholls and T Kuiken
Soon after the beginning of the SARS outbreak a novel coronavirus was detected in patients with SARS by virus isolation and molecular methods.1–3 Coronavirus particles also were detected by transmission electron microscopy in (unidentified) cells from a lung biopsy1, a broncho-alveolar lavage3 and in pneumocytes from postmortem.4 The identification of this novel coronavirus that produced severe respiratory and gastrointestinal symptoms in humans forced a rapid reappraisal of coronavirus disease in humans; until the outbreak of SARS, coronavirus infections in humans were associated primarily with the common cold.5 More tenuous associations with multiple sclerosis, pancreatitis, thyroiditis, pericarditis, nephropathy and infectious mononucleosis have also been reported. Proof that this novel coronavirus, SARS CoV, was the primary cause of SARS required production of a comparable disease in the original host or a related species. Therefore, experimental infections were performed in cynomolgus macaques (Macaca fascicularis).6,7 Besides helping to prove that SARS CoV is the primary cause of SARS, this SARS CoV-macaque model also was useful in studying the pathogenesis of the disease and in testing intervention strategies against SARS, both preventative and therapeutic. Within the lungs the acinus is the site of gas transfer. The acinus includes a respiratory bronchiole, its distal alveolar ducts
72
and alveoli.8 Two types of epithelial cells line the alveoli; the flattened type 1 pneumocyte and the more cuboidal type 2 (also known as the granular pneumocyte). The type 1 pneumocyte constitutes only 8% of the lung parenchymal cells but has an average cell surface area of 5098 mm2. The type 2 pneumocytes, which secrete surfactant, account for 16% of the lung parenchymal cells but have a mean surface area of 183 mm2. The epithelial cells are connected by tight junctions, whereas the underlying endothelial cells have leaky junctions that allow the extravasation of fluid following injury.9 The type 1 cells are regarded as nonreplicating whilst the type 2 are the progenitor cells of type 1 and replicate after injury to replace the type 1 cells. This type 2 hyperplasia is interpreted as a non-specific marker of alveolar injury and repair. In addition to the epithelial cells, alveolar macrophages are identified attached to the wall of the alveolus. These are non-specific scavenger cells that are derived from monocytes that have emigrated from the blood10 and differ in function from the interstitial macrophages.11 The SARS outbreak began in late 2002 and finished in July 2003. During that time there were over 700 deaths,12 but understanding the pathogenesis and pathology of the mortality attributed to this disease was hampered by the limited numbers of post-mortems or biopsy material from these cases. This was due to concern at the time over possible risk of infection to mortuary
Pathology and Pathogenesis
and pathology staff from a novel viral infection. In addition, cultural beliefs in a number of Asian countries made obtaining consent for autopsy difficult. In Hong Kong, most autopsies were limited to either the lung or small pieces of tissue from the lungs of patients who had spent a considerable period of time in hospital and who had been treated with multiple regimens of antibiotics, antivirals, immunomodulators and ventilator therapy. Thus there is a less than complete view of the natural course of the disease in humans. However, as a result of three published studies on the pathology of fatal SARS cases,4,13,14 it is still possible to develop a pathophysiological model for SARS CoV pneumonia. This model has been aided by the recent production of diagnostic antibodies against SARS CoV15 and in situ hybridization studies for SARS CoV RNA.16 SARS clinically involves the lower respiratory tract to a greater extent than the upper, and within the lower respiratory tract the acinus is targeted more than the trachea or bronchus.15 In two publications of a series of post-mortems4,13 emphasis has been placed on two stages or phases of SARS CoV pneumonia. The early stage, which encompasses about the first 10 days of infection, is that of an acute phase or exudative diffuse alveolar damage (DAD), characterized by a mixed inflammatory cell infiltrate and the presence of oedema and hyaline membrane formation (Plate 2). These hyaline membranes are composed of precipitated plasma proteins and the cytoplasmic and nuclear debris from sloughed epithelial cells.17 It has been demonstrated that the type 1 cells are more susceptible to viral damage than the type 2 cells.9 With the necrosis of the epithelial cells, the barrier between blood circulation and airways is removed and fluid from pulmonary capillary vessels is free to enter the alveolar space. As there have been limited numbers of cases available of SARS CoV to provide biopsy or autopsy material in this early stage of DAD it is not possible to determine if this type 1 epithelial denudation is a di-
73
rect toxic effect of the virus or secondary to a host immune response. Viral RNA and viral antigen have been identified within alveolar macrophages and epithelial cells in this exudative stage of SARS CoV pneumonia and viral particles have been identified in epithelial cells and macrophages by electron microscopy.16 The second phase of SARS CoV pneumonia appears after approximately 10 days with a change from an exudative to organizing DAD. In this period there is focal squamous metaplasia of the bronchial epithelium, increased cellularity and fibrosis in the alveolar walls as well as within the alveolar lumen. There are prominent type 2 pneumocytes present, often showing cytological atypia with enlarged nuclei and prominent eosinophilic nucleoli.4,13 One report18 described intracytoplasmic viral inclusion bodies, but two other reports4,13 were unable to confirm these findings. Cytoplasmic inclusions also were not seen in the FRhK4 and Vero-E6 cells infected with SARS CoV.13,14 One common finding from the organizing stage of DAD has been the presence of multinucleated giant cells in the alveolar lumen (Plate 3). These giant cells were already reported in two of the early publications on SARS1,3 and appear to be epithelial in some cases but macrophage-derived in others.4,13 Initially it was thought that because coronaviruses could lead to syncytium formation in cell culture19 these giant cells were a direct result of viral infection, similar to measles virus and respiratory syncytial virus infections.20 Immunoperoxidase and in situ hybridization studies, however, reveal that the organizing stage of SARS CoV pneumonia is actually associated with minimal viral presence,15 and electron microscopic studies have not identified any virus in these giant cells. On the other hand, multinucleated giant cells are not a conspicuous feature of many types of DAD, so other factors probably play a role. The tentative conclusion drawn from these studies is that the
74
Severe Acute Respiratory Syndrome
organizing stage of SARS CoV DAD is not driven by persistent viral replication but rather by a response to marked epithelial cell damage during the exudative phase, aggravated by ventilator therapy. If the organizing stage of SARS CoV is a non-specific DAD the question arises whether there are any unique features in the pathological changes of the exudative SARS CoV or whether the changes in the lung are a non-specific tissue reaction. Whilst evidence of SARS viral replication can be seen by electron microscopy, immunohistochemistry and in situ hybridization in the early stages15 it is worthwhile to note that an exudative DAD with hyaline membrane formation and cellular infiltration can be seen in other viral pneumonias due to cytomegalovirus, herpes simplex, influenza, measles, adenovirus, parainfluenza and hantavirus infections20 but most of these are associated with typical inclusions. Initially, SARS CoV was thought to be limited to the respiratory tract. However the prominent symptoms of diarrhoea21 indicated that the virus may have a wider tissue tropism. In fact, SARS CoV was detected by RT-PCR and isolated in culture from faeces and urine, suggesting that SARS is a systemic infection. Viral particles have been demonstrated by electron microscopy in the gastrointestinal tract epithelium of humans.22 Experimental infection in macaques was done to confirm that SARS CoV was the primary cause of SARS.6,7 In summary, 1 ¥ 106 TCID50 of a SARS CoV isolate from a patient who died of SARS was applied intratracheally, intranasally and on all conjunctivas of four cynomolgus macaques. Clinical signs following infection were variable, and included lethargy, a temporary skin rash, and respiratory distress. Full necropsies were performed at 6 days post infection (dpi), and a wide range of tissues was collected for histological examination. Grossly, three of four macaques had multiple foci of pulmonary consolidation and enlargement of tracheo-bronchial lymph nodes and spleen.
In the lungs, the macaques showed a similar pattern of injury to human SARS. Microscopically, the areas of pulmonary consolidation consisted of acute or more advanced stages of diffuse alveolar damage (DAD). In areas with acute DAD, the lumina of alveoli and bronchioles were variably filled with protein-rich oedema fluid, fibrin, erythrocytes and cellular debris, admixed with a moderate number of alveolar macrophages and fewer neutrophils and lymphocytes (Plate 4). There was extensive loss of epithelium from alveolar and bronchiolar walls. Some alveolar walls were lined by deep eosinophilic hyaline membranes (Plate 4). Dysplastic type 2 pneumocytes with large vacuolated nuclei, prominent nucleoli and abundant vesicular cytoplasm were frequently found attached to the alveolar walls. The epithelial origin of these cells was confirmed by keratin expression. There were occasional multinucleated giant cells (syncytia) in bronchioles and alveoli, either attached to the wall or free in the lumen (Plate 4). Based on positive CD68 staining these syncytia originated from macrophages; unlike human studies, cytokeratin-positive syncytia, indicating an epithelial origin, were not found. In areas with more advanced diffuse alveolar damage, the alveolar walls were moderately thickened and lined by cuboidal epithelial cells (type 2 pneumocyte hyperplasia), and the alveolar lumina contained mainly alveolar macrophages (Plate 4). There was epithelial regeneration in some bronchioles. Alveolar and bronchiolar walls were thickened by oedema fluid, mononuclear cells, and neutrophils, and there were aggregates of lymphocytes around small pulmonary vessels. Moderate numbers of lymphocytes and macrophages were present in the lamina propria and submucosa of the bronchial walls, and scattered neutrophils in the bronchial epithelium. Histological changes in other tissues of these three macaques were limited to lymphoid hyperplasia and sinus histiocytosis of the tracheo-bronchial lymph nodes and
Pathology and Pathogenesis
intrafollicular hyalinosis of the spleen. The fourth macaque had minimal multifocal inflammatory lesions in the pulmonary tissue. By immunohistochemistry, expression of SARS CoV antigen was observed in cuboidal alveolar epithelial cells (type 2 pneumocytes) and in intrabronchiolar and intra-alveolar syncytia. The presence of coronavirus in alveolar epithelial cells was confirmed by transmission electron mi-
75
croscopy, which revealed coronavirus-like particles measuring about 70 nm in diameter with typical internal helical nucleocapsid-like structure and club-shaped surface projections in the cytoplasm of affected cells (Fig. 10.1). These particles were similar in size and structure to coronavirus particles in Vero 118 cells infected with SARS CoV (Fig. 10.1). The macaques shed SARS CoV from sputum, nose, and pharynx from 2 days after
Figure 10.1 Electron microscopy of SARS CoV in inoculum, specimens, and tissue samples of experimentally infected cynomolgus macaques.(A) Negative contrast electron microscopy of the virus stock used to inoculate cynomolgus macaques shows the typical club-shaped surface projections of coronavirus particles. Negatively stained with phosphotungstic acid. Bar = 100 nm. (B) Morphologically identical particles were isolated from nasal swabs of infected macaques. Negatively stained with phosphotungstic acid. Bar = 100 nm. (C) Transmission electron microscopy of an infected Vero 118 cell shows viral nucleocapsids with variably electron-dense and electron-lucent cores in smooth-walled vesicles in the cytoplasm. Stained with uranyl acetate and lead citrate. Bar = 500 nm. (D) Morphologically similar particles occur in pulmonary lesions of infected macaques, within vesicles of the Golgi apparatus of pneumocytes. Stained with uranyl acetate and lead citrate. Bar = 500 nm. (Reprinted with permission from Elsevier (Kuiken et al., 2003, Lancet, 362: 263–270).)
76
Severe Acute Respiratory Syndrome
infection, based on virus isolation, RT-PCR, or both. By negative-contrast electron microscopy, coronavirus particles were seen in cell cultures obtained from nasal swabs and closely resembled those in the virus stock used to infect the macaques (Fig. 10.1). In post-mortem samples, SARS CoV was detected in the lungs of all four macaques, in trachea and tracheo-bronchial lymph node of two of four macaques, and in the cerebrum, duodenum, kidney, nasal septum, skin, spleen, stomach and urinary bladder of one of four macaques by virus isolation, RT-PCR, or both. The above results indicate that the main pulmonary lesions in macaques are comparable to those in SARS patients and to those in respiratory coronavirus infections in other species, e.g. sialodacryoadenitis virus infection in rats23 and porcine respiratory coronavirus infection in pigs.24 A contributory role of co-infections with other infectious agents, such as human metapneumovirus, cannot be excluded. The following observations suggest that the pneumocyte is the primary target cell in SARS CoV infection, although the mechanism by which damage to this cell occurs remains to be determined: 1 In several other species,23,25,26 coronavirus infections of the respiratory tract are trophic for respiratory epithelium. 2 The rare detection of SARS CoV in SARS patients involves epithelial cells.4 3 By immunohistochemistry, type 1 pneumocytes of experimentally infected macaques show widespread expression of SARS CoV antigen at 4 dpi,27 but not at 6 dpi, when expression of SARS CoV antigen is limited to occasional type 2 pneumocytes.7 This corresponds to the dynamics of infection of coronaviruses in pigs and rats, where viral infection of respiratory epithelium is maximal at 3–4 dpi, and is no longer measurable by 6–9 dpi.23,26 Lesions of SARS CoV infection in macaques7 and humans4,10 indicate type 1 pneumocyte necrosis, observed either directly by desquamation or loss of type 1
pneumocytes, or indirectly by reactive type 2 pneumocyte hyperplasia. Experimental study on the primates7 found tracheobronchial lymph node and spleen enlargement. Histologically, lymph nodes showed lymphoid hyperplasia, while the spleen showed intrafollicular hyalinosis. Lymphopenia is a consistent finding in the majority of SARS cases and may be similar in aetiology to influenza-induced leukopenia28 but this could not be performed on the primates. The primate study also found increased numbers of alveolar macrophages. The immunoperoxidase and in situ hybridization studies have shown the presence of viral RNA and protein in alveolar macrophages. It therefore appears that the macrophages are acting as the scavenger cells and not as sites for continued replication. Whether they play a role in antigen presentation has not yet been determined. In our studies on SARS we considered whether the pathological changes seen in the SARS pneumonias represent a nonspecific response to lung damage which is aggravated by ventilator therapy or one in which the immune system plays a role. As alveolar and interstitial macrophages differ in their roles and are but one facet of the immune system, bronchoalveolar lavage (BAL) from SARS patients will only provide limited information on the role of the immune response in SARS. Furthermore, in the patients with SARS, BAL also carries the increased risk for viral transmission to heath-care personnel. Even if macrophages are shown to be activated, previous studies suggested that ventilator-induced lung injury (which was instituted in many SARS patients) produces activation of macrophages in vitro.29–32 Comparative pathological studies of respiratory coronavirus infections in different species may allow us to better understand the pathogenesis of SARS in humans, especially the role of the host immune response in the pathogenesis of SARS. In particular, experimental infection of non-human pri-
Pathology and Pathogenesis
mates enables us to study the early stages of SARS CoV infection, which is rarely possible with material from SARS patients. In addition to the macaque model, successful replication of SARS CoV infection in ferrets and cats33 provides alternative animal models for the evaluation of anti-SARS therapies.
References 1 Peiris JS, Lai ST, Poon LL et al. Coronavirus as a possible cause of severe acute respiratory syndrome. Lancet 2003;361: 1319–25. 2 Drosten C, Günther S, Preiser W et al. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. N Engl J Med 2003;348: 1967–76. 3 Ksiazek TG, Erdman D, Goldsmith C et al. A novel coronavirus associated with severe acute respiratory syndrome. N Engl J Med 2003;348: 1953–66. 4 Nicholls JM, Poon LLM, Lee KC et al. Lung pathology of fatal severe acute respiratory syndrome. Lancet 2003;361: 1773–8. 5 Myint SH. Human coronavirus infections. In: Siddell S, ed. The Coronaviridae. New York: Plenum Press, 1995: 389–401. 6 Fouchier RA, Kuiken T, Schutten M et al. Aetiology: Koch’s postulates fulfilled for SARS virus. Nature 2003;423: 240. 7 Kuiken T, Fouchier RA, Schutten M et al. Newly discovered coronavirus as the primary cause of severe acute respiratory syndrome. Lancet 2003;362: 263–70. 8 Colby TY, Yousem SA. Lungs. In: Sternberg S, ed. Histology for Pathologists. New York: Raven Press, 1992: 479–97. 9 Tomashefski JF. Pulmonary pathology of acute respiratory distress syndrome. Clin Chest Med 2000;21: 435–66. 10 Holian A, Scheule RE. Alveolar macrophage biology. Hosp Pract (Off Ed) 1990;25: 53–62. 11 Franke-Ullmann G, Pfortner C, Walter P et al. Characterization of murine lung interstitial macrophages in comparison with alveolar macrophages in vitro. J Immunol 1996;157: 3097–104. 12 http://www.who.int/csr/sars/countr ytable2003_09_23/en/ (accessed 7 January 2004). 13 Franks TJ, Chong PY, Chui P et al. Lung pathology of severe acute respiratory syndrome (SARS): A study of 8 autopsy cases from Singapore. Hum Pathol 2003;34: 743–8. 14 Ding Y, Wang H, Shen H et al. The clinical pathology of severe acute respiratory syndrome (SARS): a report from China. J Pathol 2003;200: 282–9.
77
15 Wong MP, Beh PSL, Leung CY et al. SARS coronavirus (SARS-CoV) targets pulmonary alveolar epithelium and macrophages and leads to persistent proinflammatory and fibrogenic cytokine expression (submitted). 16 Nakajima N, Asahi-Ozaki Y, Nagata N et al. SARS coronavirus-infected cells in lung detected by new in situ hybridization technique. Jpn J Infect Dis 2003;56: 139–41. 17 Travis WD, Colby TV, Koss MN et al. Nonneoplastic disorders of the lower respiratory tract. First Series Fascicle 2. American Registry of Pathology, Washington DC, 2002: 49–231. 18 Ding YQ, Wang HJ, Shen H et al. Study on etiology and pathology of severe acute respiratory syndrome. Zhonghua Bing Li Xue Za Zhi 2003;32: 195–200. 19 Cavanagh D. The coronavirus suface glycoprotein. In: Siddell S, ed. The Coronaviridae. New York: Plenum Press, 1995: 73–113. 20 Travis WD, Colby TV, Koss MN et al. Nonneoplastic disorders of the lower respiratory tract. First Series Fascicle 2. American Registry of Pathology, Washington DC, 2002: 539– 727. 21 Peiris JS, Chu CM, Cheng VC et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;361: 1767–72. 22 Leung WK, To KF, Chan PK et al. Enteric involvement of severe acute respiratory syndrome-associated coronavirus infection. Gastroenterology 2003;125: 1011–17. 23 Bhatt PN, Jacoby RO. Experimental infection of adult axenic rats with Parker’s rat coronavirus. Arch Virol 1977;54: 345–52. 24 O’Toole D, Brown I, Bridges A et al. Pathogenicity of experimental infection with ‘pneumotropic’ porcine coronavirus. Res Vet Sci 1989;47: 23–9. 25 Purcell DA, McFerran JB. The histopathology of infectious bronchitis in the domestic fowl. Res Vet Sci 1972;13: 116–22. 26 Jabrane A, Girard C, Elazhary Y. Pathogenicity of porcine respiratory coronavirus isolated in Quebec. Can Vet J 1994;35: 86–92. 27 Haagmans BL, Kuiken T, Martina BE et al. Pegylated interferon-a protects type 1 pneumocytes against SARS coronavirus infection in macaques. Nat Med 2004;10: 290–3. 28 Nichols JE, Niles JA, Roberts NJ Jr. Human lymphocyte apoptosis after exposure to influenza A virus. J Virol 2001;75: 5921–9. 29 Pugin J, Dunn I, Jolliet P et al. Activation of human macrophages by mechanical ventilation in vitro. Am J Physiol 1998;275: L1040–50. 30 Vlahakis NE, Schroeder MA, Limper AH et al. Stretch induces cytokine release by alveolar
78
Severe Acute Respiratory Syndrome
epithelial cells in vitro. Am J Physiol 1999;277: L167–73. 31 Ricard JD, Dreyfuss D, Saumon G. Production of inflammatory cytokines in ventilatorinduced lung injury: a reappraisal. Am J Respir Crit Care Med 2001;163: 1176–80. 32 Dreyfuss D, Ricard JD, Saumon G. On the
physiologic and clinical relevance of lungborne cytokines during ventilator-induced lung injury. Am J Respir Crit Care Med 2003;167: 1467–71. 33 Martina BEE, Haagmans BL, Kuiken T et al. SARS virus infection. Nature 2003;425: 915.
Chapter 11
SARS Coronavirus: An Animal Reservoir? Yi Guan, Hume Field, Gavin JD Smith and Honglin Chen
Introduction The emergence of SARS and its subsequent rapid global spread caused alarm in public health circles worldwide. The unknown aetiology of this novel, atypical pneumonia was one reason for the alarm. Several putative aetiological agents were considered initially, with a previously undescribed novel coronavirus (Coronaviridae) finally shown to be responsible (Chapter 7). The origin of the outbreak and the source of the infection were both major focuses of attention, the latter even more so once the outbreak in human populations was under control. Was this a previously avirulent human coronavirus or another example of an emerging zoonotic infection? What were the risk factors for exposure? This chapter addresses these questions.
The concept of reservoir Simplistically, there are two plausible explanations of the origin of SARS coronavirus. It is either a previously unidentified human coronavirus whose virulence and/or infectivity increased as a result of genetic change or it is an animal coronavirus that jumped to an immunologically naïve human host. Consideration of the concept of reservoir is useful at this point. Haydon et al.1 define a reservoir as ‘one or more epidemiologically connected populations or environments in which the pathogen can be permanently maintained, and from which infection is
transmitted to the defined target population’. Thus, a discrete human population, an animal population, or an ecological community can be regarded as a reservoir. It is also important to recognize that, for any pathogen, multiple reservoirs are possible. An understanding of these concepts is necessary to grasp the potential complexity of the task of identifying the source of human infection, and thus for the identification of risk factors for exposure and the formulation of risk management strategies.
The search for the reservoir Epidemiological investigations indicate that the SARS outbreak originated in Guangdong Province, China, with the earliest identified case in November 2002.2,3 The search for the reservoir has therefore focused on Guangdong and has comprised both human and animal epidemiological studies. Human studies have relied heavily on case report and case interview data collected by the Guangdong Centre for Disease Control. Preliminary analysis of the early Guangdong data suggested an epidemiologic association between occupation and infection, with restaurant chefs being overrepresented (Report of the First WHO Mission to China, April 2003). A more detailed analysis indicated that people working in the food industry and people living close to markets were over-represented in early cases (Report of the Second WHO Mission to China, May 2003). None of the early cases
79
80
Severe Acute Respiratory Syndrome
lived close to livestock or farms, discounting an association with domestic animals. These findings were interpreted as best supporting a wildlife reservoir hypothesis, with the putative mode of transmission being exposure to infectious body fluids of the live or recently slaughtered animal. Genetic comparison of SARS CoV with known coronaviruses of humans and domestic animals revealed that each gene of SARS CoV had only 70% or less identity with the corresponding gene of the known coronaviruses. Phylogenetic analysis also showed that SARS CoV does not fall into any of the three known groups of coronaviruses (Chapter 8). These findings suggested that SARS CoV did not arise by mutation of human coronaviruses or by recombination between any known coronaviruses. Subsequent phylogenetic analysis by the authors revealed that human CoV isolates 229E (group 1) and OC43 (group 2) clustered with coronaviruses isolated from domestic mammals, such as pigs and mice. From an evolutionary point of view, viruses 229E and OC43 resulted from two previous independent interspecies transmission events from domestic animals to humans. This indicates that all human coronaviruses are zoonotic from domestic animals. However, SARS CoV is not closely related to other known coronaviruses but is derived from the root of group 2 coronaviruses. Thus, the plausible explanation is that SARS CoV was derived from an unknown coronavirus that existed in an unknown host, very likely a wild animal, before it jumped to humans. In contrast to recent human infections with influenza H5N1 ‘bird flu’ where transmission is so far limited largely to bird to human transmission, SARS CoV adapted to efficient human to human transmission. It was reasonable to hypothesize that the possible host of SARS CoV may be a mammal since a similar replication system and cellular factors made it much easier for SARS CoV
to adapt to humans. For an RNA virus to be present as a viable population, there must be a host species with sufficient population size to support the virus prevalence. Therefore, when investigating the possible source of SARS CoV, the author mainly focused on those wild mammalian species as exotic food in the markets but with relatively large market populations, or domestic species that were known to harbour other coronaviruses. The preliminary investigation was carried out at an animal market in Shenzhen on 7 May 2003. SARS CoV infection was found in throat and faecal swabs from three of eight species of wildlife sampled in a Shenzhen animal market (Paguma larvata, Nyctereutes procyonoides, Melogale moschata).4 The animal isolates are phylogenetically distinct from the human isolates, making it improbable that the animals were infected from humans. Interestingly, a 29nucleotide sequence present in all animal isolates is absent from the human isolates, except for an early Guangdong isolate. Another study also documented SARS CoV-like viruses from faecal swabs from farmed P. larvata (Himalayan palm civet) in Hubei Province (Hu ZH, personal communication). Independent studies carried out in several Chinese research institutions returned positive PCR results from P. larvata faecal swabs. The findings of these animal studies indicate that several wildlife species found in markets excrete a SARS-like coronavirus. Thus it is improbable that these are dead-end hosts. Two separate studies have undertaken serologic studies of humans working in animal markets.2,4 Both studies show significantly higher SARS CoV antibody prevalence in wildlife traders and animal slaughterers than in market and community controls. None of the seropositive individuals has a reported history of SARS-like symptoms, suggesting exposure to either an avirulent SARS or SARS-like virus.
Plate 1 Phylogenetic analysis of the SARS-CoV during the course of the epidemic (adapted with permission from Chinese SARS Molecular Epidemiology Consortium. Science 2004;303: 1666–1669.). An unrooted phylogenetic tree of the 61 human SARS-CoV genomes and 2 SARS-like-CoV sequences from palm civets was built with only sequences variants (including deletions) that occurred at least twice. The map distance represents the extent of genotypic difference. The 5-nt motifs indicated in boxes signify the phylogenetically related genotypes. The genomic sequences are named in concordance with their GenBank nomenclature and are represented in different colors according to the genotype clusters. Genotypes with 29-nt, 82-nt or 415-nt deletions and with the 29-nt segment are marked specifically and all other genotypes (unmarked) had the 29-nt deletion.
Facing page 80
Plate 2 Histologic appearance of early human SARS pneumonia. There is an acute phase diffuse alveolar damage present. Hyaline membranes line the walls of the alveoli and there is a mixed inflammatory cell infiltrate present in the interstitium.
Plate 3 Advaced human SARS pneumonia. The alveoli are filled with increased numbers of reactive pneumocytes and macrophages and there is fusion of cells form a syncitia.
Plate 4 Histologic lesions in lungs from cynomolgus macaques experimentally infected with SARS-CoV. A. Early changes of diffuse alveolar damage, characterized by disruption of alveolar walls and flooding of alveolar lumina with serosanguineous exudate admixed with neutrophils and alveolar macrophages. B: More advanced changes of diffuse alveolar damage, characterized by thickened alveolar walls lined by type 2 pneumocytes, and predominantly alveolar macrophages in alveolar lumina. C: The surfaces of many alveoli are covered by hyaline membranes (arrowheads). D: A characteristic change is the presence of syncytia (arrowhead), here in the lumen of a bronchiole. All slides hematoxylin and eosin stained. (Reprinted with permission from Elsevier (Kuiken et al. Lancet 2003;362: 263–270.)
SARS Coronavirus: An Animal Reservoir? 81
A second SARS outbreak averted Even though coronaviruses genetically highly related to SARS CoV were isolated from wild animals, particularly from Paguma larvata, not all investigating laboratories were able to detect the virus in these market animal species. The resulting controversy led to the lifting of the ban on wildlife trade in food markets of Guangdong in September 2003, soon after SARS was controlled in humans in July 2003. As a result, different species of wild animals, including civet, were once again traded in the markets of Guangdong. To attempt to resolve the question of the role of live animal markets in the origin of human SARS CoV, surveillance studies in Guangdong food markets were re-initiated by our group soon after the wildlife ban was lifted in September 2003. This surveillance focused on five species of wild animals including Paguma larvata, Arctonyx collaris (hog badger), Melogale moschata (Chinese ferret badger), Nyctereutes procyonoides (raccoon dog) and Meles meles (Eurasian badger). As in previous market surveillance, a SARS-like virus was also detected in each of these species. However, P. larvata still provided the main body of SARS CoV isolates, with the highest positive rate (about 76%, authors’ unpublished data). At the end of December 2003, while these surveillance efforts were still ongoing, the first suspected SARS patient re-emerged in Guangzhou, the capital city of Guangdong Province. Within a short interval, the second and third suspected SARS CoV cases appeared; SARS was resurgent in the region. Genetic analysis revealed that all of these most recent SARS cases were caused by interspecies transmission that had occurred recently, and the counterpart of human SARS CoV was detected from P. larvata in the markets (authors’ unpublished data). These findings convinced the authorities in Guangdong that they had to take action for a second time and remove all wild animals from the
food markets. After this action was taken, no new SARS case was found in humans, providing convincing evidence that the wet markets in Guangdong were the infectious source of human SARS CoV. In conclusion, the culling of civets and other wild animals in Guandong markets during early 2004 possibly averted another SARS outbreak.
A putative reservoir The repeated investigations and animal studies support the hypothesis that a wildlife market (or markets) was the origin of the SARS outbreak in humans. The findings of serological studies further support this hypothesis by providing a plausible explanation for the under-representation of wildlife traders in early cases — that is, at-risk traders had previous exposure to an avirulent SARS or SARS-like virus which provided protection against exposure to the outbreak strain. Two plausible scenarios could explain the mechanism for the emergence of SARS CoV. These are: i) increasing demand resulted in animals from previously unexploited populations (unidentified reservoir species in which SARS CoV is likely to be asymptomatic) entering markets and/or ii) a genetic change in the circulating wild animal strain, possibly as a result of passage in animals or humans in the market, resulted in increased capacity for introduction into humans. The first scenario is certainly true. In the last decade, China has undergone rapid economic growth, especially in Guangdong and other coastal areas. The rising prosperity of the people in these regions led to increased demand for exotic foods that was met by the farming of wild species, and before the SARS outbreak, China had 700 raccoon dog and 1000 civet cat farms. This generated an ecosystem change allowing those coronaviruses resident in wildlife increased contact with the human population. The second scenario is difficult to
82
Severe Acute Respiratory Syndrome
prove because the SARS CoV precursor virus and its animal reservoir have not been identified yet, but is plausible given the first scenario and the high mutation rates of RNA viruses. Both scenarios are consistent with the clustering of the early cases in the Pearl Delta region.3
the trade in wild-caught wildlife might control transmission to market and farm populations, and thus to humans; elimination of infection in the farmed wildlife population and ongoing monitoring might control transmission within this group, and thus to wildlife markets, and to humans.
A putative causal model
Research priorities
While a SARS or SARS-like virus maintained in wildlife markets is the probable source of the human outbreak, it is implausible that this cycle exists in isolation. Indeed, the emerging picture is of a virus able to infect a wide range of hosts, suggesting a complex ecology. A causal model with interacting natural, market, human, and peri-human animal components has been proposed (Fig. 11.1). Such a model is a useful tool not only for conceptualizing the likely complexity of the system, but also for identifying possible transmission control points. For example, regulation (or elimination) of
There are two key research priorities in investigating the reservoir of SARS. The first is a simple, inexpensive serological test that has been validated for the target species. The second is a comprehensive understanding of the ecology of the reservoir. Serology is well-established as a surveillance tool. Serological testing allows us to screen large numbers of animals in a relatively short time, to guide the direction of further surveillance efforts. However, serological data can be problematic since there is a lack of a suitable detection conjugate for use in the relevant wild animal species.
Domestic animals
X
X
Pets
?
Wild animal
Markets
? ? ?
ZOONOSIS
Exotic animal restaurant Figure 11.1 Ecology of SARS.
Himalayan palm civet, raccoon dog, others?
Wildlife farms
SARS Coronavirus: An Animal Reservoir? 83
There is also cross-reactivity between known coronaviruses in the Western blot test. Systematic virological surveillance is the only way to gain insight into the ecology of the SARS CoV. The findings will provide information to identify the reservoir of SARS and understand the risk factors of reintroduction to the human population. This requires a systematic approach that includes prevalence studies, longitudinal studies, and modelling. Avian influenza surveillance programmes, similar to what is needed for SARS, have been conducted successfully in Hong Kong for pandemic preparedness. In fact, the detection and prevention of the second SARS outbreak through wild animal culling was possible because of the lessons learnt from the H5N1 avian influenza in Hong Kong in 1997. In a situation where the wildlife reservoir is a trade commodity, an extension of understanding the ecology of the reservoir is an understanding of the trade. We know that the wildlife trade and farms still exist in southern China, leaving the possibility of the re-emergence of SARS. Understanding the wild animal trade is critical to its effective management, and directly related to this is an understanding of what drives the wildlife trade — a complex mix of economic, social and cultural factors. Wildlife is expensive (US$30 per kg, compared to US$1 for chicken), and there is evidence that demand and consumption have increased in recent years as economic conditions in China have improved. Why do people eat wildlife? Usually it is for perceived health benefits. For example, Paguma larvata is typically eaten in winter when fresh fruit is often unavailable. It is
believed that eating the animal (also known colloquially as the fruit fox or flower fox because of its dietary preferences) provides the same health benefits as eating fruit. In markets, wild-caught P. larvata meat attracts a price premium because people believe it is more health-giving and tastes better than its grain-fed farmed counterpart.
Conclusions Our understanding of this novel disease is not complete; however, accumulated evidence suggests that SARS resulted from a zoonosis, very probably from wildlife. Paguma larvata, and other wild animals found in markets, were the infection source for the two recognized SARS outbreaks. Since the reservoir and ecology of SARS are not well-defined the re-emergence of this disease cannot be excluded. Research efforts need to concentrate on identifying the SARS CoV reservoir and understanding its ecology better.
References 1 Haydon DT, Cleaveland S, Taylor LH et al. Identifying reservoirs of infection: a conceptual and practical challenge. Emerg Infect Dis 2002; 8: 1468–73. 2 Xu RH, He JF, Evans MR et al. Epidemiologic clues to SARS origin in China. Emerg Infect Dis 2004;10: 1030–7. 3 Zhong NS, Zheng BJ, Li YM et al. Epidemiology and cause of severe acute respiratory syndrome (SARS) in Guangdong, People’s Republic of China, in February 2003. Lancet 2003;362: 1355–8. 4 Guan Y, Zheng BJ, He YQ et al. Isolation and characterization of viruses related to SARS coronavirus from animals in southern China. Science 2003;302: 276–8.
Chapter 12
Comparative Biology of Animal Coronaviruses: Lessons for SARS Linda J Saif
Introduction The genus Coronavirus is comprised of at least three antigenic and genetic groups of coronaviruses (CoV) that cause mild to severe enteric, respiratory or systemic disease in domestic animals, wild animals, poultry and rodents, and minor colds in humans (Table 12.1). The newly emerged SARS CoV is probably one distantly related to group 2 viruses (Chapter 8). A morphological distinction between some group 2 CoVs and the other CoV groups is a double layer of surface projections, the shorter haemagglutinin (HE) and the longer spike (S) as apparent for bovine CoV (Fig. 12.1A) compared with only the spike for TGEV (a group 1 virus) (Fig. 12.1B; Table 12.1). SARS CoV lacks the HE surface protein. It has been reported1 that polyclonal antisera to the group 1 CoVs, transmissible gastroenteritis virus (TGEV) and feline infectious peritonitis virus (FIPV), but not to group 2 CoVs, cross-reacted with SARS CoV-infected cells, suggesting a potential antigenic relationship with group 1 CoV. Preliminary data (XJ Meng, personal communication) suggest that this cross-reactivity may reside in the N protein. A precursor animal SARS-like coronavirus has been identified (Chapter 11) suggesting that SARS coronavirus emerged from an animal reservoir. Other new human and animal group 1 coronaviruses have been recently documented.2–4 The emergence of SARS illustrates that
84
CoV can cause potentially fatal disease in humans as previously recognized for animal CoVs (Table 12.1). Because pneumonia and diarrhoea occur in SARS patients, this review will focus on animal CoVs that cause respiratory or respiratory/enteric disease (Table 12.1) because these may provide an insight into the pathogenesis and evolution of SARS.
Coronavirus evolution and pathogenesis Group 1 porcine CoVs: models for enteric or respiratory CoV disease New strains with altered tissue tropism can arise from existing strains through mutation. For example, the porcine respiratory coronavirus (PRCV) is a less virulent variant of TGEV and feline infectious peritonitis virus (FIPV) is the virulent variant of feline enteric coronavirus (FECoV).5,6 Alternatively, new strains may occur after recombination events such as the potential S gene recombinants between canine coronavirus (CCoV) and FECoV type 1 leading to a new FECoV serotype (type 2)6,7 or the acquisition of an influenza group C-like haemagglutinin (HE) by BCoV or its CoV ancestor.8 In addition, like SARS, new animal strains have emerged from unknown sources such as the porcine epidemic diarrhoea virus (PEDV) that first appeared in Europe and Asia between 1978 and the 1980s. It initially caused high diarrhoeal mortality in suck-
Table 12. 1 Coronavirus genetic and antigenic groups, target tissues and diseases Genetic group
Virus
Host
Disease/infection site
VN
CoV
N
S
Respiratory
Entericb
X SI
Otherc
HCoV-229E TGEV PRCV PEDV FIPV FCoV CCoV RaCoV HCoV-NL63
Human Pig Pig Pig Cat Cat Dog Rabbit Human
+ + + + + ?
+ + + ? + + + + ?
+ + + + + + + ? ?
? + +/? + + + ? ?
X upper X upper X upper/lower
HCoV-OC43 MHV RCoV (sialodocryadenitis) HEV BCoV**
Human Mouse Rat
? ? ?
+ ? ?
+ ? ?
+/-
X upper
Pig Cattle
? ?
+ +
? +
? +/-
X X lung
III
IBV TCoV (TECoV)
Chicken Turkey
? ?
+ +
? ?
? ?
X upper
X X SI
Kidney Oviduct
Pending classification
SARS Civet cat CoV
+ +
+ +
+ +
? ?
X lung X
X SI, colon X
Viraemia, kidney? Subclinical?
Putative group 2
Raccoon dog CoV
Human Himalayan palm civet Raccoon dog
+
+
+
?
?
X
Subclinical?
I
II
a
X upper
Viraemia X SI, Colon X SI X SI X SI
?? (BCoV?) X
X
Systemic Systemic
Hepatitis, CNS, systemic Eye, salivary glands CNS
X SI, Colon
85
One or two-way cross-reactions by virus neutralization (VN), and in serologic tests (ELISA, IF) versus intact virus (CoV), the nucleoprotein (N) or the spike (S) protein including cross-reactivities defined using monoclonal antibodies.? = unknown or unreported. b SI = small intestine; ?? = BCoV-like CoV from a child, Zhang et al. 1994.40 c CNS = central nervous system.
Comparative Biology of Animal Coronaviruses: Lessons for SARS
Within-group antigenic cross-reactionsa
86
Severe Acute Respiratory Syndrome
A
B
Figure 12.1 Electron micrograph of negatively stained CoV particles. (A) A typical BCoV particle showing shorter HE (top of particle, arrow) and longer S (bottom and sides) surface projections. (B) A typical TGEV particle showing single layer of longer S surface projections, similar to SARS CoV. Bar represents 100 nm.
ling pigs, but it is now endemic causing only mild or subclinical infections in previously PEDV seropositive herds.9 PEDV appears to be more closely related to human CoV 229E than to the other animal group 1 CoVs10 and, like SARS, but unlike the other group 1 CoVs, it grows in Vero cells.11
TGEV and PRCV infections as models for changes in tissue tropism TGEV (Fig. 12.1B) causes potentially fatal gastroenteritis in young pigs, targeting the small intestinal epithelial cells, and leading to severe villous atrophy and malabsorptive diarrhoea (Tables 12.1 and 12.2). The virus also replicates in the upper respiratory tract with transient nasal shedding,12 but infection or lesions in the lung are uncommon.13 The disease is mild in adults with transient diarrhoea or inappetence, but pregnant or lactating animals develop more severe clinical signs including agalactia13 similar to winter dysentery CoV infections in dairy cattle.14,15 Deletion mutants of TGEV with varying
size S gene deletions (PRCV) appeared independently in Europe in 1984 and in the US in 198913,15 with a pronounced tropism for the lower respiratory tract and little intestinal replication. A major deletion occurred at the 5’ end of the S gene (nt 45–752), ranging from 621–681 nt in size. Smaller deletions occurred preceding or in ORF 3a (encoding an undefined NS protein) leading to its lack of expression.13 Otherwise TGEV and PRCV viruses share high nt (96%) sequence identity. This is reminiscent of SARS CoV and its precursor animal virus where the human strains have acquired a deletion in the ORF8 gene region.16 Truncation of the S gene of TGEV also led to loss of antigenic site D, permitting use of monoclonal antibodies (MAbs) to site D to differentiate serologically TGEV and PRCV antibodies by blocking ELISA.13 Conventional virus neutralization (VN) tests do not discriminate between these viruses because the immunodominant neutralizing antigenic site (A) is conserved on PRCV and TGEV. The altered tissue tropism (respiratory)
Infected tissues
Viraemia Upper respiratory tract Lower respiratory tract Intestine S- gene
a
Coronavirus Macaquea
Cata
Ferreta
Pigs
SARS
SARS
SARS
TGEV-V
NRb +
NR +
NR +
+
++
+
+
+
+/-
+
+/- (1/4) Intact
+/Intact
+/Intact
+++ Intact
TGEV-A (vaccine)
+ Pt mutationsc (nt 214 and 655)
PRCV
Cattle BCoV-E
BCoV-R
+ ++
NR +
NR ++
+++
+
+++
+/Deletion (621–682 nt)
Chapters 7 and 10; bNR = Not reported; cBallesteros et al., 1997 ;17 dHasoksuz et al., 2002;36 d,eChouljeuko et al., 2001.36a
+++ (colon) ++ (colon) Pt mutationsd,e (42 aa changes at 38 sites)e
Comparative Biology of Animal Coronaviruses: Lessons for SARS
Table 12.2 Tissues infected by respiratory/enteric coronaviruses in animal hosts or models and changes in the S gene
87
88
Severe Acute Respiratory Syndrome
and reduced virulence of the PRCV variant have been attributed mainly to the S gene deletion.17 Use of recombinants between enteric and respiratory TGEV strains (attenuated TGEV) demonstrated that a substitution in aa 219 of the S protein was associated with loss of enteric tropism. PRCV spreads by the aerogenic route or via droplets such as SARS CoV. It further resembles SARS CoV in its pronounced tropism for the lung, replicating to titres of 107–108 TCID50 and producing interstitial pneumonia affecting 5–60% of the lung.13,18 Despite the invariable presence of lung lesions, many PRCV infections are clinically mild or subclinical, although PRCV strains with smaller deletions in the S gene (621 nt) and intact gene 3 reportedly produced more severe disease.19,20 Clinical signs of PRCV, like those reported for SARS, include fever and variable degrees of dyspnoea, polypnoea, anorexia and lethargy.13,18 Coughing and rhinitis are less common. Also like SARS,21 PRCV targets lung epithelial cells and alveolar macrophages.13,18 Lung infection leads to interstitial pneumonia with bronchiolar infiltration of mononuclear cells, lymphohistiocytic exudates and epithelial cell necrosis. Transient viraemia occurs and PRCV also has been isolated from nasal swabs, tonsils and trachea. Like SARS CoV1,22 PRCV replicates in undefined cells in the intestinal lamina propria unaccompanied by villous atrophy but unlike SARS PRCV infections result in limited faecal shedding and no diarrhoea. Recently faecal isolates of PRCV were found with minor (point mutations) but consistent genetic changes in the S gene compared to nasal isolates from the same pig.23 These findings suggest the presence of CoV quasispecies in a host with some strains more adapted to the intestine, a corollary potentially applicable to the faecal shedding of SARS CoV.1,22 It is notable that the more virulent TGEV infections have been displaced following the widespread dissemination of PRCV in Europe. PRCV can disappear from herds in summer and re-emerge in older pigs in the
winter,13,18 presumably circulating in pigs as subclinical infections in the summer; this fact is of interest in the study of SARS. Several cofactors exacerbate PRCV or TGEV infections or shedding. Underlying disease or respiratory co-infections, dose and route of infection and immunosuppression (corticosteroids) are all potential cofactors related to the severity of SARS. These cofactors can also exacerbate the severity of TGEV or PRCV infections.13 These cofactors may also be relevant in the super-spreader phenomenon seen in the SARS epidemic. 1 Impact of route (aerosols) and dose on PRCV infections. Studies of experimental inoculation of pigs with PRCV strains indicate that administration of PRCV by aerosol compared to the oronasal route, or in higher doses, resulted in higher virus titres shed and longer shedding.12 Similarly in two other studies, high PRCV doses induced more severe respiratory disease. Pigs given 108.5 TCID50 of PRCV had more severe pneumonia and deaths than pigs exposed by contact24 and higher intranasal doses of another PRCV strain (AR310) induced moderate respiratory disease whereas lower doses produced subclinical infections.19 The analogy to SARS CoV is that the route and dose of exposure may modulate or enhance the clinical disease.25 2 Impact of respiratory viral co-infections on PRCV infections. Co-infections of SARS CoV and other respiratory pathogens such as human metapneumovirus, rhinovirus and chlamydia have been noted (Chapter 7). The interaction between PRCV and other respiratory viruses in pigs may therefore be pertinent. Hayes et al.26 showed that sequential dual infections of pigs with the arterivirus (order Nidovirales, like CoV), PRRSV followed in 10 days by PRCV significantly enhanced lung lesions and reduced weight gains compared to each virus alone. The dual infections also led to more pigs shedding PRCV nasally for a prolonged period and, surprisingly, to faecal shedding of PRCV. The lung lesions observed resembled those in SARS victims.21
Comparative Biology of Animal Coronaviruses: Lessons for SARS
In another study, Van Reeth and Pensaert27 inoculated pigs with PRCV followed in 2–3 days by swine influenza A virus (SIV). They found that SIV lung titres were reduced in the dually infected pigs compared to those that were singly infected, but paradoxically the lung lesions were more severe in the dually infected pigs. They postulated that the high levels of IFN-a induced by PRCV may mediate interference with SIV replication, but might also contribute to the enhanced lung lesions. This is relevant to the proposed treatment of SARS patients with IFN-a (Chapter 20). 3 Impact of respiratory bacterial co-infections on PRCV infections. Respiratory viral infections enhance the potential for bacteria to colonize the lower respiratory tract in animals and humans. The outer membrane of gram-negative bacteria contains endotoxin or lipopolysaccharide (LPS) which is released in the lungs during bacterial infection or potentially after antibiotic treatments such as those commonly used in SARS patients.25 Bacterial LPS is a potent inducer of proiniflammatory cytokines. Van Reeth et al.28 showed that pigs infected with PRCV followed by a subclinical dose of E. coli LPS within 24 hours developed enhanced fever and more severe respiratory disease compared to each agent alone. They concluded that the effects were probably mediated by the significantly enhanced levels of proinflammatory cytokines induced. Thus there is a need to examine both LPS and lung cytokine levels in SARS patients as possible mediators of the severity of SARS. 4 Impact of treatment with corticosteroids on CoV infections of animals. Corticosteroids are known to induce immunosuppression and reduce the numbers of CD4 and CD8 T cells and certain cytokine levels.29 Many hospitalized SARS patients were treated with steroids to reduce lung inflammation, but there are no data to assess the outcome of this treatment on virus shedding or respiratory disease. Tsunemitsu et al.30 reported a recrudescence of BCoV faecal shedding in one of four winter dysentery BCoV-
89
infected cows treated with dexamethasone. Similarly, treatment of older pigs with dexamethasone prior to TGEV challenge led to profuse diarrhoea and reduced lymphoproliferative responses only in the treated pigs.31 These data raise issues for corticosteroid treat-ment of SARS patients related to possible transient immunosuppression leading to enhanced respiratory disease or increased and prolonged CoV shedding. Alternatively, corticosteroid treatment may be beneficial in reducing proinflammatory cytokines if they play a major role in lung immunopathology.29
Group 1 feline CoV (FCoV) as models for systemic and persistent CoV infection Historically, two types of FCoVs have been recognized: feline enteric CoV (FECoV) and FIPV. Current information suggests that the two viruses are biotypes of a prototype FCoV and that the FECoV which causes acute enteric infections in cats establishes persistent infection in some cats, evolving into the systemic virulent FIPV in 5–10% of cats.5,6 The initial site of FCoV replication is in the pharyngeal, respiratory or intestinal epithelial cells.7,32 Clinical signs include anorexia, lethargy and mild diarrhoea with villous atrophy in the jejunum and ileum in severe cases. The prolonged incubation period for FIP and its reactivation upon exposure to immunosuppressive viruses or corticosteroids suggested that FCoVs could cause chronic enteric infections in cats.7,32 Recent reports of chronic faecal shedding and persistence of FCoV mRNA or antigen in blood, ileum, colon and rectum of FCoVinfected cats for prolonged periods (up to 7 months) confirm this scenario.5 A key pathogenetic event for development of FIP is productive infection of macrophages with cell-associated viraemia and systemic dissemination of virus.7,32 Stress (immunosuppressive infections, cat density, transport to new environments) leading to immune suppression may trigger FIP in chronically infected cats, similar to its role in shipping fever CoV infections of cat-
90
Severe Acute Respiratory Syndrome
tle. Two major forms of FIP are recognized: the effusive form with a fulminant course and death within weeks to months and the non-effusive form that progresses more slowly. The effusive form is characterized by fibrin-rich fluid accumulation in peritoneal, pleural, pericardial or renal spaces with fever, anorexia and weight loss. Noneffusive FIP involves pyogranulomatous lesions with thrombosis, CNS or ocular involvement. Fulminant FIP with accelerated early deaths appears to be immunemediated in FCoV seropositive cats and can be enhanced by IgG antibodies to the S protein, although other contributing immune factors (inflammatory mediators such as cytokines, leukotrienes and prostaglandins initiated by C’ activation or released by infected macrophages) may also play a role.7,32 At least two mechanisms implicating IgG antibodies to FCoV S protein in FIP immunopathogenesis have been described. In the first, circulating immune complexes (IC) with C’ depletion in sera and IC in lesions are evident in cats with terminal FIP.7 For the second, antibody-dependent enhancement (ADE) of FCoV infection of macrophages has been described in vitro as mediated by neutralizing IgG MAbs to the S protein of FIPV, or of interest, to the antigenically related CoV, TGEV.33 Similar accelerated disease was seen in vivo in cats inoculated with recombinant vaccinia virus expressing the S protein (but not the M or N proteins) of FIPV.7,32 Thus the spectrum of disease evident for FCoV/FIPV exemplifies the impact of viral persistence and macrophage tropism on CoV disease progression and severity. The scenario of immune-enhanced disease is one to be kept in mind in the development of SARS vaccines.
Antigenic relationships and cross-species transmission: the example of group 1 coronaviruses Within group 1 CoVs, TGEV, PRCV, CCoV and FCoV share close biological, antigenic
and genetic relationships and they may represent host range mutants of an ancestral CoV.13,34 The four CoVs cross-react in VN and IF tests and with MAbs to the S, N or M proteins (Table 12.1).13 Duarte et al.10 reported that the group I CoV PEDV is genetically more closely related to human CoV 229E than to TGEV, raising unanswered questions about its origin. Both CCoV and FIPV infect baby pigs, but with only the latter virus causing diarrhoea and intestinal lesions similar to those caused by TGEV.13 Cats infected with TGEV shed the virus in faeces and seroconverted to TGEV and FIPV;13 cats exposed to CCoV remained clinically normal and did not shed the virus, but seroconverted to CCoV and FIPV.32 Cats, dogs and foxes which seroconvert to TGEV were suggested as possible subclinical carriers of TGEV serving as reservoirs between seasonal (winter) epidemics, but only virus excreted by dogs was infectious for pigs.13 Birds (Sturnus vulgaris) and flies (Musca domestica) have been proposed as mechanical vectors for TGEV.13 These observations may be relevant to considerations of the ecology of the precursor animal SARS CoV.
Group 2 bovine CoVs (BCoV): models for pneumoenteric CoV infections The shedding of SARS in faeces of most patients and the occurrence of diarrhoea in 10–27% of patients (Chapter 3)25 suggests that SARS may be pneumoenteric like BCoV. BCoV (Fig. 12.1A) causes three distinct clinical syndromes in cattle: calf diarrhoea; winter dysentery with haemorrhagic diarrhoea in adults, and respiratory infections in cattle of various ages including cattle with shipping fever (Table 12.3).14,35,38 Genetic differences (point mutations but not deletions) have been detected in the S gene between enteric and respiratory isolates, including ones from the same animal (Table 12.2).36,36a Unlike SARS and group 1 CoV, BCoV possesses a double layer of surface projections: the
Table 12.3 Summary of disease syndromes associated with BCoV infections Disease syndrome
Clinical signs
Cells infected
Lesionsa
Sheddingb
Respiratory
Enteric
Nasal
Faecal
Ages affected
Diarrhoea Dehydration Fever/anorexia
Intestinal/ nasal/± lung epithelial cells
+/lung emphysema
++ J, I, Colon Villous atrophy
2–8 days
2–8 days
Birth to 4 weeks
Winter dysentery
Haemorrhagic Diarrhoea Dehydration ± Rhinitis/dry cough Fever/anorexia
Intestinal/ nasal/± lung? epithelial cells
NR
++ J, I, Colon Enterocolitis
NR
1–4 days
6 months to adult
Calf pneumonia
Cough Rhinitis ± Pneumonia ± Diarrhoea Fever/anorexia
Nasal/± lung Tracheal ± Intestinal epithelial cells
+/pneumonia
+/J, I, Colon Villous atrophy
NR
NR
2 wks to 6 months
Shipping fever
Cough/dyspnoea ± Rhinitis ± Pneumonia ± Diarrhoea Fever/anorexia
Nasal/trachea Bronchi/alveoli ± Intestinal epithelial cells
Interstitial emphysema Bronchiolitis Alveolitis
+/J, I, Colon Villous atrophy
5–10 days (17 days)
4–8 days (17 days)
6–10 months
J = jejunum, I = ileum; NT = not tested. Shedding detected by infectivity or antigen assays; parentheses denote shedding detected by RT-PCR; NR = not reported. In experimental challenge studies, the incubation periods for disease onset and shedding ranged from 3 to 8 days. a
b
Comparative Biology of Animal Coronaviruses: Lessons for SARS
Calf diarrhoea
91
92
Severe Acute Respiratory Syndrome
shorter haemagglutinin (HE) and the longer spike (S), both of which function in viral attachment and fusion, induction of VN antibodies and immunity and haemagglutination of erythrocytes. Whether the HE influences the respiratory tropism or the broad host range of BCoV is unclear. Of interest, the HE of BCoV has homology with the HE of group C influenza viruses suggesting a prior recombination event between the two viruses.8 Of concern is whether similar recombinants could arise between SARS CoV and influenza strains if co-infections were to occur in the future.
Calf diarrhoea and calf respiratory BCoV infections Calf diarrhoea BCoV strains infect the epithelial cells of the distal small and large intestine and superficial and crypt enterocytes of the colon leading to villous atrophy and crypt hyperplasia.15,37 After an incubation period of 3–4 days, calves develop a severe, malabsorptive diarrhoea persisting for 3–8 days and resulting in dehydration and often death. Concurrent faecal and nasal shedding can occur. BCoV are also implicated as a cause of mild respiratory disease (coughing, rhinitis) or pneumonia in 2–24-month-old calves and are detected in nasal secretions, lung and often the intestines.15 More recent studies have implicated BCoV in association with respiratory disease (shipping fever) in feedlot cattle.35,38 BCoV was isolated from nasal secretions and lungs of cattle with pneumonia39 and from faeces.39a In a subsequent study, a high percentage of feedlot cattle (45%) shed BCoV both nasally and in faeces.39a Shipping fever is recognized as a multifactorial, polymicrobial respiratory disease complex in feedlot cattle with several factors exacerbating respiratory disease, including BCoV infections as well as stress.35,38,39,39a
Group 2 BCoVs: cross-species transmission The likelihood that SARS CoV is a zoonotic infection potentially transmitted from wild animals to humans is not surprising in light of previous research on interspecies transmission of BCoV including wildlife reservoirs. Although many CoVs have restricted host ranges, some such as BCoV appear to be promiscuous. In 1994, Zhang et al.40 isolated a human enteric CoV from a child with acute diarrhoea (HECoV-4408) which was genetically (99% nt identity in the S and HE gene with BCoV) and antigenically more closely related to BCoV than to HCoVOC43, suggesting this isolate is a BCoV variant. Tsunemitsu et al.41 isolated CoV, antigenically closely related to BCoV by two-way cross-neutralization tests, from captive wild ruminants in Ohio, USA including Sambar deer (Cervus unicolor), white-tailed deer (Odocoileus virginianus) and a waterbuck (Kobus ellipsiprymnus) with bloody diarrhoea resembling winter dysentery in cattle. In addition, CoVs antigenically related to BCoV were isolated from elk and wapiti (Cervus elephus) in the western USA.42 A more dramatic demonstration of the broad host range of BCoV was the experimental induction of infection and diarrhoea in SPF baby turkeys and contact controls, but not in chicks, with an enteric strain of BCoV43 documenting CoV transmission from mammalian to avian species. More recent data suggest that CoVs genetically closely related to BCoV also occur in dogs with kennel cough.44 Reasons for the broad host range of BCoV are unknown, but might relate to the presence of the HE on BCoV and its possible role in virus binding to diverse cell types.
Group 3 CoVs. Infectious bronchitis virus (IBV): model for respiratory CoV infection with other target tissues IBV is a highly contagious respiratory
Comparative Biology of Animal Coronaviruses: Lessons for SARS
disease of chickens, like SARS, spread by aerosol or possibly faecal–oral transmission, and with a worldwide distribution.45,46 Genetically and antigenically closely related CoV have been isolated from pheasants and turkeys,47,48 but in young turkeys, they cause mainly enteritis. Respiratory infections of chickens are characterized by tracheal rales, coughing and sneezing with the disease most severe in chicks.45,46 IBV also replicates in the oviduct causing decreased egg production or quality. Nephropathogenic strains cause mortality in young birds, whereas in broilers death ensues from systemic E. coli infections after IBV damage to the respiratory tract. IBV replicates in epithelial cells of the trachea and bronchi, intestinal tract, oviduct and kidney, causing necrosis and oedema with small areas of pneumonia near large bronchi in the respiratory tract and interstitial nephritis in the kidney.45,46 Of interest in SARS investigations is the persistence of IBV in the kidney and its prolonged faecal shedding since SARS CoV is detected in urine and shed longer term in faeces. Importantly, the respiratory tropism of one serotype of IBV was altered to kidney tropism by in vivo serial passage of virus via the cloacal route.49 Both diagnosis and control of IBV are complicated by the existence of multiple serotypes and the occurrence of IBV recombinants.45,46 This is unlike the scenario for most group one or two respiratory CoVs in which only one or two (FCoV) serotypes are known. Also relevant to SARS CoV is the finding that IBV strains also replicate in Vero cells, but only after passage in chicken embryo kidney cells.45
Animal respiratory or enteric CoVs: treatments and vaccines Treatments with IFN-a Interferons (IFNs) are of major interest for treatment of patients with SARS, but their potential effectiveness is unknown. Human
93
recombinant IFN-a (rHuIFN-a) inhibited FIPV replication in vitro.32 However, in vivo rHuIFN-a plus an immunomodulating drug failed to protect cats significantly against fatal FIPV disease, although the treatment suppressed clinical signs and prolonged survival time in cats.50 Similarly during a field outbreak of TGEV, 1–23-day-old pigs treated orally for 4 days with 1–20 IU of rHuIFN-a had significantly greater survival rates than had placebo pigs.51 However, in piglets given rHuIFN-a shortly after birth, there was no increased survival. Thus in vivo treatment of CoV-infected animals with rHulFN-a produced variable results.
Animal CoV vaccines 1 Enteric CoV vaccines. Major efforts will probably be focused on development of SARS CoV vaccines. An understanding of the pathogenesis of SARS CoV infections including the target organs infected and how the virus is disseminated to these organs will assist in the development of effective vaccine strategies to block viral dissemination and protect the target organs. In monogastrics (pigs, humans) which secrete SIgA antibodies in milk, vaccination is accomplished by exploiting the common mucosal immune system. Because neutralizing SIgA antibodies to TGEV in milk are a correlate of protection to TGEV, the strategy is to evoke the gut-mammary IgA axis (first described in studies of immunity to TGEV; reviewed in ref. 13, 51a) by administering attenuated TGEV (TGEV-A) vaccines orally to induce SIgA antibodies in milk via intestinal stimulation of the mother. Problems13 were encountered in the field application of this strategy such as poor titre and immunogenicity. Use of less attenuated TGEV strains or the antigenically related FIPV caused disease in baby pigs. These studies illustrate further the difficulty in priming for protective SIgA mucosal immune responses, even using live vaccines, in naïve
94
Severe Acute Respiratory Syndrome
seronegative animals. However, in comparison, killed TGEV vaccines given parenterally (IM) induced only IgG antibodies in milk and no protection against TGEV. Interestingly, a single infection of the respiratory tract of pigs or sows with the TGEV deletion mutant, PRCV, induced only partial active or passive immunity to TGEV, respectively13,52 but repeated PRCV infections of the mother induced higher IgA milk antibody responses and protection rates.53 Van Cott et al.52 found that in young pigs this was because a single PRCV infection of the respiratory tract induced few IgA antibody secreting cells (ASC) in the intestine, but higher numbers of IgG ASC in the lower respiratory tract (bronchial lymph nodes), and primed for anamnestic IgG and IgA intestinal antibody responses after TGEV challenge leading to the partial protection observed. In the field, pigs experience multiple respiratory infections with PRCV providing sufficient immunity to TGEV such
that TGE has largely disappeared from European swine herds.13 SARS CoV frequently causes intestinal infections as well as pneumonia. Experience with animal coronavirus vaccines suggest that neither killed parenteral nor respiratory applied vaccines may prevent the diarrhoeal disease or faecal shedding. Attempts have also been made to develop TGEV subunit vaccines to induce active immunity to TGEV in older pigs (Table 12.4). 54–56 It is likely that similar strategies may be devised for SARS CoV vaccines. Again problems were encountered in providing effective active immunity against TGEVinduced diarrhoea as summarized by the protection data in Table 12.4. Two subcutaneous doses of two different baculovirusexpressed S glycoprotein constructs containing the four major antigenic sites (including the immunodominant site A) elicited neutralizing antibodies in serum, but failed to induce any protection against
Table 12.4 Active immune responses and protection in pigs to recombinant TGEV spike (S) glycoprotein with or without the N and M proteinsa Inoculum b (50 mg/dose)
Inoculation Adjuvantd routec
VN Ab serume
Intestinal (MLN)f ASC responses
Diarrhoea morbidity
SA–D (1449 AA) SA–D (789 AA) SC+D (397 AA)
SC 2X
IFA
Yes
NT
100%
SC 2X
IFA
Yes
NT
100%
SC 2X
IFA
No (PC)
NT
100%
S+N+M Killed TGEV (vaccine) TGEV-V Mock
IP 2X IP 2X Oronasal IP 2X
IFA + mLT R192G Undefined None IFA + mLT
Yes Low Yes No
IgA None IgA/IgG None
Faecal shedding 38% 86% 0% 100%
aShoup et al., 1997;55 Sestak et al., 1999.54 b
Truncated forms of baculovirus expressed TGEV S glycoprotein included the four major antigenic sites (A–D) or only sites C + D were tested. For baculovirus expressed S + N + M, S (789 AA), N and M proteins were mixed and tested. TGEV-V = virulent TGEV; mock = control. c SC = subcutaneous; IP = intraperitoneal. dIFA = Incomplete Freund’s adjuvant; mLT R192G = mutant E. coli heat labile toxin lacking toxicity; Undefined = commercial vaccine adjuvant. e VN Ab = Neutralizing antibody; (PC) = post-challenge only. fMLN = Mesenteric lymph node; NT = not tested.
Comparative Biology of Animal Coronaviruses: Lessons for SARS TGEV-induced diarrhoea (Table 12.4).55 However, neutralizing antibodies induced against the recombinant S protein with antigenic site A given passively (orally) to TGEV-challenged pigs delayed the onset of diarrhoea and virus shedding.56 The data confirm earlier findings using killed TGEV vaccines, which indicated that serum neutralizing antibodies (in contrast to intestinal IgA antibodies) do not correlate with a high degree of protection against TGEV infection.13 However, in a subsequent study, partial protection against TGEV infection (faecal shedding) was induced in pigs vaccinated intraperitoneally with the S glycoprotein mixed with the N and M proteins (Table 12.4).54 Other studies of TGEV also suggested that both recombinant N proteins (T cell epitopes) and S proteins were required for maximal antibody responses to TGEV.57 Thus in spite of long-term research efforts, effective TGEV vaccines have remained elusive, but with the emergence of PRCV, nature appears to have generated its own highly effective vaccine for the more virulent TGEV infections. 2 Respiratory CoV vaccines. In spite of its economic impact, no respiratory CoV vaccines have been developed to prevent BcoVassociated pneumonia in calves or in cattle with shipping fever. The correlates of immunity to respiratory BCoV infections remain undefined. Limited data from epidemiological studies of BCoV infections in cattle suggest that serum antibody titres to BCoV may be a marker for respiratory protection. However, whether the serum antibodies are themselves correlates of respiratory protection or only reflect prior enteric or respiratory exposure to BCoV is uncertain. The only available animal CoV vaccines targeted to prevent respiratory CoV infections are IBV vaccines for chickens. Both live attenuated and killed commercial IBV vaccines are used.45,46 Attenuated vaccines are used in broilers, usually at 1 day of age and 10 days later, since only short-term (6–7 weeks) protection is needed. For layers or breeders where longer protection is needed
95
(~18 months), attenuated vaccines are used for priming at 2–3 weeks of age followed by injection of killed oil–emulsion vaccines, often at 8–10-week intervals throughout the laying cycle. The correlates and mechanisms of protection against IBV clinical disease are uncertain but neutralizing antibody does appear to be relevant.46 Evidence exists that the S1 glycoprotein, but not the N or M proteins of IBV can induce protection, although all three proteins induce cell-mediated immune responses to IBV.58 Problems encountered in vaccine protection include the existence of multiple serotypes/subtypes of IBV which may fail to cross-protect, variation in virulence among IBV field strains and the possible increase in virulence of some live vaccines after back-passage in chickens59 with the suggestion that point mutations in the genomes of attenuated vaccines may generate new epidemic strains of IBDV.60 (See Chapter 22.) 3 FIPV vaccines. Because of the immunopathogenesis of FIPV, vaccines for its control have been among the most problematical to develop. Most conventional FIPV vaccine approaches (killed and attenuated) have failed and, in fact, they induced accelerated disease and reduced survival times.7,32 These adverse effects have been attributed to antibodies to the S protein which can mediate immune complex disease or ADE of infection. A recombinant vaccinia virus expressing the S protein also mediated this effect. The efficacy of a commercially available temperature-sensitive FIPV vaccine is also debated, although there is no evidence that it causes accelerated FIP. Attempts to circumvent use of the S protein, by priming with DNA vaccines containing the M and N proteins (augmented by codelivery of plasmids encoding feline IL12) and then boosting with recombinant vaccinia virus encoding the N and M proteins also failed and even enhanced susceptibility of the vaccinated cats to FIP.61 Thus the development of safe and effective FIPV vaccines remains elusive.
96
Severe Acute Respiratory Syndrome
Concluding remarks Enteric coronaviruses alone can cause fatal infections in seronegative young animals. However, in adults respiratory CoV infections are more often fatal or more severe when combined with other factors including high exposure doses, aerosols, treatment with corticosteroids and respiratory co-infections (viruses, bacteria, LPS). These variables may also influence the severity of SARS and transmission of SARS. Studies of animal CoVs have highlighted the potential for new CoV strains to emerge as deletion mutants or recombinants from existing strains or for new strains to appear from unknown or perhaps wildlife reservoirs, the latter a likely origin for SARS CoV. A number of CoV strains, particularly ones from wild animals, remain to be characterized and the full genomic sequence is available for only a small number of human and animal CoVs. In addition, interspecies transmission of certain CoVs may not be uncommon, although the determinants of host range specificity among CoVs are undefined. Early examples of CoVs with broad host range include TGEV, CCoV and FCoV which appear to be host-range mutants of an ancestral CoV. Even more promiscuous are BCoVs which cross-infect diverse species from wild ruminants to baby turkeys and appear as genetically similar strains in dogs and even humans. Thus it is not unprecedented for new CoV strains to emerge or for interspecies transmission of CoVs to occur. As a reminder of this potential disease threat, it is estimated that 75% of emerging human pathogens are zoonotic,62 but we understand very little about CoVs or other viruses circulating in wildlife or their potential to emerge as either public or animal health threats. Development of safe and efficacious vaccines for animal CoV infections has been problematic and only partially successful. Problems encountered often relate to a lack of understanding of basic mechanisms to induce mucosal immunity by vaccines tar-
geted at preventing enteric or respiratory mucosal infections. Stimulation of protective mucosal immunity, especially priming of seronegative vaccinees, often requires use of live replicating vaccines or vectors as opposed to non-replicating killed viruses or subunit vaccines (unless applied with effective mucosal delivery systems or adjuvants) to provide optimal mucosal antigenic stimulation and to avoid tolerance induction.13,51a In addition, most vaccines for mucosal pathogens may fail to induce sterilizing immunity or prevent reinfections, as commonly observed for natural CoV mucosal infections, and the major vaccine focus may be to prevent severe disease. Although early studies of immunity to TGEV infections provided evidence for new immunologic linkages (gut–mammary axis and a common mucosal immune system, reviewed in ref 13, 51a), subsequent studies of TGEV/PRCV demonstrated compartmentalization within the respiratory and intestinal components of the common mucosal immune system, influencing the protection levels induced and future strategies for mucosal vaccines.12,52 An understanding of CoV disease pathogenesis is critical for the design of effective vaccine strategies. For SARS, many unanswered questions remain: these include the following. What is the initial site of viral replication and is SARS pneumoenteric like BCoV or primarily targeted to the lung like PRCV with faecal shedding of swallowed virus and with other sequelae contributing to the diarrhoea reported? Does SARS CoV infect the lung directly or via viraemia and does it productively infect secondary target organs (intestine, kidney) via viraemia after replication in lung? For both TGEV and IBV infections, live vaccines alone (TGEV) or for priming followed by killed vaccines for boosting (IBV) provided at least partial protection against enteric and respiratory disease, respectively. But as illustrated for IBV, live vaccines may revert to virulent if inadequately attenuated, raising safety issues. Finally, the macrophage-tropic, systemic
Comparative Biology of Animal Coronaviruses: Lessons for SARS
FIPV CoV infection of cats presents yet another vaccine dilemma in that neutralizing IgG antibodies to FIPV, not only fail to protect, but actually potentiate the immunopathogenesis of FIPV. In summary, although much progress has been made in the comparative biology of animal coronaviruses that is applicable to SARS CoV infections, much remains unknown as highlighted in this chapter. Perhaps the SARS epidemic will generate new interest in these fundamental research questions related not only to CoV infections, but also to other infectious diseases of humans and animals.
11
12
13
14
15
References 1
2
3
4
5
6
7
8
9
10
Ksiazek TG, Erdman D, Goldsmith CS et al. A novel coronavirus associated with severe acute respiratory syndrome. N Engl J Med 2003;348: 1953–66. van der Hoek L, Pyrc K, Jebbink MF et al. Identification of a new human coronavirus. Nat Med 2004;10: 368–73. Poon LL, Chu DK, Chan KH et al. Identification of a novel coronavirus in bats. Virology 2005;79: 2001–9. Woo PC, Lau SK, Chu CM et al. Characterization and complete genome sequence of a novel coronavirus, coronavirus HKV1, from patients with pneumonia. Virology 2005;79: 884–95. Herrewegh AAPM, Mahler M, Hedrich HJ et al. Persistence and evolution of feline coronavirus in a closed cat-breeding colony. Virology 1997;234: 349–63. Vennema H, Poland A, Floyd Hawkins K et al. A comparison of the genomes of FECVs and FIPVs and what they tell us about the relationships between feline coronaviruses and their evolution. Feline Pract 1995;23: 40–4. deGroot RJ, Horzinek MC. Feline infectious peritonitis. In: Siddell SG, ed. The Coronaviridae. New York: Plenum Press, 1995: 293–315. Brian DA, Hogue BG, Kienzle TE. The coronavirus hemagglutinin esterase glycoprotein. In: Siddell SG, ed. The Coronaviridae. New York: Plenum Press, 1995: 65–79. Pensaert, MB. Porcine epidemic diarrhea. In: Straw B et al., eds. Diseases of Swine, 8th edn. Ames, IA: Iowa State University Press, 1999: 179–85. Duarte M, Tobler K, Bridgen A et al. Sequence analysis of the porcine epidemic diarrhea virus genome between the nucleocapsid and
16
17
18
19
20
21
22
23
24
97
spike protein genes reveals a polymorphic ORF. Virology 1994;198: 466–76. Hoffman M, Wyler R. Propagation of the virus of porcine epidemic diarrhea in cell culture. J Clin Microbiol 1988;26: 2235–9. Van Cott JL, Brim TA, Simkins RA et al. Antibody-secreting cells to transmissible gastroenteritis virus and porcine respiratory coronavirus in gut- and bronchus-associated lymphoid tissues of neonatal pigs. J Immunol 1993;150: 3990–4000. Saif L, Wesley R. Transmissible gastroenteritis virus. In: Straw B et al., eds. Diseases of Swine, 8th edn. Ames, IA: Iowa State University Press, 1999. Saif LJ. A review of evidence implicating bovine coronavirus in the etiology of winter dysentery in cows: an enigma resolved? Cornell Vet 1990;80: 303–11. Saif LJ, Heckert RA. Enteric coronaviruses. In: Saif LJ, Theil KW, eds. Viral Diarrheas of Man and Animals. Boca Raton, Fla: CRC Press, 1990: 185–252. Guan Y, Zheng BJ, He YQ et al. Isolation and characterization of viruses related to the SARS coronavirus from animals in southern China. Science 2003;302: 276–8. Ballesteros ML, Sanchez CM, Enjuanes L. Two amino acid changes at the N-terminal of the transmissible gastroenteritis coronavirus spike protein result in the loss of enteric tropism. Virology 1997;227: 378–88. Pensaert M, Callebaut P, Vergote J. Isolation of a porcine respiratory, non-enteric coronavirus related to transmissible gastroenteritis. Vet Q 1986;8: 257–61. Halbur PG, Paul PS, Vaughn EM et al. Experimental reproduction of pneumonia in gnotobiotic pigs with porcine respiratory coronavirus isolate AR310. J Vet Diagn Invest 1993;5: 184–8. Vaughn E, Halbur P, Paul PS. Three new isolates of porcine respiratory coronavirus with various pathogenicities and spike (S) gene deletions. J Clin Microbiol 1994;32: 1809–12. Nicholls JM, Poon LL, Lee KC et al. Lung pathology of fatal severe acute respiratory syndrome. Lancet 2003;361: 1773–8. Peiris JS, Chu CM, Cheng VC et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;361: 1767–72. Costantini V, Lewis P, Alsop J et al. Respiratory and enteric shedding of porcine respiratory coronavirus (PRCV) in sentinel weaned pigs and sequence of the partial S gene of the PRCV isolates. Arch Virol 2004;149: 957–74. Jabrane A, Girard C, Elazhary Y. Pathogenicity
98
25
26
27
28
29
30
31
32
33
34
35
36
Severe Acute Respiratory Syndrome
of porcine respiratory coronavirus isolated in Quebec. Can Vet J 1994;35: 86–92. Peiris JSM, Yuen KY, Osterhaus ADME et al. The severe acute respiratory syndrome. N Engl J Med 2003;349: 2431–41. Hayes J, Sestak K, Myers G et al. Evaluation of dual infection of nursery pigs with U.S. strains of porcine reproductive and respiratory syndrome virus and porcine respiratory coronavirus. Proc. VIIth International Symposium on Nidoviruses (Corona and Arteriviruses), Lake Harmony, PA, May 20–25, 2000. Van Reeth K, Pensaert MB. Porcine respiratory coronavirus-mediated interference against influenza virus replication in the respiratory tract of feeder pigs. Am J Vet Res 1994;55: 1275–81. Van Reeth K, Nauwynck H, Pensaert M. A potential role for tumour necrosis factor-alpha in synergy between porcine respiratory coronavirus and bacterial lipopolysaccharide in the induction of respiratory disease in pigs. J Med Microbiol 2000;49: 613–20. Giomarelli P, Scolletta S, Borrelli E et al. Myocardial and lung injury after cardiopulmonary bypass: role of interleukin (IL)-10. Ann Thorac Surg 2003;76: 117–23. Tsunemitsu H, Smith DR, Saif LJ. Experimental inoculation of adult dairy cows with bovine coronavirus and detection of coronavirus in feces by RT-PCR. Arch Virol 1999; 144: 167–75. Shimizu M, Shimizu Y. Effects of ambient temperatures on clinical and immune responses of pigs infected with transmissible gastroenteritis virus. Vet Microbiol 1979;4: 109–16. Olsen CW. A review of feline infectious peritonitis virus: molecular biology, immunopathogenesis, clinical aspects, and vaccination. Vet Microbiol 1993;36: 1–37. Olsen CW, Corapi WV, Jacobson RH et al. Identification of antigenic sites mediating antibody-dependent enhancement of FIPV infectivity. J Gen Virol 1993;74: 745–9. Horzinek MC, Lutz H, Pedersen NC. Antigenic relationships among homologous structural polypeptides of porcine, feline and canine coronaviruses. Infect Immun 1982;37: 1148– 55. Storz J, Stine L, Liem A et al. Coronavirus isolation from nasal swab samples in cattle with signs of respiratory tract disease after shipping. J Am Vet Med Assoc 1996;208: 1452–5. Hasoksuz M, Sreevatsan S, Cho KO et al. Molecular analysis of the S1 subunit of the spike glycoprotein of respiratory and enteric bovine coronavirus isolates. Virus Res 2002;84: 101–9.
36a Chouljenko VN, Lin XQ, Storz J et al. Comparison of genomic and predicted amino acid sequences of respiratory and enteric bovine coronaviruses isolated from the same animal with fatal shipping pneumonia. J Gen Virol 2001;82: 2927–33. 37 Van Kruiningen HJ, Khairallah LH, Sassevelle VG et al. Calfhood coronavirus enterocolitis: a clue to the etiology of winter dysentery. Vet Pathol 1987;24: 564. 38 Lathrop SL, Wittum TE, Brock KV et al. Association between infection of the respiratory tract attributable to bovine coronavirus and health and growth performance of cattle in feedlots. Am J Vet Res 2000;61: 1062–6. 39 Storz J, Purdy CW, Lin X et al. Isolation of respiratory bovine coronavirus, other cytocidal viruses, and Pasteurella sp from cattle involved in two natural outbreaks of shipping fever. J Am Vet Med Assoc 2000;216: 1599–604. 39a Cho KO, Hoet A, Loerch SC et al. Evaluation of concurrent shedding of bovine coronavirus via the respiratory and enteric route in feedlot cattle. Am J Vet Res 2001;62: 1436–141. 40 Zhang XM, Herbst W, Kousoulas KG. Biologic and genetic characterization of a hemagglutinating coronavirus isolated from a diarrhoeic child. J Med Virol 1994;44: 152–61. 41 Tsunemitsu H, Reed H, El-Kanawati Z et al. Isolation of coronaviruses antigenically indistinguishable from bovine coronavirus from a Sambar and white tailed deer and waterbuck with diarrhea. J Clin Microbiol 1995;33: 3264–9. 42 Majhdi F, Minocha HC, Kapil S. Isolation and characterization of a coronavirus from elk calves with diarrhea. J Clin Microbiol 1997;35: 2937–2942. 43 Ismail MM, Cho KO, Ward LA et al. Experimental bovine coronavirus in turkey poults and young chickens. Avian Dis 2001;45: 157–63. 44 Erles K, Toomey C, Brooks HW et al. Detection of a novel canine coronavirus in dogs with respiratory disease. Proc. IXth Intl Symp. on Nidoviruses, p. 49, The Netherlands, May 24–29, 2003. 45 Cavanagh D, Naqi S. Infectious bronchitis. In: Saif YM et al., eds. Diseases of Poultry, 11th edn. Ames, IA: Iowa State University Press, 2003: 101–19. 46 Cook J, Mockett APA. Epidemiology of infectious bronchitis virus. In: Siddell SG, ed. The Coronaviridae. New York: Plenum Press, 1995: 317–35. 47 Guy JS. Turkey coronavirus is more closely related to avian infectious bronchitis virus
Comparative Biology of Animal Coronaviruses: Lessons for SARS
than to mammalian coronaviruses. Avian Pathol 2000;29: 206–12. 48 Ismail M, Cho KO, Hasoksuz M et al. Antigenic and genomic relatedness of turkey-origin coronaviruses, bovine coronaviruses and infectious bronchitis virus of chickens. Avian Dis 2001;45: 978–84. 49 Uenaka T, Kishimoto I, Uemura T et al. Cloacal inoculation with the Connecticut strain of IBV: an attempt to produce nephropathogenic virus by in vivo passage using cloacal inoculation. J Vet Med Sci 1998;60: 495–502. 50 Weiss RC, Cox NR, Oostrom-Ram T. Effect of interferon or Propionibacterium acnes on the course of experimentally induced feline infectious peritonitis in specific-pathogen-free cats and random source cats. Am J Vet Res 1990;51: 726–33. 51 Cummins JM, Mock RE, Shive BW et al. Oral treatment of TGE with natural human interferon alpha: a field study. Vet Immunol Immunopathol 1995;45: 355–60. 51a Saif LJ. Mucosal immunity: an overview and studies of enteric and respiratory coronavirus infections in a swine model of enteric disease. Vet Immunol Immunopathol 1996;54: 163–9. 52 Van Cott J, Brim T, Lunney J, Saif LJ. Contribution of antibody secreting cells induced in mucosal lymphoid tissues of pigs inoculated with respiratory or enteric strains of coronavirus to immunity against enteric coronavirus challenge. J Immunol 1994;152: 3980– 90. 53 Sestak K, Lanza I, Park SK et al. Contribution of passive immunity to porcine respiratory coronavirus to protection against transmissible gastroenteritis virus challenge exposure in suckling pigs. Am J Vet Res 1996;57: 664–71.
54
55
56
57
58
59
60
61
62
99
Sestak K, Meister RK, Hayes JR et al. Active immunity and T-cell populations in pigs intraperitoneally inoculated with baculovirusexpressed transmissible gastroenteritis virus structural proteins. Vet Immunol Immunopathol 1999;70: 203–21. Shoup DI, Jackwood DJ, Saif LJ. Active and passive immune responses to transmissible gastroenteritis virus (TGEV) in swine inoculated with recombinant baculovirusexpressed TGEV spike (S) glycoprotein vaccines. Am J Vet Res 1997;58: 242–50. Tuboly T, Nagy E, Derkyshire JB. Passive protection of piglets by recombinant baculovirus induced TGEV specific antibodies. Can J Vet Res 1995;59: 70–2. Anton IM, Gonzalez S, Bullido MJ et al. Cooperation between TGEV structural proteins in the in vitro induction of virus-specific antibodies. Virus Res 1996;46: 111–24. Ignjatovic J, Galli L. The S1 glycoprotein but not the N or M proteins of avian infectious bronchitis virus induces protection in vaccinated chickens. Arch Virol 1994;138: 117–34. Hopkins SR, Yoder HW Jr. Reversion to virulence of chicken passaged infectious bronchitis vaccine virus. Avian Dis 1986; 30: 221–3. Kusters JG, Niesters HGM, Bleumink NM et al. Molecular epidemiology of IBV in the Netherlands. J Gen Virol 1987;68: 343. Glansbeek HL, Haagmans BL, te Lintelo EG et al. Adverse effects of feline IL-12 during DNA vaccination against feline infectious peritonitis virus. J Gen Virol 2002;83: 1–10. Taylor LH. Risk factors for human disease emergence. Philos Trans R Soc London Ser B 2001;356: 983–90.
Chapter 13
Epidemiology and Transmission of SARS Angela Merianos, Robert Condon, Hitoshi Oshitani, Denise Werker and Roberta Andraghetti
Chronology of the epidemic Severe acute respiratory syndrome (SARS) has been aetiologically linked to a novel coronavirus named the SARS-associated coronavirus (SARS CoV). Genome sequence analyses support the hypothesis that SARS is a zoonosis with an as yet unknown animal reservoir.1 The first cases of SARS emerged in mid-November 2002 in the southern Chinese province of Guangdong. On 11 February 2003, the World Health Organization (WHO) received the first official report of an outbreak of severe atypical pneumonia of unknown aetiology in the province affecting 305 people, 30% of whom were health-care workers (HCW).2 The international spread of SARS in 2003 was associated with air travel and characterized by ‘superspreading’ events, which initiated outbreaks in a number of countries. A 64-year-old physician, who had treated SARS patients and later died of SARS, carried the virus out of Guangdong on 21 February 2003. He stayed on the ninth floor of a hotel in Kowloon, Hong Kong Special Administrative Region (SAR) of China on 21–22 February while symptomatic and was the source case for the global spread of SARS3 and for the outbreak in Hong Kong SAR.4 At least 16 symptomatic secondary cases occurred among guests at the hotel who stayed on the same floor, some of whom seeded SARS outbreaks in the hospital systems of Hanoi City, Vietnam, Singapore and
100
the Greater Toronto Area (GTA), Canada5 upon their return to those areas. SARS transmission was amplified in acute care settings in these sites before the risk of SARS was known and transmission-based infection control precautions implemented. The first SARS case in Vietnam (a travelling businessman who stayed at the Kowloon hotel concurrently with the Guangdong physician) was admitted to the Hanoi French Hospital on 26 February.6 Approximately 20 hospital staff became ill with an acute febrile respiratory illness after his admission. Dr Carlo Urbani from the WHO country office in Hanoi was the first person to recognize the severity of the public health threat that proved to be SARS and to notify the outbreak within the Organization. In Hong Kong, the brother-in-law of the Guangdong physician and a ChineseCanadian business man who also stayed at the Kowloon hotel were infected and became the source cases for chains of transmission ultimately affecting four hospitals by 17 March.4 In response to the reports from Vietnam and Hong Kong SAR, the WHO issued a global alert describing the syndrome on 12 March.7 This alert was followed on 14 March by reports from Singapore, Canada and Taiwan, China, of clusters of cases with a similar clinical presentation. The index case of SARS in Singapore8 was a previously healthy 23-year-old woman who had stayed on the ninth floor of the
Epidemiology and Transmission of SARS
Kowloon hotel between 20 and 25 February and was admitted to the Tan Tock Seng Hospital, Singapore on 1 March on day 5 of her illness.9 More than 100 cases over several generations of disease transmission were subsequently linked to this patient. The index case of the Toronto outbreak, a 78-year-old woman with a history of type 2 diabetes and coronary heart disease, returned to Canada on 23 February following a 3-night stay at the Kowloon hotel (18 to 21 February). She became symptomatic on 25 February and died at home on 5 March. She infected several members of her family and her attending physician. Her son, who had a history of type 2 diabetes and hypertension and who became symptomatic on 27 February died in a Toronto hospital on 13 March and was the source case for nosocomial spread of SARS CoV in the GTA. The first recognized SARS patient in Taiwan was a 54-year-old businessman who travelled to Guangdong on 5 February, and returned to Taipei by way of Hong Kong SAR on 21 February.10 He became symptomatic on 25 February and was hospitalized 12 days later. These events provided mounting evidence that SARS was spreading globally through international air travel and on 15 March WHO issued the first of two travel advisory notices.
The origins of SARS There is evidence that natural infection with SARS CoV may occur in a number of animal species indigenous to China and parts of south-east Asia. The WHO epidemiological investigation in Guangdong in April 2003 determined that throughout the epidemic a high proportion of community cases had no reported contact history; 42.8% of SARS cases without a history of exposure to known cases worked in wildlife markets or kitchens in several of the affected cities of Guangdong.11 It is postulated that food and wildlife handlers were
101
occupationally exposed to SARS CoV-like viruses from an animal host that became human adapted.12 A Chinese government team has also released results showing that 66 of 508 (13%) animal handlers tested at markets in Guangdong had antibodies against SARS CoV.13,14 Serological studies of animal and vegetable traders within the Guangdong market showed that 40% (8/20) of the wild animal traders, 20% (3/15) of the wild animal butchers and 5% (1/20) of the vegetable traders were seropositive for SARS CoV. None of those tested had reported SARS-like symptoms in the preceding 6 months. On 23 May 2003 research teams in Hong Kong SAR and Shenzhen, China announced the results of a joint study of wild animals for human consumption taken from a market in southern China.13 The study detected several coronaviruses genetically similar to SARS CoV in two of the animal species tested, the Himalayan masked palm civet (Paguma larvata) and raccoon dog (Nyctereutes procyonoides). The study also found antibodies against the SARS CoV in one Chinese ferret badger (Melogale moschata). These and other wild animals are considered delicacies throughout southern China. Two fully sequenced civet coronaviruses showed an additional 29 base pair sequence compared to human SARS CoV. A number of studies are under way in China to determine the prevalence of SARS CoV infection in animals, and the host range. A number of additional species have tested positive by polymerase chain reaction (PCR) or serology — cynomolgus macaques, fruit bats, snakes and wild pigs — however, these findings may be the result of test cross-reactivity and require further validation.13,14 In addition, domestic cats (Felis domesticus) living in the Amoy Gardens apartment block in Hong Kong SAR were also found to be infected with SARS CoV.15 More recently, ferrets (Mustela furo) and domestic cats were infected with SARS CoV experimentally and found to transmit the virus efficiently to previously uninfected
102
Severe Acute Respiratory Syndrome
animals housed with them.16 These findings indicate that the reservoir for this pathogen may involve a range of animal species. The eradication of SARS is unlikely if the virus is zoonotic. Priority areas for action include establishing the origins of SARS CoV, the animal host range and viral ecology, factors leading to emergence of the virus (changes in the agent, host factors, farming and slaughtering practices and wildlife utilization) and models for the dynamics of infection. There is considerable speculation about the ongoing risk of interspecies transmission of SARS CoV, from animals to humans and to other animals. Behavioural and environmental risk factors for interspecies transmission require further research, including linked epidemiological studies in humans and animals at risk of infection and the molecular epidemiology of SARS CoV-like viruses in both populations.
The global epidemic The epidemic curve of the 2002–03 epidemic, excluding unit data from the People’s Republic of China, is presented in Fig.13.1 by week of onset. Following closure of the global database for the 2002–03 epidemic of SARS, WHO reported 8096 cases of probable SARS from 26 countries;17,18 53% were female and 21% HCW. The largest outbreak occurred in the People’s Republic of China with 5327 cases and 349 deaths. The global case fatality ratio was 9.6% (774 deaths directly attributed to SARS) but varied from 0% to more than 50% depending on the age group affected and location. On 5 July 2003, 4 months after the first cases were reported outside mainland China, WHO announced that the last chain of human transmission had been interrupted. Most cases occurred in HCW and their close contacts; however, secondary community transmission did occur, especially
Probable cases of SARS by week of onset Worldwide* (n = 5910), 1 November 2002 – 10 July 2003 Number of cases 160
WHO issues first travel advisory 15 March
140
WHO issues global alert 12 March 120
100
80
60
40
20
0 01-Nov-02 22-Nov-02 13-Dec-02 03-Jan-03 24-Jan-03 14-Feb-03 07-Mar-03 28-Mar-03 18-Apr-03 09-May-03 30-May-03 20-Jun-03 11-Jul-03 Date of onset
Figure 13.1 Progress of 2002–3 SARS epidemic. Adapted from World Health Organization. Epidemic curves — Severe acute respiratory syndrome (SARS). *This graph does not include 2527 probable cases of SARS (2521 from Beijing, China), for whom no dates of onset are currently available.
Epidemiology and Transmission of SARS
in more mature outbreaks and in some circumscribed environments such as the outbreak in the Amoy Gardens housing estate in Hong Kong19 and during air travel20 where transmission is partially attributed to aerosols. Sustained local transmission occurred in Toronto, Canada (incidence rate of 5.1 per 100 000), Beijing city, China (19.6), Hong Kong SAR (25.6), Taipei, Taiwan, China (23.1), Singapore (5.9) and Hanoi City, Vietnam (2.3).21 All age groups were affected (age range 0–100 years, median age 42 years). Multivariate analysis of risk factors associated with SARS-related mortality from Hong Kong SAR found that age over 60 years,22,23 elevated lactate dehydrogenase at presentation,22,23 the presence of diabetes mellitus or heart disease,24 and the presence of other comorbid conditions24 were independent predictors of SARS mortality. A WHO global multivariate analysis of SARS mortality in 1870 cases from Hong Kong SAR, Singapore and Vietnam by affected area, age group, sex and occupation (HCW or other) found that increasing age and male sex were independently associated with higher mortality while being a HCW was protective after adjustment (Anker M, unpublished data). There was no statistically significant difference in the mortality by geographical area. Since July 2003, there have been four occasions when SARS has reappeared, resulting in 17 cases. Three of these incidents were attributed to breaches in laboratory biosafety and resulted in one or more cases (Singapore,25 Taiwan, China26 and Beijing, China27,28). Fortunately only one of these incidents resulted in secondary transmission outside of the laboratory. The most recent incident was a cluster of 11 cases, one of whom died, in three generations of transmission affecting family and hospital contacts of a laboratory worker. For this reason, WHO strongly urges countries to undertake an inventory of all laboratories working with cultures of live SARS CoV or storing infectious clinical specimens. All countries
103
should ensure adherence to biosafety levels and safe procedures29 and that appropriate monitoring and investigation of illness in laboratory workers is undertaken. The fourth incident (Guangzhou, Guangdong province, China30,31) resulted in four cases arising over a 6-week period. Three of the cases were attributed to exposure to animal1 or environmental sources while the source of exposure is unknown in the other case. There was no further community transmission.
Incubation period The estimates for the incubation period for SARS converge at 2–10 days (range 1–14 days). Estimates are derived from an analysis of SARS cases with single point exposures or exposure over a well-defined interval. Most countries reported a median incubation period of 4–5 days, and a mean of 4–6 days.21 Cases with an incubation period outside the 2–10-day interval have not necessarily been subjected to rigorous and standardized investigation, including serological confirmation. Donnelly et al.32 analysed 1425 cases notified to 28 April in the Hong Kong SAR for whom epidemiological, demographic and clinical data were linked. It remains unclear whether the route of transmission influences the incubation period and the infective dose in humans is not known.
Infectious period Transmission efficiency appears to be greatest from severely ill patients and those experiencing rapid clinical deterioration, usually during the second week of illness. However, there have been reports of transmission of SARS CoV in the first week of illness.5 Further elucidation of the risk of transmission from laboratory-confirmed cases with mild illness, and transmission during the prodromal period is needed. There has been no observed transmission before the onset of symptoms.33 Data
104
Severe Acute Respiratory Syndrome
from Singapore show that few secondary cases occur when symptomatic patients are isolated within 3–4 days of illness onset. 34 This inference of infectivity by time since onset derived from epidemiological observations correlates very closely with laboratory and clinical data on cases. Maximum virus excretion from the respiratory tract occurs on about day 10 of illness and then declines. Peiris et al.35 examined sequential nasopharyngeal aspirates/throat and nose swabs, and stools from patients with SARS using quantitative reverse transcriptase (RT)-PCR. Virus shedding in stool begins later than in respiratory secretions but in both peak viral excretion occurred in the second week of illness followed by a decline in the proportion of specimens testing positive. Clinical deterioration, when it occurs, also coincides with peak infectivity. Data linkage is required to determine whether there is a direct relationship between clinical severity and viral load and excretion.36 There are no reports of transmission beyond 10 days of fever resolution consistent with the total period of isolation following defervescence recommended by WHO. Based on the evidence now available, the WHO guidelines on hospital discharge and convalescence remain valid.37 Few serial clinical specimens have been collected and some centres experienced difficulty in linking clinical, laboratory and epidemiological data to build up a complete picture of the interaction between the SARS CoV, its human host and transmission environments. Viral shedding studies linked to the clinical progression of disease should be conducted on all new cases of SARS. Additional epidemiological and laboratory studies are needed to describe fully the period of communicability, including quantitative virology.
SARS transmission The primary mode of transmission appears to be direct mucous membrane (eyes, nose, and mouth) contact with infectious respira-
tory droplets and/or through exposure to fomites. Cases have occurred primarily in persons with close contact38 with those very ill with SARS, especially in health-care settings prior to the implementation of hospital-wide infection control precautions39 and household settings. Patients with atypical presentations have been important in prolonging nosocomial outbreaks or fuelling new ones.40 In Hong Kong SAR,41 secondary household attack rates were calculated for 1214 SARS case-patients and their household members, stratified by two phases of the epidemic. Secondary infection occurred in 14.9% (22.1% versus 11% in earlier and later phases) of all households and 8% (11.7% versus 5.9% in the earlier and later phases) of all household members. Duration before hospitalization, visiting the SARS patient in hospital (and mask use during the visit) and frequency of close contact were independent predictors of secondary transmission in a multivariate analysis. In Singapore, a secondary household transmission study of 114 households involving 417 contacts reported an attack rate of 6.2%,42 similar to that reported in Beijing (4.6%).43 Transmission to social and casual contacts has occasionally occurred as a result of short but intense exposure to a case of SARS (in hospitals,38,44 very rarely in other workplaces,8,45 aeroplanes,46 taxis8 and a train47). Community-acquired infection has been associated with a religious gathering.48 The basic reproduction number, R0 (the average number of secondary infectious cases produced by an infectious case), was 2–4 during the 2003 outbreak. This is consistent with a disease spread by direct contact or larger virus-laden droplets that travel only a few metres rather than by lighter airborne particles.21,32 By contrast, in diseases transmitted by aerosols, such as measles and influenza, a single person can infect an entire room by coughing. Aerosolizing procedures in hospitals (intubation,49 nebulization,50 suction and ventilation), and other events51 that promote aerosolization of infectious respira-
Epidemiology and Transmission of SARS
tory droplets or other potentially infectious materials (such as faeces or urine) in hospitals or other settings,50 may amplify transmission. Survival of SARS CoV51a needs further investigation in various settings and under a variety of conditions (e.g. in fomites and difficult-to-clean surfaces including carpets) as do methods for cleaning surfaces without generating dangerous aerosols. Appropriate respiratory precautions should be sustainable in a fully functioning hospital and there is a need to establish a ‘new norm’ in respiratory precautions.37,52 Health-care facilities are complex systems allowing many opportunities for disease transmission through the interaction between an infective agent, individuals and their environment. To prevent or interrupt SARS CoV transmission all health facilities should ensure they are applying standard precautions at all times, with the adoption of additional transmission-based precautions for the investigation and management of individuals with an acute respiratory illness based on a risk assessment of the population risk and individual risk of SARS at the local level. Laboratory biosafety has emerged as an important risk factor for SARS transmission in the post-outbreak period.29 On 10 September 2004, a new case of SARS was detected in Singapore.25 The patient was a 27-year-old post-doctoral student working on West Nile virus at a microbiology laboratory which also worked with live SARS CoV. His infection was attributed to crosscontamination. The second case was that of an experienced researcher working with SARS CoV in Taiwan, China, following a spill which he attempted to decontaminate without the use of appropriate personal protective equipment.26 The third laboratory transmission incident occured in Beijing, China.27,28 The outbreak was attributed to the use of inadequately inactivated SARS CoV in the context of poor biosafety systems and practices.52a WHO has issued guidance on laboratory safety specific to SARS CoV.53 The role of faecal–oral transmission is un-
105
known and several animal coronaviruses are spread via the faecal–oral route;54 however, there is no evidence currently that this mode of transmission plays a key role in SARS. In Hong Kong SAR, Peiris et al. reported watery diarrhoea in 55 (73%) of 75 cases from the Amoy Gardens outbreak35 and Leung et al. reported that 20.3% of 138 patients had watery diarrhoea on admission.55 The onset of diarrhoea occurred at a mean 7.5 days of illness with a maximum frequency of 6.3 stools per day. Diarrhoea was less frequent in other series; 16% of 1315 cases on the Hong Kong SAR Hospital Authority database.21 In Vietnam, approximately 50%56 of cases had diarrhoea during their illness (6.5% with diarrhoea at admission57) with the most severe cases all having diarrhoea. In Guangzhou City, Guangdong, 8.6% of 662 probable and suspect cases of SARS had diarrhoea at onset; diarrhoea at any time during the course of illness was not documented. In Taiwan, approximately 57% of cases had diarrhoea at any time as did 28% of probable cases and 19% of suspect cases in Ontario, Canada. Although late diarrhoea may be related to antibiotic treatment rather than part of the natural history of the disease, diarrhoea may remain important for infectivity given that viral excretion was greatest in stool. Under certain circumstances, such as in health-care settings or other closed environments, contamination of inanimate materials or objects by infectious respiratory secretions or other body fluids seems to play a role occasionally in disease transmission. Infection control measures that have reduced contact, droplet and aerosol spread of SARS CoV have been shown to reduce greatly the risk of transmission.58 Despite considerable opportunity there have been no reports of food or waterborne transmission; however studies are needed to define further the potential role of these routes. Children are at lower risk of SARS than are adults; most children experience a milder illness than adults do and there is
106
Severe Acute Respiratory Syndrome
little evidence of transmission from children.59,60,60a These findings are in contrast to the secondary attack rates among adults.61 In addition, there have been no reported cases of vertical transmission.62 However, SARS CoV infection has been associated with both maternal and fetal morbidity and mortality.63
Superspreading events The spread of SARS has been characterized by a number of ‘superspreading events’ defined as a transmission event that generates many more than the average number of secondary cases. Superspreading events have been reported from all sites with sustained local transmission of SARS, especially before the implementation of strict hospital infection control.12,64–66 It is uncertain whether these ‘superspreading events’ are due to special conditions conducive to virus transmission (superspreading environments), to some characteristic of the source case such as high viral load or the capacity to excrete large amount of virus or to a characteristic of the virus making it more transmissible. Detailed investigations of ‘superspreading events’ are needed to elucidate further the relative importance of environment, host and vector, given that most SARS cases generate fewer than three secondary cases. In Singapore, the number of secondary cases decreased with each generation but may in part be the result of earlier case detection and isolation.
Controlling SARS SARS was controlled by traditional hospital infection containment67 and public health measures68 — early case detection and isolation, strict infection control and the use of personal protective equipment, contact tracing and quarantine and community mobilization. There is an urgent need to evaluate quantitatively the effectiveness of population-based prevention and control measures.69 It is also essential that early and
specific laboratory tests be developed so that measures to prevent onward transmission may be rapidly introduced. In the inter-epidemic period where the risk of false positive results is high, positive results reported by national reference laboratories should be rapidly verified at a WHO SARS Reference and Verification Network laboratory.29 The goal of WHO is to eliminate SARS as a public health risk. Breaches in laboratory biosafety and the re-emergence of SARS from interspecies transmission remain the two most likely sources of new infections. The WHO SARS Research Advisory Committee70 which first convened in October 2003 has developed a list of public health research priorities, including further elucidation of the epidemiology, clinical, diagnostic, social and economic aspects of SARS CoV infection, research into the effectiveness of public health interventions implemented during the 2003 epidemic, the zoonotic origin of the virus and its ability to jump between species. WHO has also established committees to assist in vaccine development and has maintained its clinical, epidemiological, laboratory, and animal networks. These WHO activities, together with similar activities in partner institutions of the Global Outbreak Alert and Response Network, will greatly assist in preventing further epidemic activity should SARS CoV re-emerge from its zoonotic reservoir.
Acknowledgements This chapter acknowledges the significant contribution made by the WHO SARS Epidemiology Working Group and the WHO SARS Epidemiology Network to the global epidemiology of SARS: • Andrea Ammon, Udo Buchholz, Gérard Krause, Klaus Stark (Robert Koch-Institut, Germany); • Larry Anderson, Ray R Arthur, Scott F Dowell, Jairam R Lingappa, Sara Lowther, Sonja Olsen, Umesh Parashar, Lauren Stockman (Centers for Disease Control
Epidemiology and Transmission of SARS
•
•
•
• •
• • •
•
•
• •
• •
• •
and Prevention, Department of Health and Human Services, USA); Roy M Anderson (Department of Infectious Disease Epidemiology, Imperial College, UK); Dounia Bitar, Isabelle Bonmarin, Daniel Lévy-Bruhl, Bénédicte Decludt, JeanClaude Desenclos (Institut National de Veille Sanitaire, France); Gabrielle Breugelmans, Alain Moren, Doris Radun, (European Programme for Intervention Epidemiology Training, European Union); Philippe Cavailler (Epicentre, France; Médecins Sans Frontières, Switzerland); Margaret Chan, Wilina Lim, Marina Sum, Thomas Tsang, (Department of Health, Hong Kong Special Administrative Region of China); Suok Kai Chew, Lynn James, Stefan Ma (Ministry of Health, Singapore); Stephen Corber, Marlo Libel (WHO Regional Office for the Americas); Shelley Deeks, Arlene King, Jeanette Macey, Susan Squires, Theresa Tam, Tom Wong, Ping Yan (Population and Public Health Branch, Health Canada, Canada); Valerie Delpech, Nigel Gay, Angus Nicoll, John Watson, (Health Protection Agency, UK); Yon Fleerackers, Jeffrey Gilbert, Julie Lyn Hall, Chin-kei Lee (WHO Representative Office in the People’s Republic of China, People’s Republic of China); Bernardus Ganter, Richard Pebody (WHO Regional Office for Europe, Denmark); Yu Hongjie, Yang Weizhong (China Center for Disease Control and Prevention, People’s Republic of China); Peter Horby (WHO Representative Office in Vietnam, Vietnam); Jirapat Kanlayanaphotporn, Kumnuan Ungchusak, (Ministry of Public Health, Thailand); Stephen Lambert (WHO Regional Office for the Western Pacific, Philippines); Mark Lipsitch (Department of Epidemiology, Harvard School of Public Health, USA);
107
• John Mackenzie, Aileen Plant (Australian Biosecurity Cooperative Research Centre for Emerging Infectious Disease, Australia; Centre for International Health, Curtin University of Technology, Australia); • Babatunde Olowokure (Department of Communicable Disease Surveillance and Response, WHO Geneva, Switzerland); • Leslee Roberts, Jenean Spencer (Department of Health and Ageing, Australia); • Mick G Roberts (Institute of Information and Mathematical Sciences, Massey University, New Zealand); • Xu Shen (Ministry of Health, People’s Republic of China); • Chin-Shui Shih, Ih-Jen Su (Center for Disease Control, Taiwan, China); • Sean Tobin (WHO Representative Office in Cambodia, Cambodia); • Jacco Wallinga, (Centre for Infectious Disease Epidemiology, National Institute of Public Health and the Environment, Netherlands).
References 1
2
3
4
5
6
7
8
Guan Y, Zheng BJ, He YQ et al. Isolation and characterization of viruses related to the SARS coronavirus from animals in southern China.Sciencexpress/www.sciencexpress. org/4 September 2003/Page 1/10.1126/ science. 1087139 World Health Organization. Update 95 - SARS: Chronology of a serial killer. http://www.who. int/csr/don/2003_07_04/en/ Centers for Disease Control and Prevention. Update: Outbreak of Severe Acute Respiratory Syndrome — Worldwide, 2003;52: 241–8. Tsang KW, Ho PL, Ooi GC et al. A cluster of cases of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348: 1977–85. Poutanen SM, Low D, Henry B et al. Identification of severe acute respiratory syndrome in Canada. N Engl J Med 2003;348: 1995–2005. World Health Organization. Update — Severe Acute Respiratory Syndrome (SARS), 16 March 2003. World Health Organization. WHO issues a global alert about cases of atypical pneumonia, 12 March 2003. Centers for Disease Control and Prevention.
108
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Severe Acute Respiratory Syndrome
Severe Acute Respiratory Syndrome — Singapore, 2003. MMWR 2003;52: 405–11. Hsu L-Y, Lee C-C, Green JA et al. Severe acute respiratory syndrome (SARS) in Singapore: clinical features of index patient and initial contacts. Emerg Infect Dis 2003;9: 713–17. Twu S-J, Chen T-J, Chen C-J et al. Control measures for severe acute respiratory syndrome (SARS) in Taiwan. Emerg Infect Dis 2003;9: 718–20. Breiman RF, Evans MR, Preiser W et al. Role of China in the quest to define and control severe acute respiratory syndrome. Emerg Infect Dis 2003;9: 1037–41. Chinese Center for Disease Control and Prevention. Overview of epidemics and responses to the severe acute respiratory syndrome (SARS) in the People’s Republic of China. 16 June, 2003. Cyranoski D, Abbott A. Virus detectives seek source of SARS in China’s wild animals. Nature 2003;423: 467; doi:10.1038/423467a News24. SARS found in pigs and snakes. 4 June 2003 http://www.news24.com/News24/ World/News/0,,2–10–1488_1368983,00.ht ml Ng SKC. Possible role of an animal vector in the SARS outbreak at Amoy Gardens. Lancet 2003;362: 570–2. Martina BE, Haagmans BL, Kuiken T et al. Virology: SARS virus infection of cats and ferrets. Nature 2003;425: 915. World Health Organization. Global surveillance for severe acute respiratory syndrome (SARS). Wkly Epidemiol Rec 2003; 78: 100–110. World Health Organization. Summary of probable SARS cases with onset of illness from 1 November 2002 to 31 July 2003 (Based on data as of the 31 December 2003). http://www. who.int/csr/sars/country/ table2004_04_21/ en/ Yu IT, Li Y, Wong TW et al. Evidence of airborne transmission of the severe acute respiratory syndrome virus. N Engl J Med 2004;350: 1731–9. Olsen SJ, Chang HL, Cheung TY et al. Transmission of the severe acute respiratory syndrome on aircraft. N Engl J Med 2003;349: 2416–22. World Health Organization. Consensus document on the epidemiology of severe acute respiratory syndrome (SARS). WHO/CDS/CSR/ GAR/2003.11, WHO 2003. Choi KW, Chau TN, Tsang O et al. Outcomes and prognostic factors in 267 patients with severe acute respiratory syndrome in Hong Kong. Ann Intern Med 2003;139: 715–23. Wong RS, Wu A, To KF et al. Haematological manifestations in patients with severe acute
24
25
26
27
28
29
30
31
32
33
34
35
36
37
respiratory syndrome: retrospective analysis. BMJ 2003;326: 1358–62. Chan JW, Ng CK, Chan YH et al. Short term outcome and risk factors for adverse clinical outcomes in adults with severe acute respiratory syndrome (SARS). Thorax 2003;58: 686–9. World Health Organization. Severe acute respiratory syndrome (SARS) in Singapore, 10 September 2003. http://www.who.int/csr/ don/2003_09_10/en/ World Health Organization. Severe Acute Respiratory Syndrome (SARS) in Taiwan, China,17 December 2003. http:// www.who.int/csr/ don/2003_12_17/en/ World Health Organization. SARS: one suspected case reported in China, 22 April 2004. http://www.who.int/csr/don/2004_04_22/ en/ World Health Organization. Update 7 — China’s latest SARS outbreak has been contained, but biosafety concerns remain, 18 May 2004. http://www.who.int/csr/don/2004_ 05_18a/en/ World Health Organization. Summary of the discussion and recommendations of the SARS Laboratory Workshop, 22 October 2003. http://www.who.int/csr/sars/guidelines/ en/SARSLabmeeting.pdf World Health Organization. Suspected severe acute respiratory syndrome (SARS) case in southern China, 28 December 2003. http://www. who.int/csr/don/2003_12_28/en/ World Health Organization. Update 5 — New case of laboratory-confirmed SARS in Guangdong, China, 31 January 2004. http:// www.who.int/csr/don/2004_01_31/en/ Donnelly CA, Ghani AC, Leung GM et al. Epidemiological determinants of spread of causal agent of severe acute respiratory syndrome in Hong Kong. Lancet 2003,361: 1761–6. Vu TH, Cabau JF, Nguyen NT et al. SARS in Northern Vietnam. N Engl J Med 2003;348: 2035. Lipsitch M, Cohen T, Cooper B et al. Transmission dynamics and control of severe acute respiratory syndrome. Science 2003; 300: 1966–70. Peiris JSM, Chu CM, Cheng VC et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;361: 1761–6. Lee N, Hui DS, Wu A et al. A major outbreak of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348: 1986–94. World Health Organization. WHO hospital discharge and follow-up policy for patients who have been diagnosed with severe acute
Epidemiology and Transmission of SARS
38
39
40
41
42
43
44
45
46
47
48
49
50
respiratory syndrome (SARS). Revised 28 March 2003. http://www.who.int/csr/sars/ discharge/en/ Close contact is defined as having cared for, lived with or having had direct contact (<1 metre) with respiratory secretions or other body fluids of a suspect or probable case of SARS. Varia M, Wilson S, Sarwal S et al. for the Hospital Outbreak Investigation Team. Investigation of a nosocomial outbreak of severe acute respiratory syndrome (SARS) in Toronto, Canada. Can Med Assoc J 2003; 169: 285–92. Tomlinson B, Cockram C. SARS: experience at Prince of Wales Hospital, Hong Kong. Lancet 2003;361: 1486–87. Lau JT, Lau M, Kim JH et al. Probable secondary infections in households of SARS patients in Hong Kong. Emerg Infect Dis 2004;10: 235–43. Goh DL-M, Lee BW, Chia KS et al. Secondary household transmission of SARS, Singapore. Emerg Infect Dis [serial online] 2004 Feb [date cited]. http://www.cdc.gov/ncidod/ EID/vol10no2/03–0676.htm Centers for Disease Control and Prevention. Efficiency of quarantine during an epidemic of severe acute respiratory syndrome — Beijing, China, 2003. MMWR 2003;52: 1037. World Health Organization. SARS outbreak in the Philippines. Wkly Epidemiol Rec 2003;78: 189–92. Health Canada. Learning from SARS. Renewal of public health in Canada. A report of the National Advisory Committee on SARS and Public Health, October 2003. http://www.hc-sc.gc.ca/english/pdf/sars/ sars-e.pdf Olsen SJ, Chang HL, Cheung TYY et al. Transmission of severe acute respiratory syndrome on aircraft. N Engl J Med 2003;349: 2414–20. Chiu RWK, Chim SSC, Lo YMD. Molecular epidemiology of SARS — from Amoy Gardens to Taiwan. N Engl J Med 2003;349: 1875–6. Health Canada Summary of Severe Acute Respiratory Syndrome (SARS) Cases: Canada and International, 16 April, 2003. http:// www.hc-sc.gc.ca/pphb-dgspsp/sars-sras/ eu-ae/sars 20030416_e.html Christian MD, Loutfy M, McDonald LC et al. Possible SARS coronavirus transmission during cardiopulmonary resuscitation. Emerg Infect Dis [serial online] 2004 Feb [date cited]. Available from: URL: http:// w w w. c d c . g o v / n c i d o d / E I D / v o l 1 0 n o 2 / 03–0700.htm Wong T-W, Lee C-K, Tam W et al. Cluster of SARS among medical students exposed
109
to single patient, Hong Kong. Emerg Infect Dis [serial online] 2004 Feb [date cited]. http://www.cdc.gov/ncidod/EID/ vol10no2/ 03–0452.htm 51 Ministry of Health, Welfare and Food, Hong Kong SAR. Transcript of SHWF on the findings of an investigation of severe acute respiratory syndrome outbreak at Amoy Gardens (Parts 1 and 2), 17 April 2003. http://www.info. gov.hk/gia/general/200304/17/0417290. htm and http://www.info.gov.hk/gia/ general/200304/17/ 0417308.htm 51a Rabenau HF, Cinati J, Morgenstern B et al. Stability and inactivation of SARS coronavirus. Med Microbiol Immunol (Berl) 2004 Apr 29 (Epub ahead of print). 52 Schabas R. SARS: prudence, not panic. Can Med Assoc J 2003;168: 1432–3. Published online www. cmaj.ca 23 April, 2003. 52a World Health Organization, Western Pacific Region. Investigation into China’s recent SARS outbreak yields important lessons for global public health, 2 July 2004. http://www.wpro.who.int/sars/docs/ update/update_07022004.asp 53 World Health Organization. WHO postoutbreak biosafety guidelines for handling of SARS-CoV specimens and cultures, 18 December 2003. http://www.who.int/csr/sars/ biosafety2003_12_18/en/ 54 McIntosh K. Coronaviruses: a comparative review. Curr Top Microbiol Immunol 1974;63: 85–129. 55 Leung WK, To KF, Chan PKS et al. Enteric involvement of severe acute respiratory syndrome-associated coronavirus infection. Gastroenterology 2003;125: 1011–17. 56 Minutes of the 7 May World Health Organization Ad Hoc Working Group on the Epidemiology of SARS. 57 Vu HT, Leitmeyer KC, Le DH et al. Clinical description of a completed outbreak of severe acute respiratory syndrome (SARS) in Vietnam, February–May, 2003. Emerg Infect Dis [serial online] 2004 Feb [date cited]. http://www.cdc.gov/ncidod/EID/vol10no2/ 03–0761.htm 58 Seto WH, Tsang D, Yung RW et al. Advisors of Expert SARS group of Hospital Authority. Effectiveness of precautions against droplets and contact in prevention of nosocomial transmission of severe acute respiratory syndrome (SARS). Lancet 2003;361: 1519–20. 59 Hon KLE, Leung CW, Cheng WTF et al. Clinical presentations and outcome of severe acute respiratory syndrome in children. Lancet. Published online 29 April, 2003. http://image.thelancet.com/extras/ 03let4127web.pdf
110
60
Severe Acute Respiratory Syndrome
Wang M, Du L, Zhou D-H et al. Study on the epidemiology and measures for control of severe acute respiratory syndrome in Guangzhou City (English Abstract). In: Collection of papers on SARS published in CMA Journals. Beijing: Chinese Medical Association, 2003: 50. 60a Chan WM, Kwan YW, Wan HS et al. Epidemiologic linkage and public health implication of a cluster of Severe Acute Respiratory Syndrome in an extended family. Pediatr Infect Dis J 2004 Dec;23(12): 1156–9. 61 Peiris JSM, Lai ST, Poon LLM et al. and members of the SARS study group. Coronavirus as a possible cause of severe acute respiratory syndrome. Lancet 2003;361: 1319–25. 62 Ho LC. SARS and pregnancy. Experience in Hong Kong March–May 2003. SARS Clinical management Workshop, 13–14 June 2003, China, Hong Kong SAR. 63 Wong SF, Chow KM, de Swiet M. Severe acute respiratory syndrome and pregnancy. BJOG 2003;110: 641–2. 64 Wong T-W, Lee C-K, Tam W et al. Cluster of SARS among medical students exposed to single patient, Hong Kong. Emerg Infect Dis [serial online] 2004 Feb [date cited]. http://www.cdc.gov/ncidod/EID/vol10no2/ 03–0452.htm
65
66
67
68
69
70
Shen Z, Ning F, Zhou W et al. Superspreading SARS events, Beijing, 2003. Emerg Infect Dis [serial online] 2004 Feb [date cited]. http:// www.cdc.gov/ncidod/EID/vol10no2/03–07 32.htm World Health Organization. Update 83 — One hundred days into the outbreak, June 18. http:// www.who.int/csr/don/2003_06–18/en/ Lloyd-Smith JO, Galvani AP, Getz WM. Curtailing transmission of severe acute respiratory syndrome within a community and its hospital. Proc R Soc Lond B Biol Sci 2003; 270:1979–89. World Health Organization Communicable Diseases. Surveillance and Response. Severe acute respiratory syndrome (SARS): Status of the outbreak and lessons for the immediate future, Geneva, 20 May 2003. http://www.who.int/ csr/media/sars_wha.pdf Chau PH, Yip PSF. Monitoring the severe acute respiratory syndrome epidemic and assessing effectiveness of interventions in Hong Kong Special Administrative Region. J Epidemiol Community Health 2003;57: 766–9. World Health Organization. WHO SARS Scientific Research Advisory Committee concludes its first meeting, 22 October 2003. http://www.who.int/csr/sars/archive/ research/en/
Chapter 14
Transmission Dynamics and Control of the Viral Aetiological Agent of SARS Gabriel M Leung, Anthony J Hedley, Tai Hing Lam, Azra C Ghani, Christl A Donnelly, Christophe Fraser, Steven Riley, Neil M Ferguson and Roy M Anderson
Introduction In the global response to SARS, orchestrated by the World Health Organization (WHO), there were five priority tasks at the start of the outbreak. These were the identification of the aetiological agent of the new disease SARS, the development of diagnostic tests to detect the virus, the development and assessment of treatment protocols to reduce morbidity and mortality, determination of the key epidemiological parameters that affected spread and persistence and the formulation and implementation of appropriate public health interventions. Most of these tasks were completed rapidly, including the identification of the viral aetiological agent (Chapter 7), the delineation of the full genome sequence of the virus (Chapter 8), the evaluation of key epidemiological parameters1 and the impact of different interventions.2,3 We have also recently developed and validated a clinical prediction rule for SARS triage and diagnosis at the point of first contact.4 With respect to clinical interventions to reduce morbidity and mortality, too little has been done to date to merge clinical patient databases from different settings and countries to provide sufficient power to perform rigorous evaluations taking due account of the many confounding factors such as age and various comorbidities. In any such analyses due note must be taken of differences between countries in the way clinical symptoms were recorded.
Despite these problems, simple public health interventions such as isolation of sick patients suspected of having SARS, contact tracing and the concomitant quarantining of contacts, restrictions on travel and behavioural change within communities (either self-imposed or encouraged by WHO, government departments or local health authorities), brought the global epidemic under control by midsummer of 2003. Post its recognition as a global threat in mid-March 2003, SARS was successfully contained in less than 4 months. On 5 July 2003, WHO reported that the last human chain of transmission of SARS had been broken. During the 4-month period of the major epidemic 8098 cases were reported from 29 countries. Most countries had no or a few cases while a few had in excess of 1000 cases (Fig. 14.1). The epidemic was largely in China and surrounding countries or administrative regions such as Taiwan, Singapore and Hong Kong. Although the total number of cases is small by comparison with other major respiratory tract infections, the case fatality rate was high1 and the economic impact in the Asian region and in Canada was very high (http://www.dfait-maeci.gc.ca/mexico city/economic/may/sarsbriefMay03.pdf; http://www.intlrisk.com/pdf/reports/may 2003-sars.pdf). The resurgence of SARS remains a distinct possibility, given its uncertain origins and the possible existence of an animal reservoir. In the coming year, all countries have
111
112
Severe Acute Respiratory Syndrome
N u m b e r o f c o u n trie s
250 200 150 100 50 0 0
1-10
11-100
1011000
10015000
Number of reported SARS cases
been requested by WHO to remain vigilant for the recurrence of SARS and maintain their capacity to detect and respond to the re-emergence of the disease should it occur (Chapter 23). This chapter reviews our current understanding of the key epidemiological determinants of the transmission dynamics of SARS CoV, and attempts to evaluate what interventions worked best in different settings, using mathematical models as a template for analysis. We also consider gaps in current knowledge and priorities for future research to improve predictions of observed pattern and intervention impact.
Data needs for the study of transmission dynamics The study of the transmission dynamics of an infectious agent is typically based on simple or complex mathematical frameworks that may be either deterministic or stochastic in form.5 The goals in model formulation and analysis can be many and varied. They include delineating what needs to be measured for a better understanding of an observed pattern, what are the key determinants of this pattern, and how might different interventions introduced at various times post the emergence of an epidemic influence the future incidence of infection
>5000
Figure 14.1 The frequency distribution of reported SARS cases per country as recorded by WHO. The two special administrative regions of China and Taiwan are recorded separately from mainland China.
and associated disease? For a new pathogen the first of these goals is of central importance in guiding data collection and analysis as well as the formulation of policies to protect public health. The following sections describe the key steps in data collection and analysis, with the aim of providing robust estimates of the most important parameters (and their distributions) that influence transmission and control. Much variability is often associated with the parameters that influence the typical course of infection in a patient and transmission between patients. The distributions must be estimated with the aim of characterizing precisely variability in such properties as the incubation period of the disease, the infectious period and how it changes over time post infection, duration post onset of clinical symptoms to isolation in a healthcare setting, and duration to recovering or death.
The construction of a patient database The creation of a patient database that integrates socio-demographic detail with information on both clinical aspects, such as treatment and outcome, and data of an epidemiological nature, such as contact tracing information and behavioural
Transmission Dynamics and Control of the Viral Aetiological Agent of SARS
questionnaire data, is central to the realtime analysis and control of infectious disease outbreaks. Ideally, the electronic database should have one central point of control (such as in a government department or designated research centre), be web-based with password protection for remote data entry, and be designed to act as a registry and monitoring system to inform policy formulation and analysis day by day throughout the epidemic. All patients suspected of having contracted SARS within a country or administrative region should be entered into this database with a unique patient identifying code. Appropriate measures should be taken to protect patient identity and to conform to existing data protection legislation. At a global level, such country-wide data should ideally be shared with WHO, to inform their policy formulation and advisory notices, either in its entirety or in a pared down form which includes case numbers over time and basic epidemiological plus clinical information. Few countries achieved such real-time collection and synthesis of information during the SARS outbreak. Arguably, the best example of good practice occurred in the Hong Kong Special Administrative Region of China, where the Health, Welfare and Food Bureau of the government (in collaboration with University-based public health professionals), created a system called SARSID (SARS Integrated Database) to collect and collate case information. Even in this setting, however, problems were encountered in linking case information with clinical treatment data held in inpatient acute care settings, and with the results of contact tracing which were held in a separate database managed by the police force. In many countries, the appropriate databases and associated analyses were only fully assembled and completed after the end of the epidemic. As a consequence of the SARS experience (and experiences with other rapidly developing infectious disease outbreaks — see Riley et al.6) the development of appropriate software and the train-
113
ing of personnel to use it should be a priority for all government health departments.
Case definition and clinical symptoms The case definition of SARS has changed since the emergence of the 2002–03 epidemic. However, the following definition developed by WHO was used throughout the main period of spread in 2003 (http://www.who.int/csr/sars/casedefinition/en/). Two criteria were used in this definition as follows. (1) A person presenting after 1 November 2002 with history of high fever (>38°C) and cough or breathing difficulty, and one or more of exposures during the 10 days prior to onset of symptoms. Exposures are defined as close contact (having cared for, lived with, or had direct contact with respiratory secretions or body fluids) with a person who is a suspect or probable case of SARS; history of travel to an area with recent local transmission of SARS; residing in an area with recent local transmission of SARS. (2) A person with an unexplained acute respiratory illness resulting in death after 1 November 2002, but on whom no autopsy has been performed and one or more exposures 10 days prior to onset of symptoms. Exposures are defined as close contact, with a person who is a suspect or probable case of SARS; history of travel to an area with recent local transmission of SARS; residing in an area with recent local transmission of SARS. A probable case was defined by WHO to have three criteria as follows. (1) A suspect case with radiographic evidence of infiltrates consistent with pneumonia or respiratory distress syndrome (RDS) on chest X-ray (CXR). (2) A suspect case of SARS that is positive for SARS coronavirus by one or more assays. (3) A suspect case with autopsy findings consistent with the pathology of RDS without an identifiable cause. Since all SARS diagnoses were based on exclusion of other possible causes (e.g. influenza), a case could be excluded if an
114
Severe Acute Respiratory Syndrome
alternative diagnosis was made at a later stage post admission to hospital. Throughout the epidemic and over the months after its termination on 5 July 2003, many cases have been reclassified on the basis of information from virological assays or serological tests based on sera drawn from patients after cessation of clinical symptoms. Clinical symptoms at admission to hospital, and during hospital stays for confirmed SARS cases (based on virological confirmation) have been recorded in various studies (Chapters 3 and 4). At admission high fever, malaise, cough and headache seem to be the most common symptoms. For confirmed SARS cases, high and prolonged fever and diarrhoea are typical symptoms. Virus can be isolated from sputum, urine and faeces during the mid- to late clinical phase of symptoms.7,8
Some key distributions Incubation period The incubation period is defined as the time from infection to onset of clinical symptoms of disease. This duration is often influenced by factors such as infecting dose of virus, host genetic background, route of exposure and age of patient. The determination of the form of this distribution and its summary statistics at the start of an epidemic of a novel infectious agent are of great importance, given their significance to the duration of quarantine and to contact tracing. The average incubation period also influences the time-scale of the development of the epidemic by its impact
on the rate at which secondary cases are generated. Infection events cannot be observed, but data on patients with short and well defined periods of exposure to known SARS cases can be used to estimate the distribution of the incubation period. Estimation of the incubation period is based on sets of patients with a short exposure period of known date to a suspect SARS case. This information may be collated via contact tracing or from questionnaires. The estimates for SARS are based on approximately 200 cases from 5 regions that experienced moderate to severe epidemics (Hong Kong, mainland China (Beijing and Guangdong), Taiwan, Singapore and Canada). The data are summarized in Table 14.1. Donnelly et al.5 report a mean of 6.37 (subsequently revised to 4.2 after the end of the epidemic), with roughly 95% of all estimates lying between 2 and 14 days. An observed distribution based on 70 cases in Guangdong and Beijing is recoded in Fig. 14.2. Gamma distributions have been used to describe the observed pattern.
Time from onset of clinical symptoms to admission to hospital Very early on in the SARS epidemic it was understood that reducing the time from onset of clinical symptoms to admission to hospital with subsequent isolation was an important measure to reduce the net rate of transmission within a community or country.1 Most countries compiled statistics on this distribution, with Hong Kong, Taiwan and mainland China having access to such information as the epidemic progressed.
Table 14.1 Incubation period of SARS Country/region
Mean (days)
Number of cases
Reference
Hong Kong Mainland China Singapore Canada
4.25 4.5 5.3 4.8
113 70 46 42
Unpublished data (9) (10) (11)
Transmission Dynamics and Control of the Viral Aetiological Agent of SARS
115
18
Number of cases
16
Figure 14.2 Incubation period distribution from 70 cases with a known and short exposure period to a suspect SARS case from Guangdong, China (Chinese CDC, 2003).
14 12 10 8 6 4 2 0 0
1
2
3
4
5
6
7
8
9
10
11
12
Days post exposure
0.2 0.18 Data
0.16
Fitted gamma distribution
P ro b ab i l i ty (P DF )
0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 0
5
10
15
20
25
30
35
Time from onset to hospital admission (days) Figure 14.3 Probability distribution from onset of clinical symptoms to admission to hospital from Hong Kong early in the epidemic.5
Onset and admission times are both observable events. However, allowance must be made in the early phase of a new epidemic in the analysis for censoring due to incomplete observation. If censoring is not taken into account, the distribution will be biased towards short onset-to-admission times, because patients are only eligible to be included in the hospital-based database on admission to hospital. Patients with recent onsets
and long onset-to-admission times are less likely to have been admitted to hospital and thus to be included in any analysis. One such distribution, compiled post the cessation of the epidemic to eliminate the problem of censoring, is recorded in Fig. 14.3. As a consequence of public health announcements using the press and media, many regions managed to encourage individuals with symptoms of severe respiratory tract infection to report rapidly to hospitals. For
116
Severe Acute Respiratory Syndrome
Table 14.2 Mean number of days from onset to admission in Hong Kong Date (2003)
Mean number of days from onset to admission
26 February– 4 March 5–11 March 12–18 March 19–25 March 26 March–1 April 2–8 April 9–15 April 16–22 April 23–29 April 30 April–6 May 7–13 May 14–20 May
9.3 5.1 4.4 4.0 3.0 2.5 2.8 2.3 2.9 2.5 2.1 1.0
140 120
Frequency
100 80 60 40 20 0 1
7
13 19 25 31 37 43 49 55 61 68 74 83 94 106 Time (days)
example, in Hong Kong the mean period shortened greatly over the course of the epidemic, as illustrated in Table 14.2.
Figure 14.4 Distribution of times from admission to discharge from hospital for 1363 SARS patients from Hong Kong.
stay is an important statistic for the effective management of a SARS outbreak, since it describes one aspect of hospital resource utilization.
Time from admission to hospital to discharge
Time from admission to death
The duration of stay in hospital for those who recovered was typically long with a mean in excess of 25 days in Hong Kong. These times and the fit of a gamma distribution to the raw data are plotted in Fig. 14.4. Longer durations of stay were typically related to patient age and the presence of comorbidities. If older patients recovered they had long stays in hospital. Duration of
SARS is a very pathogenic disease with a high case fatality rate. Times from admission to death are again of importance to health-care planners in terms of resource utilization within the hospital setting. In the Hong Kong setting the mean time was 36 days but with a high variance (Fig. 14.5). Age and other comorbidities greatly influenced the mortality rate.
Transmission Dynamics and Control of the Viral Aetiological Agent of SARS
117
14 12
Frequency
10 8 6 4
Figure 14.5 The distribution of times from admission to hospital to death for 238 SARS patients from Hong Kong.
2 0
1
7
13 19 25
31 37 43
49 55
61 68 74 83 94 106
Time (days)
Table 14.3 Case fatality rates (CFR) from 1755 SARS patients in Hong Kong Gender
Age group (years)
Case fatality rate (%)
Lower 95% confidence bound (%)
Upper 95% confidence bound (%)
Male
<30 30–44 45–59 60–74 75+ <30 30–44 45–59 60–74 75+
0.5 12.3 21.9 43.0 65.8 0.0 5.5 10.2 40.3 71.8
0.01 8.0 14.7 32.6 54.9 0.0 3.0 6.0 29.3 57.7
2.7 16.7 29.0 53.5 76.6 1.4 8.0 14.4 51.2 85.9
Female
Case fatality rates (CFRs) SARS induces much morbidity and mortality in vulnerable patients. The term case fatality rate or CFR is widely used to describe the proportion of those who acquire an infection that eventually die from the disease induced by the aetiological agent. Note that the CFR is not strictly a rate — it is a simple proportion or percentage. The first published study of case fatality in a sample of SARS patients for which outcome was known, or adjusted for by appropriate statistical methods for censoring of the data, was that of Donnelly et al.5 Using a modified Kaplan-Meier-like non-parametric method this study gave estimates of 6.8% for patients younger than 60 years and 55.0%
for patients greater than 60 years. At the end of the first epidemic of SARS CoV it is possible to examine the mortality associated with the disease in more detail. Overall, a WHO summary suggests a global average of around 15%. However, this figure hides much variation — there was little mortality in the young and high levels in the elderly. Multivariate analyses pick out age effects and the presence of comorbidities (such as pre-existing heart or respiratory tract disease) as the most important determinants of the outcome. Table 14.3 shows age and gender specific estimates of the CFR from 1755 patients from Hong Kong. Similar patterns have been recorded in Singapore and Taiwan. Lower rates were reported from mainland China (Beijing and Guangdong)
118
Severe Acute Respiratory Syndrome
90
Number of cases
80 70 60 50
Primary cases
40
Secondary cases
30 20 10 0 0
1
2
3
4
5-6
7-8
9+
Days from onset of clinical symptoms
but no detail is available as yet concerning laboratory confirmation of reported SARS cases. In most studies laboratory testing has now become an integral part of mortality assessment to ensure that a reported SARS case, based on clinical features, is confirmed by either virus isolation or serology post recovery.
Infectious period and the typical course of infection A key determinant of the pattern of spread of an infectious agent is the infectiousness distribution. This must be measured during the incubation period prior to the onset of clinical symptoms, and post this onset. Two approaches are possible, namely, the study of secondary case generation and the measurement of viral load given the assumption that this is proportional to infectiousness to contacts. For SARS CoV the efficiency of transmission to close contacts appears to be greatest from patients with overt clinical symptoms, usually during the second week post onset. A study reported at the WHO meeting in Geneva in May 2003 but not as yet published9 based on cases in Singapore suggests that peak infectiousness as judged by the generation of secondary cases is around days 7–8 post onset of clinical symptoms (Fig. 14.6). The main conclusion of this analysis is in good agreement with data from virological studies,10,11 of changes over time post onset of clinical symptoms of
Figure 14.6 Secondary cases generated by primary cases stratified by time from onset of clinical symptoms based on SARS case reports from Singapore.13
SARS CoV. Viral load data from quantitative RT-PCR as well as rates of RT-PCR positivity in nasopharyngeal aspirates and nose or throat swabs progressively increase to peak around day 10 of illness and decline thereafter to a low level at about day 23. Virus in stools seems to start later than in respiratory excretions with a peak between days 12 and 14 and a slower decline thereafter. These results are summarized in Fig. 14.7. Virus also detected in urine indicated wide organ involvement in pathogenesis.7,8 Retrospective quantitative studies of viral shedding in patients post recovery and discharge from hospital, based on collected samples of respiratory tract excretions, faeces and urine, are currently under way in a number of countries. For the study of transmission dynamics, quantitative measurements are of great importance in order to define a distribution of infectiousness pre- and post onset of illness. The viral shedding studies that are available to date suggest that transmission could occur via close contact involving respiratory tract excretions, and via faecal or urine contamination of surfaces.
Routes and settings of transmission The epidemics of SARS CoV in different countries with a moderate to large number of cases were characterized by a small number of superspreading events, where one case generated large numbers of secondary cases. It is important to note that these events were
Figure 14.7 Percentage of patients shedding SARS-CoV in respiratory tract secretions, stool and urine as determined by RT-PCR, stratified by days post onset of illness (from references 7, 8, 13).
Percentage of sam ples positi ve for SARS-CoV
Transmission Dynamics and Control of the Viral Aetiological Agent of SARS
119
100 90 80 70 60 50
Respiratory tract
40
Urine
Stool
30 20 10 0
most probably created by different combinations of person-related characteristics (e.g. high viral shedding) and environmental factors (e.g. contamination by fomites or close contact in a health-care setting). Evidence from Guangdong province in China, during the very early stages of the global epidemic, suggests that most cases seem to have occurred in food handlers (individuals who slaughter, handle and sell animals and meat, and those who prepare and serve food).8,12 Post this initial stage, the primary mode of person to person transmission appears to be via direct contact with respiratory droplets and exposure to fomites in settings where close contact occurred either in a household or a healthcare facility. In the total epidemic worldwide approximately 21% of recorded cases were in health-care workers.13 Other transmission events occurred in the general population (often unknown in nature) either in workplaces, taxis or airplanes. The localized nature of transmission in defined environments suggests an agent of low transmissibility that requires close contact, either with an ill patient or with a recently contaminated surface. There is no direct evidence of faecal–oral transmission but the high and persistent viral shedding in faeces suggests this could be a route of significance. The exact route of transmission is always difficult to determine and the relative importance of respiratory, faecal and urine excretions remains unclear at present. For the purposes of model construction, the most important factors are the settings in
0-2
3-5
6-14
15-17
21-23
Days post onset of illness
which transmission occurs, the incubation and infectious periods and the average rate at which secondary cases are generated.
Epidemic patterns In regions with significant numbers of cases (see Fig 14.1), the typical pattern of growth post introduction was an initial period of stuttering chains of transmission, interspersed with one or two major superspreading events. This was followed by a period of exponential growth slowing to reach a peak and then a period of steady decline with perhaps one or more superspreading events leading to temporary resurgence, before the epidemic decayed with a cessation in chains of transmission. The decay phases were of varying durations, depending on the efficacy of the interventions introduced in the different regions. These patterns are displayed in Fig. 14.8 for Hong Kong, Taiwan, mainland China and Singapore. The epidemic in mainland China had two phases, with initial spread in Guangdong province followed by a major epidemic in Beijing.
Mathematical models of transmission In the emergence of a new disease, mathematical models of the transmission dynamics of infectious agents are valuable tools in making assessments of both what needs to be measured to understand the spread of the epidemic better and what interventions used alone or in combination are most
Number of cases
60
-0
20
*As of 10 July 2003, 21 additional probable cases of SARS have been reported from Hong Kong SAR, China, for whom no dates of onset are currently avaliable. Source: Department of Health, Hong Kong Special Administrative Region of China
Probable cases of SARS by date of onset Singapore, 1 February–10 July 2003 (n = 206)
14
7-
ar -M
25
Date of onset
Date of report *As of July 2003, 5327 probable cases of SARS have been reported from China. This graph includes 223 probable cases of SARS who had been discarded and for whom dates of report could not be identified. Source: Ministry of Health, China and WHO
Ap 3 r-0 3 -A pr 3- -03 M ay -0 16 -M 3 ay -0 29 -M 3 ay 11 -03 -J un -0 24 3 -J un 0 7Ju 3 n03
3
3 -0
-0 ar
-M
12
-F
eb
-0
03 b-
-F
Fe 1-
7-
14
-0
ay
ay
-M
10
-M
pr -A
26
24
3
3
pr -A
12
-M
-0
3
-0 ar
-0 ar -M
16
29
3
-0
-0
ar
1-
15
M
eb
b-
-F
Fe
-0 Ju 3 n03 21 -J un 5- 03 Ju n03
0 3
0 -0 3
20
3
100
3
40
27
200
80
eb
300
03
Number of cases
400
Probable cases of SARS by date of onset Taiwan, China, 1 February–10 July 2003 (n = 671)
30
-M a 2- r Ap 14 r -A p 26 r -A pr 8M a 20 y -M ay 1Ju n 13 -J un 25 -J un 7Ju l
ar M
9-
eb
Fe
Date of onset
Date of onset Source: Ministry of Health, Singapore
21
7-
1-
3 -0
ar -M
25
ar -M
12
eb
-0
3 -0
-F
27
eb
bFe
-F
-F
0 b
0 -0 Ap 3 r-0 20 3 -A pr 0 3M 3 ay -0 16 3 -M ay -0 29 3 -M ay 11 -03 -J un -0 24 3 -J un 7- -03 Ju n03
5 3
2
14
1-
10
25
4
15
-F
6
20
13
8
25
eb
Number of cases
10
03
Number of cases
12
Source: CDC Taiwan, China and WHO
Figure 14.8 Epidemic patterns in Hong Kong, Taiwan, mainland China and Singapore (source WHO).
Severe Acute Respiratory Syndrome
100
500
1-
Probable cases of SARS by date of onset Hong Kong SAR, China, 1 February–10 July 2003 (n = 1734*)
120
120
Probable cases of SARS by date of report China, 1 February–10 July 2003 (n = 5550*)
600
Transmission Dynamics and Control of the Viral Aetiological Agent of SARS
likely to succeed in bringing it under control. Models provide a template to integrate epidemiological and biological data. Examples include the design and evaluation of childhood disease immunization programmes, predicting the demographic impact of the HIV epidemic in different regions, and analysing the spread and control of the 2001 foot-and-mouth epidemic in Britain.9,10 In the early to mid stages of the SARS epidemic, a number of research groups formulated mathematical models of viral spread of varying complexity. On the one hand was the deterministic susceptibleinfected-recovered frameworks (SIR),14 while at the other, more complex stochastic frameworks with some representation of spatial structure and mixing.15 Approaches of intermediary complexity were also adopted.16 Epidemics can be viewed as chain reactions of disease spread, based on contact between people and possible transmission. The chain is an expanding one if each primary case, on average, generates more than one secondary case. The average number of secondary cases generated in a susceptible population is termed the basic reproductive number, R 0. As the epidemic develops the effective reproductive number at time t, Rt, which describes the generation of secondary cases in a partially susceptible population, decreases in value and eventually settles to unity if the disease becomes endemic. The pattern of epidemic growth is governed by two factors: the number of secondary cases generated by one primary case at the start of the epidemic, R0, and the average time taken for the secondary cases to be infected by a primary case, termed the generation time or Tg. Essentially, R 0 determines how intensive a policy will need to be to control the epidemic, whereas both Tg and R 0 determine the time available to implement suitably intensive controls. For diseases that are highly infectious with short incubation periods, such as measles (R 0 > 17, Tg < 11 days), population-wide control requires the long-term reduction of
121
the recruitment of susceptible people, through widespread childhood immunization. In contrast, for less infectious agents (R 0 around 2–10), that have longer incubation periods (Tg >20 days), if an outbreak is detected in its earliest stages, there is sufficient time for localized control measures to be successfully applied. One aim in model development for SARS spread in defined regions was therefore the estimation of both the basic reproductive number and the generation time in order to assess how difficult it might be to bring the epidemic under control. Longitudinal monitoring of the magnitude of a further parameter, the effective reproductive number at time t, Rt, also provides crucial information on the success of the interventions that have been implemented to date. For an epidemic to be under control the magnitude of Rt must be less than unity — each primary case generating on average less than one secondary case such that the chains of transmission stutter to extinction. What is the best method to estimate R 0 in an emerging crisis? One simple approach is to estimate the doubling time of the epidemic, td, in its early stages, taking due account of the stochastic fluctuations that typically occur in the early stages of any epidemic’s development.10 Estimation of the key parameter, R 0, can be achieved by fitting an exponential growth equation. For simple epidemics of a directly transmitted respiratory or gastro-intestinal pathogen, during early growth in a totally susceptible population, the doubling time td is related to the magnitude of R 0 by the simple expression: t d = ln(2 ) D ( R0 - 1)
[14.1]
where D is the average duration of the latent plus infectious periods (essentially the generation time Tg). In the absence of knowledge of D, a range of assumptions will have to be made for its value to get some idea of the magnitude of R 0. This approach is illustrated diagrammatically in Fig 14.9. In the
122
Severe Acute Respiratory Syndrome
Exhaustion of susceptibles
Rate of new infections Random
Equilibrium, or recurrent epidemics
y μ e(R0-1)/Tg t establishment
exponential growth
Time endemicity
Figure 14.9 In initial stages transmission leads to exponential growth (once infected numbers are great enough to make random effects small), until the epidemic begins to run out of people to infect. This effect is encapsulated in the effective reproduction number at time t, Rt, which declines from its maximum, R0, as the susceptibles are depleted. The diagram illustrates the classic epidemic profile.9,18
Susceptibles (S)
Infected but not infectious (L)
Asymptomatic but infectious (I)
Hospitalized and eventually dies (HD)
Symptomatic and infectious (Y)
Hospitalized and recovers (HR)
Recovers and leaves
early stages of the SARS epidemic, contact tracing data, especially that based on superspreading events, suggested that the value of D was somewhere between one and a few weeks.17 The simplest models are based on the classification of the population into various disease states such as susceptible, infected but not yet infectious (latent), infectious and recovered or dead. Equations for each state were constructed denoting rates of flow between the states. These rates may be represented as constants (with an exponential waiting time distribution for movement between states) or by distributed variables
Figure 14.10 Schematic flow diagram of transmission within a population. Susceptible individuals, S, become infected and enter a latent class, L. They then progress to a short asymptomatic and potentially infectious stage (see text), I, before the onset of symptoms and progression to class Y. It is assumed every infected individual eventually attends hospital and either recovers, HR, or dies, HD.
given availability of appropriate data (see Figs 14.2 to 14.5). For SARS, the simplest structure is represented in Fig. 14.10, which pictures in a flow chart an additional class to those outlined above to reflect isolation or quarantine, such that a patient is infectious but unable to transmit. Such models may be stratified by factors such as age, spatial location or environmental setting (i.e. within a health-care setting), and may be formulated within a deterministic or stochastic framework. Population heterogeneity ideally needs to be accounted for — reflecting, for example, variation in infection risk with spatial loca-
Transmission Dynamics and Control of the Viral Aetiological Agent of SARS
tion, variation in contact rates between groups, and between-case variability in infectiousness.10,15 Heterogeneity in Rt between cases appears to be particularly important for SARS due to the occurrence of ‘superspreading events’ (SSEs). These are defined as rare events where, in a particular setting, an individual may generate many more than the average number of secondary cases. The Hong Kong epidemic, for example, was characterized by two large clusters of cases, together with ongoing transmission to close contacts. In the first cluster, at least 125 people were infected on or soon after 3 March in the Prince of Wales Hospital (PWH) by the index patient for the Hong Kong epidemic.19 In the second cluster, an unknown number of people were infected from a probable environmental source in the Amoy Gardens (AG) estate (Kwung Tong district). Following mixing with fellow residents, families and friends, over 300 people became infected. Examination of local reports of SARS investigations support the distinction between these two large superspreading events and the other contact-based infections, where many occurred in a hospital setting. It may be that the distinction between typical infection events and SSEs reflects the two different routes of transmission so far identified as likely for this virus; namely, respiratory exudates and faecal–oral contact. However, much still remains uncertain about possible routes of transmission in these SSE settings. Choice of a suitable model framework must be governed by the degree to which the investigator wishes to capture varying degrees of heterogeneity. A variety of approaches are possible, ranging from a simple deterministic compartmental approach with mixing between patches or settings, to a spatially explicit, individual-based stochastic simulation structure; Riley et al.15 adopted a stochastic, metapopulation compartmental model, given the quality of the data available in real-time from the Hong Kong region. A
123
metapopulation approach was considered to be appropriate because the incidence of SARS varied substantially by geographical district in Hong Kong. A stochastic model was employed since chance fluctuations in case numbers can be large in the early stages of an epidemic. Stochastic models predict both average trends and variability, so that a more robust assessment can be made to examine what changes are likely to be caused by chance and what changes genuinely reflect process and the impact of interventions. They classified the population of each district of Hong Kong into susceptible, latent, infectious, hospitalized, recovered and dead individuals. Data were available to characterize the distributions around these transitions from one compartment to the next (see Figs 14.2 to 14.5). Epidemiological coupling between districts was assumed to depend on their adjacency. Incubating and infectious categories were further divided into multiple stages, chosen in number and duration so as to match accurately the estimated delay distributions determining disease progression and diagnosis (Figs 14.2 to 14.5).5 Multiple realizations of the stochastic model were performed, both for parameter estimation and to generate predicted case incidence time-series. The mean time from the onset of symptoms until hospital admission and subsequent isolation of suspect SARS cases reduced significantly over the course of the epidemic. Changes in the onset-to-hospitalization distribution were treated as an input to the model. They assumed that infectiousness began just before the onset of symptoms and remained constant during the symptomatic phase. With hindsight this was a reasonable assumption. More sophisticated assumptions could now be made given knowledge of how infectiousness changes post onset of illness10 (Fig.14.7). Simulations were seeded explicitly with the PWH cluster. Model fit to the observed case-incidence data in Hong Kong was qualitatively good, both in terms of capturing the temporal development of overall incidence and the pattern of spatial
124
Severe Acute Respiratory Syndrome
spread (Fig. 14.11). Fitting the model to observed trends provided estimates of both Ro, temporal changes in Rt, and the generation time Tg,2 and as such this approach has many advantages over the use of simpler model constructs given good availability of data in real time. The disadvantages of this more sophisticated and statistically robust approach are the need for good computational facilities and the availability of a preexisting model structure that could be quickly adapted for the new disease in an emerging crisis.
160 140 Daily incidence
120 100 80 60 40 20 0 Figure 14.11 Figure adapted from Riley et al.15 showing the fit of a meta-population stochastic mathematical model (solid line) of SARS transmission to observed time series of case reports from Hong Kong. The vertical bars denote the 95% prediction intervals based on an average of 1000 model realizations. Prediction intervals are generated from extreme realizations at the boundaries of the multivariate confidence intervals.
The case reproductive number, R0 Various estimates for R 0 were published during the mid- to end-phases of the global epidemic. These were based on various methods as briefly alluded to in the previous section and on data for a variety of regions. Encouragingly, however, the estimates were in good agreement, despite the different methods employed and the various epidemics in different regions, with an average value of approximately 3 independent of setting (Table 14.4). It is important to note that these estimates were based on case reports of individuals with overt clinical symptoms that led to an initial diagnosis of SARS. In many situations case definition was based solely on clinical criteria and not on confirmation by virological or serological diagnosis. In addition, they do not account for transmission events that created mild symptoms of disease not recorded as SARS. Given the low incidence of SARS in children, relative to their proportional representation in affected communities, it is possible that many transmission events went unrecorded. However, serological surveys in children and other potentially exposed groups post the end of the epidemic in specific regions or countries have not revealed many asymptomatic infections. For example, in Hong Kong one survey in 200 children revealed only two positive tests and these individuals were recorded as suspect SARS cases.13 Further
Table 14.4 Estimates of the basic reproductive number, R0 Reference
Region
R0
Comments
Riley et al., 20036
Hong Kong
2–3
Excluding superspreading events
Lipsitch et al., 20037 Wallinga, 2003 (unpublished; reported in Ref. 10)
Canada and Singapore Singapore
3 3.3
Transmission Dynamics and Control of the Viral Aetiological Agent of SARS
surveys are under way amongst hospital staff and individuals from environments in which superspreading events occurred. To date there are no confirmed cases of transmission from asymptomatic individuals.13 The main conclusion to draw from these estimates of R0 is that SARS CoV is of low transmissibility by comparison with other directly transmitted viruses such as influenza A (R0 of around 7 or more)18 and the measles virus (R0 of around 15–18 prior to wide-scale immunization).10 Further confirmation of this low level of transmissibility is provided by analyses of SARS cases within those placed in quarantine as a result of close contact (either in health-care, family, work or transport settings) with a suspect SARS case. In Taiwan, for example, of 131 132 people placed in quarantine, only 45 were recorded as probable SARS cases (CDC Taiwan, 2003). Further research on the estimation of R0 is currently under way, based on retrospective analyses of contact tracing data from Singapore, Taiwan and Hong Kong. The factors that triggered superspreading events remain poorly understood at present. Such events were of great importance in the epidemics in Canada, Singapore, Hong Kong and Taiwan. In Singapore, for example, 103 cases in the early stages of the epidemic were probably generated by contact with only 5 source cases.3 In Hong Kong, a similar pattern pertained, with 125 people infected by one index case in the Prince of Wales hospital and over 300 infected in one event in the Amoy Gardens residential estate.19 The causes probably involve both environmental factors and determinants of the infectiousness of the index patient (the amount of virus excreted in respiratory tract exudates, faeces and urine, and the duration of shedding). The importance of environmental factors is well illustrated by the high proportion of infections worldwide that occurred in healthcare settings (approximately 21% of all reported SARS cases).13 It may be that the distinction between superspreading events
125
and typical transmissions reflects the route of transmission, but much remains uncertain. In the estimation of R0 it is important to recognize that this key epidemiological parameter is a mean drawn from a distribution with high variance. Those events in the right-hand tail of the distribution of secondary case generation may or may not constitute a separate class generated by distinct environmental or case infectiousness factors.
The impact of interventions The importance of measuring R0 is in part related to the need to gain insight into how difficult it might be to bring a given epidemic under control. Infectious agents with high R0 values and short generation times are typically difficult to control, by comparison with agents with low R0 values. To evaluate how different control interventions might impact on any given epidemic during its emergence and early growth we ideally need to use a mathematical model of transmission that embeds estimates of transmission efficiency and the details concerning the typical course of infection, to explore both what works best and in what combination, and the degree to which a specific intervention must be applied. For example, in the case of SARS obvious questions are how important is it to reduce the time between onset of clinical symptoms to isolation or quarantine within a health-care setting, and what time interval should be the target (i.e. within 1 or 2 days)? For international and governmental agencies, the questions may be more complex, such as the value of introducing travel restrictions to and from affected areas, and the effectiveness of screening of passengers at airports for elevated temperature. Such measures may be costly to introduce and may have grave economic consequences for affected regions. It is therefore highly desirable to have available some sort of quantitative template to permit assessments to be made during the heat of the crisis.
126
Severe Acute Respiratory Syndrome
For SARS the options for intervention within a country were limited to public health measures, in the absence of a vaccine or effective therapies. There are essentially seven intervention categories, namely: (1) restrictions on entry to the country and screening at the point of arrival for fever; (2) isolation of suspect cases; (3) the encouragement of rapid reporting to a health-care setting post the onset of defined clinical symptoms, with subsequent isolation; (4) rigorous infection control measures in health-care settings; (5) restrictions on movements within a country (restricting travel, limiting congregations such as attendance at school) and (6) contact tracing plus isolation of contacts. To study in quantitative terms their respective impacts on transmission, or the impacts of different combinations applied with varying efficiencies, each category must be captured within the mathematical model. One simple illustration is given in the flow chart in Fig. 14.10, where a category denoting infectious but isolated is represented. Figure 14.12 depicts schematically how an epidemic is influenced by changes in the magnitude of the effective reproductive number at time t, Rt. Interventions may reduce the magnitude of Rt but fail to reduce its value
to less than unity. In these circumstances the incidence of new infections declines, as if the epidemic is waning, but in reality the trend is to a new endemic state with the infection still persisting. During a crisis, it is difficult to interpret correctly whether or not changes in incidence post the introduction of control measures indicate (a) decay to probable extinction or (b) continued transmission but at a reduced rate. It is in these circumstances that mathematical models can play an important role as a template for the repeated estimation of Rt through time. Once Rt drops below unity, and stays there, the epidemic is under control provided no relaxation in implementation of the introduced interventions takes place. An application of this approach in Hong Kong is recorded in Fig. 14.13. This figure records the estimated change in the effective reproductive number in three blocks of time post the onset of the epidemic in Hong Kong.15 Combinations of reductions in onset-to-hospitalization, in population contact rates (mixing) and with health-care setting transmission (improved 4.00 3.50 3.00
1.00 0.50 30-Apr
2-Apr
0.00 16-Apr
Time
1.50
19-Mar
Epidemic stopped (R0 < 1)
2.00
5-Mar
Epidemic slowed (R0 < 1)
19-Feb
0 0.5 1 1.5 2 3 4 5 6
No controls
Rt XSS
Rate of new infections
2.50 R after controls
Day Figure 14.12 Control interventions introduced at time t (the vertical grey line) can have varying effects depending on how much they affect the effective reproductive number at time t, Rt. The number of cases may decline without the epidemic decaying to extinction if Rt is not reduced to below unity in value.
Figure 14.13 Changes in the effective reproductive number, Rt (excluding superspreading events) over time in Hong Kong (see Riley et al.6 Confidence intervals are also shown. The selfsustaining threshold Rt = 1 is shown as a light grey line.
Transmission Dynamics and Control of the Viral Aetiological Agent of SARS
infection control procedures) reduced the effective reproductive number to around 1.0 by 21 March, 0.9 by 26 March and then to 0.14 by 10 April. Teasing out the relative contributions of different interventions is more difficult to achieve if the only quantitative outcome measure is that of cases of disease reported each day. Ideally, other data are required such as changes in travel patterns, the decay in the interval between disease onset and isolation, and the fraction of all contacts of a suspect case traced within a defined time interval. For SARS, aside from case numbers, very few data were available day by day during the course of the epidemic in any given setting, except for information on the temporal change in the distribution of times between onset and isolation (see Fig. 14.3 and Table 14.2). Various attempts have been made to dissect the differing impacts of various interventions but with limited success to date. Some of these analyses were based on fitting models to case data to estimate parameters reflecting the efficacy of defined controls,15 while others are more abstract in the sense that parameter values were changed within a model and conclusions drawn on model-predicted trends.10,14 Conclusions drawn from both approaches depend on model structure and parameter assignments, and as such should be accepted with caution. Pooling the results from the published analyses (largely from references 10, 14, 15), the following conclusions can be drawn. Isolation and quarantine, contact tracing, improved infection control procedures and selfimposed movement and mixing restrictions that limited contact were very effective in combination. They induced the dramatic changes in Rt post the peaks in the epidemic in all regions that were badly affected. Reductions in the onset to admission times were important in most settings due to the late onset of infectiousness post onset of symptoms of disease. High impact can also be attributed to changes in contact rates (mixing) and better infection control
127
plus isolation and quarantine of symptomatic patients in hospital settings. Spatially explicit models also suggest that restriction on movement between locations within defined communities played a useful role in limiting spread. Transmission in most regions was highly localized. Although somewhat controversial at present, detailed model based analyses of the effectiveness of contact tracing for SARS suggest that it was far less effective than commonly perceived. More retrospective analyses are urgently required in this area, using the larger contact tracing databases. A similarly urgent priority is analysis of the effectiveness of travel directives. Preliminary unpublished studies suggest that they were effective in restricting between country spread. However, this conclusion remains to be confirmed by detailed analyses of travel patterns between major cities before, during and after the SARS epidemic.
Why did the epidemic come under control? Global success in the control of SARS was partly due to certain epidemiological and biological characteristics of the infectious agent. In the absence of effective vaccines or treatment, understanding the factors which make containment of SARS feasible is important for evaluating how best to control future outbreaks of newly evolved pathogens. Two public health policy options exist for controlling the spread of a novel directly transmitted infectious disease agent: (a) effective isolation of symptomatic individuals (which includes rapid hospitalization post onset of clinical symptoms) and (b) tracing and quarantining of the contacts of symptomatic cases. Both measures rely on rapid dissemination of information to facilitate accurate diagnosis of the symptoms of the disease. For SARS, the timing of the onset of symptoms relative to peak infectivity is probably the most crucial factor in the success of simple public health interventions aimed at
128
Severe Acute Respiratory Syndrome
reducing transmission. In SARS patients, peak viral load in respiratory secretions appears to occur between 5 and 10 days after the onset of symptoms.7 Although viral load does not always predict infectivity, the very low levels measured in the days immediately after the onset of symptoms suggest that peak infectivity occurs somewhat later. Also, no confirmed cases of transmission from asymptomatic patients have been reported to date in detailed epidemiological analyses of clusters of SARS cases.13 This suggests that for SARS there is a period after symptoms develop during which people can be isolated before their infectiousness increases. It is during this period that transmission can be very effectively interrupted by isolation and quarantine. The second feature is the low transmissibility of SARS CoV, with its moderately low basic reproductive number, with the exception of the settings in which superspreading events occurred. A recent study by Fraser et al.,6 attempts to analyse why SARS was so effectively contained, using a generic mathematical model for directly transmitted agents. The approach adopted is comparative, and centres on the definition of two key properties of transmission and the typical course of infection, namely, the basic reproductive number R0, and the proportion of the area under the viral load curve in a typical patient that occurs prior to the onset of easily diagnosable symptoms. This proportion is assumed to reflect infectiousness before and after the onset of symptoms. They show that the proportion of infections that arise prior to the onset of symptoms (or via asymptomatic infection) is as strong a predictor of success of the simple public health control measures as the inherent transmissibility of the aetiological agent as measured by R0. From published studies, they estimate these quantities for two moderately transmissible viruses, SARS CoV and HIV, and for two highly transmissible viruses, smallpox and pandemic influenza A. They conclude that SARS and smallpox are easier
to control using these simple public health measures. This study therefore suggests that in an emerging epidemic of a novel agent, both clinical epidemiological studies of pathogen load and clinical symptoms, plus contact tracing to assess when transmission occurs during the typical course of infection should be priorities.
Discussion What lessons are to be drawn from analyses of transmission dynamics about how to manage better the next epidemic outbreak involving a novel pathogen? The first important lesson concerns data capture. For effective ongoing public health policy formulation, information must be captured country-wide on a daily basis. Unique patient identifiers should be used to record all socio-demographic, clinical, treatment and epidemiological information (e.g. contact tracing — who got infected from whom). Careful thought is needed to define data fields and apply effective capture across all health-care settings. Web-based, passwordprotected systems need to be ready to be put into action, with information fed daily to one centralized database for analysis and interpretation. No one country or region had such a system in place prior to the emergence of SARS. Ideally some common database/software structure should be used across all countries and regions, such that WHO could capture information relevant for the global management of an outbreak. The second lesson concerns the appropriate use of statistical techniques to estimate key distributions that shape the pattern of the epidemic. Of particular note in the SARS outbreak were inaccurate reports of case fatality rates. Low estimates were derived using inappropriate methods. During the rapid increase of cases early in an epidemic, the right-hand tails of observed distributions of such properties as time to death, time to recovery and the incubation period may be censored when compared with the true distributions. This is because
Transmission Dynamics and Control of the Viral Aetiological Agent of SARS
not enough time has passed to record accurately the correct number of long waiting times. The third lesson concerns clinical epidemiology. As noted in the last section of this chapter, a key property that influences the likelihood of success of control is the typical relationship between onset of clinical symptoms and viral load as a surrogate marker of infectiousness. With some notable exceptions,11 not enough attention was paid to quantifying these properties day by day for large samples of patients. The fourth lesson concerns the need for analyses and model development prior to the emergence of a new problem. Generic models should be constructed for a wide variety of pathogen types (including vector transmitted viruses), with differing biological and transmission properties. More than one type of model is needed to deal with problems at a local level (perhaps even within a health-care setting), at a community scale and within an international context. Prior exploration of what control option or combination of options works best in defined situations is a clear priority. One approach is that adopted by Fraser et al.6 but other model templates and methods of analysis would be highly desirable, using more sophisticated stochastic individual based models. We need to understand much better how different interventions impact on agents with given properties, particularly if it is possible to put such measures in a cost-effectiveness framework. The issues of travel directives and airport screening for fever immediately come to mind. We understand very little at present about how effective they were in limiting the spread of SARS CoV. Ideally, mathematical models would be used to provide a framework within which to evaluate possible intervention strategies during the course of future epidemics. Estimation day by day of the effective reproductive number Rt (the average rate of generating secondary cases) provides a quantitative measure of success or failure.
129
The fifth and final lesson concerns the potential value of contact tracing data. The speedy and accurate collection of contact tracing data serves many purposes. In the first place it obviously acts as a public health measure in the sense that it gives an opportunity to isolate or quarantine those in close contact with a suspect case. Equally important, however, is the fact that it provides a source of information to estimate key parameters and distributions. The incubation period is one, as are estimates of the distribution and mean of the effective reproductive number. In most settings that were badly affected by SARS, the percentage of contacts traced within a few days was very limited. We need to learn more about how best to do such tracing for directly transmitted agents, and how best to analyse and interpret the data given uncertain denominators. In badly affected regions, the SARS epidemic caused much suffering, significant mortality, great disruption to social and work activities and considerable economic losses. Draconian public health measures involving the isolation and quarantining of hundreds of thousands of people, and tight restrictions on travel had to be put in place in some countries. However, it was brought under control — and relatively quickly — with WHO playing a vital role in co-ordinating the international response. The quick and effective response of WHO to the SARS crisis did much to restore faith amongst the many critics of the effectiveness of international agencies with large bureaucracies and limited resources for action. But it is difficult to escape the conclusion that the world community was very lucky this time round, given the very low transmissibility of the agent, plus the fact that fairly draconian public health measures could be put in place with great efficiency in Asian regions where the epidemic originated. Given the litigious nature of people in North America in particular, and to a lesser degree in Western Europe, the control of superspreading events in these
130
Severe Acute Respiratory Syndrome
regions might have presented greater problems if mass quarantining had been required. In the next global epidemic of a directly transmitted short generation period infectious agent we may not be so lucky, either in terms of the biology of the agent or the region of its origin. Thus one of the major dangers arising from the effective control of SARS is complacency, spawning sentiments of the type — ‘we have been successful once — we will be again’. This may be far from the truth.
References 1 Peiris JS, Lai ST, Poon LL et al. Coronavirus as a possible cause of severe acute respiratory syndrome. Lancet 2003;361: 1319–25. 2 Ksiazek TG, Erdman D, Goldsmith CS et al. A novel coronavirus associated with severe acute respiratory syndrome. N Engl J Med 2003;348: 1953–66. 3 Drosten C, Gunther S, Preiser W et al. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. N Engl J Med 2003;348: 1967–76. 4 Marra MA, Jones SJ, Astell CR et al. The genome sequence of the SARS-associated coronavirus. Science 2003;300: 1399–404. 5 Donnelly CA, Ghani AC, Leung GM et al. Epidemiological determinants of spread of causal agent of severe acute respiratory syndrome in Hong Kong. Lancet 2003;361: 1761–6. 6 Fraser C, Riley S, Anderson RM et al. Factors that make an infectious disease outbreak controllable. Proc Nat Acad Sci 2004;101: 6146–51. 7 Peiris JS, Chu CM, Cheng VC et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;361: 1767–72. 8 Chan KH, Poon LLM, Cheng VCC et al. Detection of SARS coronavirus (SCoV) by RT-PCR, culture, and serology in patients with acute
9
10
11
12 13
14
15
16
17
18
19
respiratory syndrome (SARS). Emerg Infect Dis 2003;10: 294–9. Ferguson NM, Donnelly CA, Anderson RM. The foot-and-mouth epidemic in Great Britain: pattern of spread and impact of interventions. Science 2001;292: 1155–60. Lipsitch M, Cohen T, Cooper B et al. Transmission dynamics and control of severe acute respiratory syndrome. Science 2003;300: 1966–70. Leung GM, Rainer TH, Lau FL et al. for Hospital Authority SARS Collaborative Group. A clinical prediction rule to identify emergency room attendees with severe acute respiratory syndrome (SARS) during an outbreak. Ann Intern Med 2004;141: 333–42. Centers for Disease Control, China: Personal communication 2003. World Health Organization. Consensus document on the epidemiology of severe acute respiratory syndrome (SARS). 2003. http:// www.who.int/csr/sars/en/WHOconsensus. pdf Galvani AP, Lei X, Jewell NP. Severe acute respiratory syndrome: temporal stability and geographic variation in case-fatality rates and doubling times. Emerg Infect Dis 2003;9: 991–4. Riley S, Fraser C, Donnelly CA et al. Transmission dynamics of the etiological agent of severe acute respiratory syndrome (SARS) in Hong Kong: the impact of public health interventions. Science 2003;300: 1961–6. Lloyd-Smith JO, Galvani AP, Getz WM. Curtailing transmission of severe acute respiratory syndrome within a community and its hospital. Proc R Soc Lond [B] 2003;270: 1979–89. Health Canada. Latest Canadian numbers on SARS. http://www.hc-sc.gc.ca/pphb-dgspsp/ sars-sras/cn-cc/20030903_e.html Ferguson NM, Galvani AP, Bush RM. Ecological and immunological determinants of influenza evolution. Nature 2003;422: 428–33. Lee N, Hui D, Wu A et al. A major outbreak of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348: 1986–94.
Chapter 15
The Seasonality of Respiratory Virus Diseases: Implications for SARS? JC de Jong and WL Lim
Seasonality of infectious diseases
Seasonality of respiratory virus diseases
Human infectious diseases differ widely in the extent of fluctuations in incidence. Peak incidences may occur at different times of the year (seasonal variation). With some diseases, presumably due to widespread postepidemic immunity, high incidences occur only at intervals of 2 or more years (cyclic variation). With the same agent, seasonality may be different for different regions, especially when the climates are different. Although the underlying mechanisms still remain largely unexplained, the seasonality for many infectious diseases is stable and well-documented.1,2 Mass immunization reduces the reproductive rate and hence increases the period between epidemics.3 Knowledge of the phenomenon of seasonality is not only useful in clinical management of patients but also essential for the design and implementation of prevention strategies and for planning outbreak control. In the case of a new emerging disease such as SARS, seasonality and immune interactions between related coronaviruses may have an influence on disease reemergence. It is relevant therefore to attempt to understand the reasons for virus seasonality in general.
In the tropical region of Hong Kong there are usually two peaks of respiratory virus diseases, a major one in February–March and a minor one in July (Fig. 15.1). In temperate climates these diseases generally have a marked predilection for the winter. Consequently, their large impact on human health is apparent by the annual rises in morbidity and mortality from total respiratory disease and even mortality from all causes in the winter.4 Influenza virus is thought to be mainly responsible for the concomittant increase in mortality, but other respiratory viruses such as respiratory syncytial virus and rhinovirus may also play a significant role.
Influenza The best known example of seasonal variation of an infectious disease is perhaps influenza. In the (sub)tropics influenza can occur throughout the year but an increase in the incidence of influenza is generally observed in late winter/early spring and in the tropics again in the summer months or during the rainy season (Fig. 15.2).5,6 In temperate and colder regions of both hemispheres the disease occurs as winter epidemics and even the pandemic influenza virus strains of 1918, 1957, and 1968 had in these areas a major impact only in the winter.7,8 Influenza epidemics are variable in
131
132
Severe Acute Respiratory Syndrome
Figure 15.1 Monthly respiratory virus isolations at the Government Virus Unit of Hong Kong (1996–2002). Detected viruses include influenza viruses, parainfluenza viruses, respiratory syncytial virus, rhinoviruses, and adenoviruses.
Figure 15.2 Influenza in Hong Kong. The monthly numbers of influenza virus isolations at the Government Virus Unit of Hong Kong are given over the period 1991–2003. The large difference in the numbers of virus isolates before and after 1997 is partly due to a sampling artefact. In the autumn of 1997, sporadic H5N1 viruses were detected in Hong Kong and the influenza surveillance therefore was enhanced during 1998. In the autumn of 1997, a new H3N2 virus variant (A/Sydney/5/97-like virus strains) emerged worldwide and boosted influenza activity in the next year. As a consequence, in 1998, independent from the size of the influenza epidemic, many more patients were sampled than usual. This effect slowly decreased in subsequent years but in 2003 the epidemic of SARS with enhanced surveillance boosted the sampling size once more and resulted in increased numbers of influenza virus isolates.
size as well as in date of start and duration (Fig. 15.3). Although it is difficult to distinguish clinically between influenza and other respiratory diseases, simultaneous virus isolations from cases of influenza-like illness link the clinical and virological data to each other in time (Fig. 15.4). In this way the seasonality of influenza was demonstrated in the Netherlands over three decades, during which period the size of the influenza epidemics gradually decreased (Fig. 15.5).
Parainfluenza virus diseases Worldwide, parainfluenza viruses can be isolated around the year.9 In Hong Kong, these viruses are also recovered throughout the year with increased incidence in the winter (Lim, unpublished data). This is particularly so for parainfluenza virus 1. Interestingly, this also holds true for the incidence of parainfluenza virus 3 which is in contrast to the reports from the temperate zone for this virus (see below). In moderate climates, diseases by parainfluenza viruses 1 and 2 show a predilection for the autumn or winter and parainfluenza
The Seasonality of Respiratory Virus Diseases: Implications for SARS?
133
Figure 15.3 Influenza in the Netherlands, 1998/99, 1999/2000, and 2000/01 seasons. Triangles stand for the 1998/99 season, squares for the 1999/2000 season and diamonds for the 2000/01 season. The weekly incidence of clinical influenza-like illnesses (ILI) per 10,000 inhabitants was monitored in the stable nation-wide sentinel system of general practices operated by NIVEL (Netherlands Institute for Primary Health Care, Utrecht, the Netherlands).
Figure 15.4 Influenza in the Netherlands, 1999/2000 season. Line: ILI per 10,000 inhabitants. Bars, from bottom to top: dark grey: type B, blank: A/H3, black: A/H1 and light grey: A/not subtyped. Source of the data for the incidence of ILI: see legend of Fig. 3. Source of the data for the numbers of influenza virus isolates: the National Influenza Centre of the Netherlands, Rotterdam, the Netherlands.
Figure 15.5 Number of influenzalike illnesses (ILI) per 10,000 inhabitants in the Netherlands from the 1969/70 through the 2002/03 season. Source of the data: see legend of Fig. 15.3.
virus 3 diseases for the spring or summer.2,10,11 In some countries there was evidence for mutual exclusion of parainfluenza viruses 1 and 2 on the one hand and parainfluenza virus 3 on the other9,11 (also de Jong, unpublished data). Earlier reports suggested mutual exclusion of parainfluenza viruses 1
and 2 viruses.10 These observations seem to imply interesting cases of interference of the cyclic variation of a virus disease by the circulation of another virus. There could be an immunological explanation for this phenomenon. Although the three parainfluenza virus types belong to two different
134
Severe Acute Respiratory Syndrome
nucleotide sequence clades of the Paramyxoviridae family, there are related antigenic determinants present on the H site of the HN proteins of the type 1, type 2, and type 3 viruses,12 which are responsible for the frequent occurrence in humans of heterotypic complement-fixing, haemagglutinationinhibiting or neutralizing antibody titre rises within the parainfluenza virus group.13
been reported for the epidemic adenoviruses 3, 4 or 7. For instance, in the 1960s the highest rates of respiratory adenovirus disease among military recruits in the USA, usually caused by the epidemic adenoviruses, were found in winter and spring. In other countries however, adenoviruses 3 and 7 epidemics have been described to be most frequent in summer.2,17,18
Respiratory syncytial virus (RSV) disease
Rhinovirus disease (common cold)
In temperate climates, RSV epidemics occur in the late autumn, winter, or early spring but are absent during the summer. The size, date of start and duration of the epidemics are very constant. In England and the USA the peak is in January, February or March2,14 and in the Netherlands in December or January. In contrast, high RSV disease rates occur annually in spring and summer in such tropical regions as Hong Kong and Singapore.5 The incidence begins to rise in March, peaks steeply in April and remains at a comparatively high level until September.
Human metapneumovirus disease The disease caused by the newly discovered human metapneumovirus (hMPV) displays the same seasonal pattern as most other respiratory virus diseases do and is mainly restricted to the winter months15 but in Hong Kong this virus has a summer seasonality.16
Respiratory adenovirus diseases Adenoviruses are circulating worldwide around the year. Adenoviruses causing respiratory tract disease are recovered throughout the year in Hong Kong. Decreased incidence is noted in this region in the months of August, September and October. In moderate climates, the so-called endemic adenoviruses (types 1, 2, 5 and 6) can be detected around the year.17 In contrast, sometimes marked seasonal prevalence has
In the USA, rhinovirus disease peaks in early autumn and spring.19,20 In the Netherlands, a peak among patients consulting a general practitioner is observed in September. Thereafter, the incidence decreases but remains fairly high until the end of the winter and in spring and summer, rhinoviruses are rarely detected in the Netherlands.
‘Classical’ coronavirus diseases Coronavirus diagnostics of the ‘classical’ human coronavirus types 229E and OC43 have not often been included in studies on respiratory virus diseases. In the USA, France and Finland, isolation and antibody titre rises were rarely reported outside the period from December until the end of May.21–23 In the Netherlands, these viruses were detected by PCR in the period from October until the end ofMay. In children, two peaks in late autumn to early winter and early summer were reported.24 In the USA, outbreaks of type 229E followed a 2–4-year cycle and outbreaks due to type OC43 occurred every other year when the incidence of 229E is low.21 Very rarely sizeable epidemics of types 229E and OC43 occurred in the same year, indicating an interference phenomenon similar to that observed with parainfluenza viruses (see above). Also the explanation of the interference could be similar. Although the two types belong to two different nucleotide sequence clades of the Coronaviridae family and the three antigens of the two types have been found to be unrelated to one another
The Seasonality of Respiratory Virus Diseases: Implications for SARS?
by crossed immunoelectrophoresis, children and adults infected with coronaviruses frequently show antibody responses to both viruses at the same time, thereby implying some antigenic relationship that has escaped the scrutiny of more rigorous tests.25 Whether this elusive antigenic relationship extends to SARS CoV remains to be seen. This possibility should be kept in mind when interpreting results of serologic diagnosis or seroepidemiological studies.
Seasonality of SARS Obviously, no seasonal pattern can be inferred for SARS at present. It is only possible to compare the dates of the reported detections of the illness with the seasonality of classical coronavirus disease. The first case of SARS was recorded on 16 November 2002 in Guangdong, where the outbreak peaked in February 2003 and disappeared in May of the same year. In Hong Kong the syndrome was first detected in February 2003, outbreaks occurred in March, diminished in magnitude in April and disappeared by June of the same year. While these observations are superficially compatible with a seasonal pattern similar to that of the classical coronavirus disease in temperate zones described above, the active case finding and isolation measures clearly contributed to the decline of SARS. Thus it is unclear if SARS CoV manifests a natural seasonality in humans. There is the interesting possibility of interference between SARS CoV and the classical coronaviruses, similar to that observed with the parainfluenza viruses 1–3 and the classical coronaviruses 229E and OC43 (see above).
Routes of transmission of respiratory virus infections Seasonality of a virus disease is the consequence of the seasonal variation of the transmissibility of the virus. A short description of the routes of transmission is therefore included in this chapter.
135
Respiratory virus infections are mainly contracted indoors. They spread by aerosols, large droplets or direct contact (hand contact, kissing), or indirect contact via fomites (shared inanimate objects such as door handles, taps, telephones and toys). Among small children, contact infection may be the most frequent route of transmission. Dissemination can occur by large droplets, effective only at distances of up to 1 or 2 metres that can be considered to be an extension of contact spread. Aerosol transmission can result from inhalation of tiny particles (‘droplet nuclei’) with a diameter of 1–5 m, which allows the virus to infect persons at distances of several metres and to start infection of the lower airways directly. The main route of transmission may depend on the epidemiological setting, such as target group (children, adults, elderly people), climate and weather, intensity of ventilation, level of hygiene, and location (family, school, hospital, etc.). It is therefore virtually impossible to establish which route is most important for a particular virus in the society at large. Nevertheless, on the basis of epidemiological observations and intervention and simulation experiments, some speculations will be given below, mainly concerning adults.1,26 Influenza viruses can spread by contact but are thought to disseminate usually via large droplets or inhalation of an aerosol. Peak titres of influenza viruses in nasal discharge can reach 108 TCID50 (50% tissue-culture-infectious-doses) per ml and inhalation of as few as 3 TCID50 can result in infection. Parainfluenza viruses and RSV spread by direct person-to-person contact or large droplets. Respiratory adenoviruses are transmitted like influenza viruses or by contact infection of the eyes resulting in pharyngoconjunctival fever. Rhinoviruses disseminate by contact or by large droplets. Classical coronaviruses may be transmitted by droplet nuclei or large droplets.21 SARS CoV usually disseminated only to close contacts such as those within families or in hospitals, indicating spread by large
136
Severe Acute Respiratory Syndrome
droplets or (indirect) contact.27 The virus has been detected in respiratory secretions, faeces, blood and urine and has been shown to survive for 1–4 days in faeces and for at least 36 hours on smooth surfaces at room temperature.27 These data indicate that fomites are able to transmit the virus.
Possible mechanisms of seasonality of respiratory diseases The mechanisms of seasonality of respiratory diseases are poorly understood. It should be noted that infection is not always followed by disease. The innate or specific host response can control the infection before it induces symptoms. The resulting limited infection is called subclinical, asymptomatic, inapparent, or abortive infection. Thus, 25% of influenza virus infections, 27% of parainfluenza virus infections of 0–5-year-old children and 89% of parainfluenza virus infections of 20–39-yearolds were estimated to be subclinical.14 Symptomatic persons are more likely to spread infection than asymptomatic persons, presumably because they shed more virus; this has been demonstrated for influenza virus.28 The proportion of subclinical infections may be season dependent and seasonality of disease could therefore differ theoretically from seasonality of infection. Important factors controlling the efficiency of the spread of respiratory virus infections (transmission) and the spread of disease (contagion) include: 1 The virus concentration in respiratory secretions 2 The duration of virus shedding 3 The viscosity of the secretions 4 The frequency and intensity of coughing, sneezing, and brushing teeth29 5 The ability of the virus to survive in aerosols or on the surface of hands or fomites 6 The level of hygiene 7 The intensity of ventilation 8 The route of infection
9 The minimal infectious dose 10 The innate and specific immunity of the host 11 Social factors such as crowding and mobility of people. All these factors may be subject to seasonal influences. Factors 5 and 10 will be briefly discussed below.
Effect of humidity of virus survival An environmental factor that may play a role in the seasonality of virus infections is the relative humidity (RH) of the ambient air. Influenza virus has been shown to be stable in droplet nuclei aerosols with a low RH (below 45%) and rapidly inactivated in humid aerosols with an RH above 50% .30,31 In the same experimental conditions, it was shown that poliovirus has a reverse pattern in its RH sensitivity, being stable in aerosols with an RH over 50% and rapidly inactivated at an RH under 45%.30 Similar results were obtained for these viruses after drying on solid surfaces.32 Epidemiologically, in temperate climates in both hemispheres influenza has a predilection for the winter, when the RH indoors is usually lower than 45%, whereas poliomyelitis also mirrored this seasonality and caused — before widespread vaccination — epidemics during the summer, when the RH indoors is usually higher than 50%. Interestingly, similar to influenza virus, coronavirus 229E is most stable in aerosols of low RH and has a predilection for the winter.33 The same type of seasonality may apply to SARS in temperate zones. However, in tropical areas, influenza and RSV disease do not display this type of seasonality, nor a predilection for periods with a relatively low RH. If RH is a seasonal factor at all, it is in these regions overruled by other factors.
Innate and specific immunity of the host Some evidence has been obtained that the activity of immunological systems and the host susceptibility to certain infectious dis-
The Seasonality of Respiratory Virus Diseases: Implications for SARS?
eases are influenced by the daily light/dark cycle, an effect which may be mediated by the pattern of melatonin secretion.34 With some viruses, notably influenza viruses, antigenic variants emerging in some part of the world may extensively but unnoticeably spread to other regions during interseasonal periods, only to start an epidemic when the appropriate conditions in that region are present. This theory of ‘seeding’ of the virus can explain the occurrence of sporadic cases and even outbreaks of influenza during the summer35 and the often almost simultaneous onset of influenza epidemics in large areas (e.g. North America, Europe). Person-to-person transmission is too slow to account for this phenomenon. Although the epidemic appears to spread, in this theory the virus does not have to because it is already present at an imperceptible level in the entire area prior to the rise in influenza activity.34,36
137
sonal variations in the dietary and cultural habits of persons partaking of exotic wild game foods. In addition, seasonal factors may affect how efficiently isolated instances of zoonotic transmission may initiate selfsustaining chains of human transmission as occurred in late 2002. It is still unknown whether cooler weather promotes the transmission of SARS coronavirus in humans as it does with most respiratory viruses, including the classical coronaviruses. In December 2003 and January 2004, SARS did re-emerge in Guangdong with four persons in the community acquiring infection presumably from an animal source. However, there was no transmission to others.38
Acknowledgement The authors are grateful for the comments and technical help from Dr R.A.M. Fouchier.
Will SARS return? The extensive measures of case finding, infection control precaution, contact tracing and quarantine taken by the various countries involved in the worldwide outbreak of SARS contributed substantially to the disappearance of the disease in June. This success was made possible by the low transmissibility of the virus. The basic reproductive number was estimated at 2–4 and only a few secondary cases occurred when symptomatic cases were isolated within 5 days of illness27 (Chapters 13 and 14). The possibility of the return of SARS from a human ‘reservoir’ depends on the proportion of asymptomatic cases and their infectiousness. Asymptomatic infection has been shown to be rare27,37 and persistence in the human population is unlikely. However, it could possibly be reintroduced from an animal reservoir (Chapter 11). In the instance of SARS, where re-emergence is likely to arise from an animal source, there is the additional factor of seasonality of the precursor virus in animals and also the sea-
References 1 Evans AS, ed. Viral Infections of Humans: Epidemiology and Control, 3rd edn. New York: Plenum Medical Book Company, 1989. 2 Noah ND. Cyclical patterns and predictability in infection. Epidemiol Infect 1989;102: 175–90. 3 Nokes DJ, Anderson RM. Mathematical models of infectious agent transmission and the impact of mass vaccination. Rev Med Microbiol 1992;3: 187–95. 4 Assaad F, Cockburn WC, Sundaresan TK. Use of excess mortality from respiratory diseases in the study of influenza. Bull WHO 1973;49: 219–33. 5 Chew FT, Doraisingham S, Ling AE et al. Seasonal trends of viral respiratory tract infections in the tropics. Epidemiol Infect 1998;121: 121–8. 6 Dosseh A, Ndiaye K, Spiegel A et al. Epidemiological and virological influenza survey in Dakar, Senegal: 1996–1998. Am J Trop Med Hyg 2000;62: 639–43. 7 Potter CW. Chronicle of influenza pandemics. In: Nicholson KG, Webster RG, Hay AJ, eds. Textbook of Influenza. Oxford: Blackwell Science, 1998: 3–18. 8 Nguyen-Van-Tam JS. Epidemiology of influenza. In: Fields BN, Knipe DM, Howly PM, eds.
138
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Severe Acute Respiratory Syndrome
Fields Virology. New York: Raven Press, 1985: 207–16. Glezen WP, Loda FA, Denny FW. Parainfluenza Viruses. In: Evans AS, ed. Viral Infections of Humans: Epidemiology and Control, 3rd edn. New York: Plenum Medical Book Company, 1989: 493–507. Glezen WP, Denny FW. Epidemiology of acute lower respiratory disease in children. N Engl J Med 1973;288: 498–505. Knott AM, Long CE, Hall CB. Parainfluenza viral infections in pediatric outpatients: seasonal patterns and clinical characteristics. Pediatr Infect Dis J 1994;13: 269–73. Chanock RM, McIntosh K. Parainfluenza viruses. In: Fields BN, Knipe DM, Howly PM, eds. Fields Virology. New York: Raven Press, 1985: 1241–53. Chanock RM. Parainfluenza viruses. In: Lennette EH, Schmidt NJ, eds. Diagnostic Procedures for Viral, Rickettsial and Chlamydial Infections. Washington: American Public Health Association, 1979: 611–32. Monto AS, Sullivan KM. Acute respiratory illness in the community: frequency of illness and the agents involved. Epidemiol Infect 1993;110: 145–60. Osterhaus A, Fouchier R. Human metapneumovirus in the community. Lancet 2003;361: 890–1. Peiris JS, Tong WH, Chan KH et al. Children with respiratory diseases associated with metapneumovirus in Hong Kong. Emerg Infect Dis 2003;9: 628–33. De Jong JC, Bijlsma K, Wermenbol AG et al. Detection, typing, and subtyping of enteric adenoviruses 40 and 41 from fecal samples and observation of changing incidences of infections with these types and subtypes. J Clin Microbiol 1993;31: 1562–9. Foy HM. Adenoviruses. In: Evans AS, ed. Viral Infections of Humans: Epidemiology and Control, 3rd edn. New York: Plenum Medical Book Company, 1989: 77–94. Gwaltney JM Jr. Rhinoviruses. In: Evans AS, ed. Viral Infections of Humans: Epidemiology and Control, 3rd edn. New York: Plenum Medical Book Company, 1989: 593–615. Couch RB. Rhinoviruses. In: Fields BN, Knipe DM, Howly PM, eds. Fields Virology. New York: Raven Press 1985: 795–816. Monto AS. Coronaviruses. In: Evans AS, ed. Viral Infections of Humans: Epidemiology and Control, 3rd edn. New York: Plenum Medical Book Company, 1989: 153–67. Aymard M, Chomel JJ, Allard JP et al. Epidemiology of viral infections and evaluation of the potential benefit of OM-85 BV on the virologi-
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
cal status of children attending day-care centers. Respiration 1994;61 Suppl 1: 24–31. Mäkelä MJ, Puhakka T, Ruuskanen O et al. Viruses and bacteria in the etiology of the common cold. J Clin Microbiol 1998;36: 539–542. Isaacs D, Flowers D, Clarke JR et al. Epidemiology of coronavirus respiratory infection. Arch Dis Child 1983;58: 500–3. McIntosh K. Coronaviruses. In: Fields BN, Knipe DM, Howly PM, eds. Fields Virology. New York: Raven Press, 1985: 1323–30. Musher DM. How contagious are common respiratory tract infections? N Engl J Med 2003;348: 1256–66. WHO. Consensus document on the epidemiology of SARS, Geneva 16–17 May 2003, http://www.who.int/csr/sars/en/WHOconsensus.pdf, WHO, October 2003. Fox JP, Kilbourne ED. Epidemiology of influenza — Summary of influenza workshop IV. J Infect Dis 1973;128: 361–86. Miller L. Studies of the aerobiology of dentistry. In: Hers JFP, Winkler KC, eds. Airborne Transmission and Airborne Infection. Utrecht: Oosthoek, 1973: 494–503. Hemmes JH, Winkler KC, Kool SM. Virus survival as a seasonal factor in influenza and poliomyelitis. Nature 1960;188: 430–1. Schulman JL, Kilbourne ED. Experimental transmission of influenza virus infection in mice. II. Some factors affecting the incidence of transmitted infection. J Exp Med 1963;118: 267–75. Buckland FE, Tyrrell DAJ. Loss of infectivity on drying various viruses. Nature 1962;195: 1063. Ijaz MK, Brunner AH, Sattar SA et al. Survival characteristics of airborne human coronavirus 229E. J Gen Virol 1985;66: 2743–8. Dowell SF. Seasonal variation in host susceptibility and cycles of certain infectious diseases. Emerg Infect Dis 2001;7: 369–74. Kohn MA, Farley TA, Sundin D et al. Three summertime outbreaks of influenza type A. J Infect Dis 1995;172: 246–9. Hope-Simpson R, Golubev D. A new concept of the epidemic process of influenza A virus. Epidemiol Infect 1987;99: 5–54. Lim WL. Presentation at the epidemiology breakout session, WHO Global Conference on SARS, Kuala Lumpur, Malaysia, 17–18 June 2003. Liang G, Chen Q, Xu J et al. Laboratory diagnosis of four recent sporadic cases of community acquired SARS, Guangdong province, China. Emerg Infect Dis 2004;10: 1774–81.
Chapter 16
Public Health Response: A View from Singapore Chorh Chuan Tan
Introduction The SARS outbreak in Singapore was an unprecedented event in the country’s history. As was the case for China, Hanoi, Hong Kong, Toronto and Taiwan, SARS surfaced ‘out of the blue’ in Singapore and swiftly developed into a national crisis. It was a severe test not only of the public health and medical systems but also of the government and people of Singapore. The rapid containment of the epidemic was the result of the government’s strong leadership, the dedication and professionalism of healthcare workers, the early involvement and co-ordinated response of multiple governmental and other agencies, and the strong support of the people of Singapore. This chapter describes the Singapore experience in containing SARS and covers: 1 The epidemiology, transmission pattern and main phases of the SARS outbreak in Singapore. 2 The formulation of Singapore’s public health policies and how containment strategies were shaped by three key public health lessons learnt in the early part of the outbreak. 3 The containment of the outbreaks in hospitals. 4 The management of the Pasir Panjang Wholesale Market (PPWM) outbreak. 5 Effectiveness of the isolation and quarantine measures, and temperature screening of travellers at the airport and seaports.
6 Key challenges in controlling the epidemic and how these were managed. 7 The steps taken to enhance preparedness against future outbreaks.
Abbreviations MOH, Ministry of Health, Singapore NUH, National University Hospital PPWM, Pasir Panjang Wholesale Market SGH, Singapore General Hospital TTSH, Tan Tock Seng Hospital
Description of the SARS outbreak in Singapore The index case of the Singapore outbreak The SARS outbreak in Singapore had its beginnings when two pairs of Singaporeans travelled to Hong Kong as tourists between 20 and 25 Feb 2003. Three of these individuals developed fever and respiratory symptoms and were admitted to two hospitals in Singapore between 1 and 3 March 2003. However, further transmission of infection only occurred with one of these three patients, Patient A. Patient A returned to Singapore on 25 February and developed fever and dry cough on the same day and was then admitted to a general hospital, Tan Tock Seng Hospital (TTSH), on 1 March 2003, was isolated on 6 March 2003 and was only categorized as a SARS case on 15 March 2003. The mystery of how Patient A
139
140
Severe Acute Respiratory Syndrome
acquired the infection was only solved later when it emerged that Patient A had stayed on the same floor of hotel M in Hong Kong as the Guangdong doctor who was the index case for the large outbreaks in Hanoi, Hong Kong and Toronto.1
Epidemiology of the Singapore SARS outbreak The Singapore SARS outbreak lasted about 11 weeks — the index case, Patient A, was admitted on 1 March 2003 and the last case of the outbreak had onset of symptoms on 5 May and was isolated on 11 May 2003. WHO removed Singapore from its list of countries with local transmission of SARS on 31 May 2003. During this period, a total of 206 probable SARS cases were diagnosed on the basis of the WHO case definition.2 With the availability of tests for the SARS coronavirus, the WHO case definition was revised on 1 May 2003 to include laboratory diagnosis as a criterion for defining a probable SARS case.3 Locally validated PCR and serology tests for the SARS coronavirus were available in Singapore about 5 April and 2 May 2003 respectively, and were used to assist in the diagnosis of all patients admitted for possible, probable and suspect SARS specimens. In the post-epidemic period, the Singapore Ministry of Health (MOH) conducted a review of all the suspect and observational cases admitted to TTSH. This study was completed on 15 July 2003 and identified 32 additional SARS antibody or culture positive patients among 600 admitted to TTSH as suspect SARS cases. These 32 patients had onset of illness in March or April; none were noted on public health investigations to have transmitted to others. A further 700 patients admitted for observation for possible SARS all tested negative for SARS antibodies. In line with WHO guidelines,3 the 32 suspect cases with positive serological results were retrospectively reclassified as probable
SARS cases on 15 July 2003, bringing the total number of probable SARS cases to 238. Figure 16.1 gives the epidemic curve for all 238 cases. Table 16.1 gives the basic characteristics of these cases, i.e. age, gender, health-care worker status, and whether infection was acquired in the hospital. Thirty-three patients died of SARS including five health-care workers.
Non-outbreak-associated case One additional case in Singapore was identified after the outbreak. On 3 September 2003, a doctoral student working on West Nile Virus was admitted with fever and myalgia to the Singapore General Hospital (SGH). He was placed in isolation and tested for SARS CoV infection. On 8 September, stool and sputum specimens tested positive for SARS coronavirus by PCR testing. Serial serum samples showed seroconversion to the SARS coronavirus. Investigations by an external expert panel revealed that the patient acquired SARS due to accidental contamination in the laboratory where he worked, which also handled SARS viral isolates. There was no evidence of secondary transmission from this case.4
Transmission patterns During the SARS outbreak, public health lessons were drawn from and containment measures were formulated based on the original 206 probable SARS cases diagnosed using the initial WHO case definition. The discussions that follow will be based on these original 206 cases. The transmission pattern of the 206 probable SARS cases is summarized schematically in Fig. 16.2 and includes 6 clusters of cases all linked to Patient A. The first cluster of 109 patients at TTSH started on 1 March 2003 when patient A was admitted to TTSH. The cluster at OV nursing home was initiated by a nursing home patient who was briefly admitted to TTSH on 16–17 March. This patient was later admitted to Changi
Public Health Response: A View from Singapore
141
16 14
Number of cases
12 10 8 6 4 2
ay 6M
29 -A pr
22 -A pr
15 -A pr
8Ap r
1Ap r
25 -M ar
18 -M ar
11 -M ar
4M ar
25 -F eb
0
Date of onset
newly reclassified
originally classified
Figure 16.1 Epidemic curve for Singapore SARS outbreak. 206 cases (in lighter shade) were diagnosed during the outbreak based on the WHO case definition while 32 suspect cases (in darker shade) were subsequently re-categorized as probable cases based on positive serological results.
General Hospital on 25 March 2003. A cluster of 53 patients at SGH can be traced to Patient D who was an in-patient at TTSH and was discharged on 20 March and then admitted to SGH on 24 March for gastrointestinal bleeding. Patient D’s SARS infection had a highly atypical presentation and the diagnosis of SARS was not suspected. This missed diagnosis resulted in spread in SGH and the cluster of 53 cases. The social group cluster of seven cases occurred because an infected SGH health-care worker was not recognized initially as having SARS, was granted medical leave, and transmitted infection to a group of friends. Patient D also was the indirect source of infection for the PP Market Cluster of 12 cases. Patient D’s brother visited him in SGH on 31 March, contracted infection, and transmitted to others through his work as a vegetable hawker in Pasir Panjang Wholesale Market (PPWM). Patient D’s brother worked for
brief periods during the time he was unwell. Later, patient D’s brother was admitted to the National University Hospital (NUH), was not suspected as having SARS, and transmitted to others in NUH causing a cluster of 10 cases in that hospital. The index case, patient A, and these 6 clusters accounted for all but 8 of the original 206 probable cases. Of these eight unaccounted for cases, six acquired infection outside of Singapore, i.e. were imported cases, and the other two had no history of contact with another SARS patient despite intensive epidemiological investigation.
Main phases of the Singapore SARS outbreak and containment The course of the SARS outbreak in Singapore was dynamic and influenced by both transmission patterns and the public health response to the outbreak. Since the public
142
Originally classified (n = 206)
Retrospectively reclassified suspect cases (n = 32)
Combined cases (n = 238)
Male (n = 71)
Male (n = 6)
Female (n = 135)
Total (n = 206)
Female (n = 26)
Total (n = 32)
Male (n = 77)
Female (n = 161)
Total (n = 238)
Age group 0–4 5–14 15–24 25–34 35–44 45–54 55–64 65+
2 0 8 12 20 11 10 8
3 2 27 39 23 18 14 9
5 2 35 51 43 29 24 17
0 0 0 1 1 0 1 3
0 3 2 14 1 4 0 2
0 3 2 15 2 4 1 5
2 0 8 13 21 11 11 11
3 5 29 53 24 22 14 11
5 5 37 66 45 33 25 22
HCW Non-HCW
13 58
71 64
84 122
0 6
13 13
13 19
13 64
84 77
97 141
Acquired in hospital Number %*
46 66%
16 64%
20 65%
50 66%
109 84%
*index cases have been taken out from the denominators. HCW: Health-care worker.
155 78%
4 67%
125 81%
175 76%
Severe Acute Respiratory Syndrome
Table 16.1 (a) Details of probable SARS cases in the Singapore outbreak. 206 cases were diagnosed during the course of the outbreak, based on the WHO casedefinition. A further 32 suspect cases admitted to TTSH in March or April 03 were subsequently re-categorized as probable SARS cases based on positive serology results
Table 16.1 (b) Deaths due to SARS Originally classified (n = 206)
Female (n = 14)
Total (n = 32)
Male (n = 1)
Female (n = 0)
Total (n = 1)
Male (n = 19)
Female (n = 14)
Total (n = 33)
Age group 0–4 5–14 15–24 25–34 35–44 45–54 55–64 65+
0 0 1 2 4 2 4 5
0 0 0 2 1 5 2 4
0 0 1 4 5 7 6 9
0 0 0 0 0 0 0 1
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 1
0 0 1 2 4 2 4 6
0 0 0 2 1 5 2 4
0 0 1 4 5 7 6 10
HCWs Non-HCWs
3 15
2 12
5 27
0 1
0 0
0 0
3 16
2 12
5 28
Public Health Response: A View from Singapore
Male (n = 18)
Combined cases (n = 238)
Retrospectively reclassified suspect cases (n = 32)
143
144
Severe Acute Respiratory Syndrome
Patient zero
TTSH cluster
Imported/unlinked 6+2
109
OV nursing home SGH cluster
Social group cluster
7 NUH cluster
7
53
PP market cluster
10
health response was multifaceted and occurred at different times during and after the outbreak, it is useful to consider the outbreak and the immediate post-outbreak period in terms of five main phases, as illustrated in Fig. 16.3. In the first phase, from 1 March to 15 March, the Singapore outbreak was developing and some, albeit incomplete control measures were taken with some patients. With the global alert from WHO on 13 March and the issuance of the first casedefinition of SARS on 15 March, public health control efforts were pursued in earnest in Singapore and elsewhere. The second phase, from 15 March to 30 April, can be characterized as containing outbreaks in hospitals and other healthcare facilities, the main settings of disease transmission (Fig. 16.2).5 Although stringent containment measures continued in hospitals and health-care facilities until July 2003 and beyond, Phase 2 was considered to extend only to 30 April, since the last two probable SARS cases were isolated on 28 April and 5 May respectively (Fig. 16.1). Phase 3, 19 April to 4 May, is defined as the period used to contain the PPWM cluster of cases. This cluster was one of the few anywhere in the world, of more widespread community SARS transmission. Phase 3 overlaps with Phase 2, and extends from 19 April to 4 May.
12
Figure 16. 2 SARS transmission pattern for 206 probable SARS cases in the Singapore outbreak (excludes 32 suspect cases subsequently reclassified as probable cases based on serological testing).
Phase 4, from May 1 to July 15 involved a period of very intense active case-finding. This effort was driven by the desire to ensure that there were no undetected chains of SARS transmission in patients with chronic medical conditions who had had undetected exposure to SARS in SARS-affected hospitals and atypical clinical presentations of the infection. This concern was heightened in late May 2003 by the resurgent SARS outbreak in Toronto.6 During this phase, MOH public health physicians monitored for and investigated fever clusters in all the nursing homes in Singapore over a 2-month period, looking for possible SARS cases. Hospitals also monitored for and investigated suspicious clusters of febrile patients in their wards. At the same time, TTSH, SGH and NUH reviewed a total of 10 000 patients who had been in their hospitals during and immediately after the time of active SARS transmission. Patients were telephoned to check if they or their household contacts had any fever or symptoms suggestive of SARS and some were recalled for medical review. Many patients, particularly those in TTSH, had repeated coronavirus testing (PCR or serology) prior to discharge and these results were reviewed. The case records of patients who had been through ‘hot wards’ in TTSH or SGH were reviewed to double-check for possible contact with known probable SARS cases. Despite this intensive exercise, no
Public Health Response: A View from Singapore
145
#1: Outbreak brewing #2: Hospitals’ outbreak containment #3: PPWM [IMH scare]
1
-
15/3
19/4
30/4
11/5
31/5
15/7
#4: Intensive case-finding #5: Enhanced surveillance Systems consolidation & building Figure 16.3 Phases of the Singapore SARS outbreak and public health control.
fresh or atypical SARS cases or clusters were uncovered. During this phase, there was one SARS scare. A cluster of patients and staff in the Institute of Mental Health in Singapore were reported to have fever and respiratory symptoms on 13 May 2003. Full containment measures were implemented but fortunately, the cause of the outbreak was ascertained to be influenza. In Phase 5, from 15 July onwards, the most stringent public health control measures were gradually stopped and replaced by an ongoing heightened level of preparedness. An enhanced surveillance system for new cases and system for the detection and response to SARS were developed and/or co-ordinated. Contingency planning and preparedness exercises were carried out and continue to the present time.
Singapore’s public health containment policies and response Today, so much is known about SARS and the coronavirus responsible for it, that it is easy to forget that in March and early April 2003, very little was understood about SARS: the spectrum of clinical presentations, the infectiousness of the new disease, the effectiveness of various personal protective equipment and the identity of the causative agent. In particular, the lack of a reliable diagnostic test greatly hampered control measures. Consequently, public health containment measures were implemented based on the limited information available and rapidly adjusted as new information emerged and as the outbreak twisted and turned in its evolution. Dr
146
Severe Acute Respiratory Syndrome
David Heymann, then WHO Executive Director for Communicable Diseases, captured this well when he likened the process to ‘building the ship as we sailed’.
Singapore’s ‘prevent-detect-isolatecontain’ strategy To contain the SARS epidemic and prevent fresh clusters of infection, Singapore adopted what it termed its ‘preventdetect-isolate and contain’ strategy. The objectives were to prevent SARS from spreading to unaffected health-care facilities and areas, and to prevent fresh clusters from arising from new imported SARS cases; to detect suspicious cases early and to isolate them; and to contain clusters of infection quickly through swift epidemiological investigation and quarantine of infectious patients and contacts. This containment strategy was based on a number of critical assumptions, which fortunately proved to be largely correct, namely: 1 SARS patients are infectious only when they develop symptoms, of which fever is the most important marker. The corollary is that there is little or no significant asymptomatic SARS infection and transmission. 2 SARS is transmitted predominantly through close contact.
Three key public health lessons from Phase 1 of the outbreak Three key public health lessons learnt during the first phase of the outbreak had a great impact on MOH’s public health containment policies.
and transmitted in the health-care setting, particularly in hospitals and nursing homes. In Singapore, of the 13 cases notified to MOH by 15 March, 7 were healthcare workers. Of the 206 probable cases during the whole outbreak, 155 (78%) were acquired in hospital including 84 healthcare workers. This assumption had two very important implications. Firstly, SARS could rapidly degrade the ability of the country’s hospital sector to provide essential clinical services. Secondly, hospitals were acting as major amplifiers of SARS infection. This would accelerate the course of an outbreak and increase the risk of spillage into the community.
‘Superspreading incidents’ can singlehandedly result in or markedly worsen a SARS outbreak For reasons that are still unclear, some SARS patients appear to be very efficient transmitters of infection. Whether this is due to late recognition of the SARS patient coupled with inadequate infection control practices, or to differences in transmission mode and viral shedding in some patients, a single superspreading incident can be the source of a major outbreak. In Singapore, five patients accounted directly for infection in 121 of the 206 probable SARS cases. As summarized in Table 16.2, three superspreading incidents (Patients A, B and C) were responsible for the rapid escalation of the TTSH outbreak, one superspreading incident (Patient D) caused the SGH outbreak while the last superspreading incident (Patient E) accounted for the PPWM and NUH clusters.
SARS is predominantly a nosocomial infection From the outset, SARS appeared to be predominantly a nosocomial infection. This has been borne out by the subsequent experience in Singapore and elsewhere, where apart from the Amoy Gardens outbreak in Hong Kong, SARS has mainly been acquired
SARS patients with atypical clinical presentations pose the greatest public health threat The WHO case-definition for probable SARS requires the presence of fever (>38°C), respiratory symptoms and the presence of a posi-
Public Health Response: A View from Singapore
147
Table 16.2 Singapore’s five superspreaders Patient
Date of symptom onset
Directly linked SARS cases
Note
A B C D E
25 Feb 03 7 Mar 03 12 Mar 03 26 Mar 03 5 Apr 03
22 21 26 40 12 121
Index case for TTSH cluster TTSH cluster TTSH cluster Index case for SGH cluster Index for NUH & PPWM clusters
8
70
7
Primary Secon dary S/P
6
50
Figure 16.4 Number of primary cases (lighter shade) by time from symptom onset to isolation, number of secondary cases infected by such cases (darker shade) and mean number of secondary cases per primary case. (Figure reproduced from Science 2003;300:1966–70.)
Cases
5 40 4 30 3 20
2
10
1
0
Secondary cases/primary case
60
0 0
tive contact or travel history and signs of pneumonia or respiratory distress syndrome on chest X-ray. The early experience with SARS already indicated that the initial presentation of cases is much broader and non-specific. For example, in patients in the Singapore outbreak who eventually developed probable SARS, about 30% did not have respiratory symptoms initially and a significant proportion presented with other symptoms such as diarrhoea. The certainty of clinical diagnosis of SARS increases with the duration of observation but this is obviously not acceptable from the public health point of view. Atypical presentations of SARS were a major factor contributing to the continued propagation of the Singapore SARS epidemic. As shown in Fig. 16.4, late detection and isolation of a SARS patient, more than 8 days from the time of onset of symptoms,
1
2 3 5-6 7-8 4 Time from onset to isolation (days)
>8
is associated with a markedly increased risk of secondary transmission.7 Patients with atypical clinical presentations of SARS are very difficult to recognize and diagnose. It is therefore easy for them to be missed until late in the course of illness.
Superspreading incidents and atypical SARS cases had a profound impact on Phase 1 of the Singapore outbreak Figure 16.5 summarizes the key events in Phase 1 of the Singapore outbreak. On 6 March 2003, two hospitals notified MOH of three cases of atypical pneumonia who had travelled to Hong Kong in late February. On the same day, MOH had been informed by WHO that several health-care workers looking after an American patient with pneumonia in Hanoi, had developed a similar respiratory illness. In view of this, MOH alerted the two Singapore hospitals to
148
Severe Acute Respiratory Syndrome
3 Singaporean tourists to Hong Kong admitted to 2 hospitals in Singapore MOH Notified
19/2/03
11-3/3
6/3
‘Bird flu’ outbreak reported in Hong Kong; WHO calls for heightened global surveillance
MOH SARS taskforce formed
MOH notified of 6 SARS cases
13/3
MOH notified of 13 SARS cases
15/3
WHO – global alert new atypical pneumonia
WHO – SARS Case definition Travel advisory
Figure 16.5 Key events in Phase 1 of the Singapore outbreak.
isolate the three patients, to maintain strict infection control measures and to start contact tracing. Based on the information available at the time, it was thought that these were most likely cases of ‘bird flu’.8 On 13 March, WHO’s global alert about cases of atypical pneumonia in hospitals in Hong Kong and Hanoi sounded the alarm. In response, MOH alerted all hospitals and doctors to look out for patients with pneumonia who had recently travelled to Hong Kong, Hanoi or Guangdong province. MOH also advised the public that persons returning from these areas should seek medical attention immediately if they developed flu-like symptoms. By 15 March, MOH had been notified of a further 13 cases of pneumonia in Singapore, including 7 hospital staff who had attended to Patient A. Recognizing the seriousness of the situation, MOH convened a Taskforce chaired by the Director of Medical Services Figure 16.6 illustrates the profound impact which three superspreading incidents had on Phase 1 of the outbreak.
Patient A was admitted to TTSH on 1 March 2003. Looking back, by the time Patient A was isolated on 6 March, infection had been transmitted to 22 individuals who eventually developed probable SARS, including 10 health-care workers, 2 in-patients and 10 visitors. One of the infected health-care workers, Patient B, was admitted to TTSH for suspected dengue fever on 10 March. Before Patient B was isolated on 13 March, 20 other contacts had been infected. Hence, by the time the WHO global alert was raised, 2 superspreading incidents alone had directly transmitted infection to 42 contacts. One of the contacts infected by Patient B was Patient C, the first atypical SARS case encountered. Patient C had been admitted to TTSH on 10 March with fever and pneumonia on a background history of diabetes and ischaemic heart disease. Blood cultures grew gram-negative bacteria. Her condition deteriorated on 12 March and she was transferred to the coronary care unit with a clinical diagnosis of worsening communityacquired pneumonia and heart failure. She
Public Health Response: A View from Singapore
1 March
6 March
1 health-care worker (Patient B) 9 health-care workers
Patient A admitted to TRSH
13 March
1 inpatient (Patient C – SARS with atypical pressentation)
149
20 March
26 contacts
20 contacts
2 inpatients
10 visitors
Figure 16.6 Impact of three superspreaders (shaded boxes), of whom 1 had a highly atypical clinical presentation, on Phase 1 of the Singapore outbreak. Only secondary transmission from superspreaders is shown.
required mechanical ventilation and was only isolated on 20 March. This patient turned out to have concomitant SARS infection which had been clinically ‘masked’ by her other severe medical problems. By the time she was isolated, she had already transmitted infection to 26 contacts.5
Impact of these three public health lessons on containment strategies These three public health lessons influenced the ‘prevent-detect-isolate and contain’ strategy in the following ways: (a) Because hospitals, and nursing homes, were the major battlefields in the war against SARS, rapid containment of hospital SARS transmission was the key to control of the SARS epidemic. Stringent measures to prevent and detect SARS cases in hospitals were vital to prevent fresh outbreaks in previously unaffected facilities. Hence, within the framework of its overall control strategy, Singapore placed particular emphasis and resources on the prevention, detection and containment of hospital and nursing home clusters of SARS infection. This decision was particularly important in March and April 2003 when
human resources, containment systems and critical supplies such as N95 masks, were very stretched and prioritization was essential. (b) The realization that a single superspreading incident can cause a large outbreak of SARS and that atypical presentations of SARS can occur in patients with other chronic medical conditions, led to the adoption of a very ‘wide-net’ surveillance, isolation and quarantine policy. The intention was to detect all suspicious cases as early as possible and to isolate them. To achieve this, a very broad definition for suspicious cases was adopted, over and above the WHO case-definition for suspect and probable SARS, as follows: • Any health-care worker with fever and/or respiratory symptoms. • Patients with atypical pneumonia for which the cause had not been determined. • Clusters of health-care workers or patients with fever, especially in a healthcare setting. • Unexplained respiratory deaths. • Patients with fever >38°C without other causes and which the attending clinician assesses is suspicious of SARS.
150
Severe Acute Respiratory Syndrome
Such cases had to be isolated and investigated and transferred to TTSH if necessary. Contacts of such cases would be traced and placed on daily telephone surveillance or home quarantine depending on the specifics of each case. (c) Special measures were taken to reduce the risk of outbreaks due to patients with atypical SARS presentations, as discussed below. The fear of spread through undetected atypical cases was also the main reason for the intensive exercise mounted in Phase 4 of the outbreak.
Containment of the outbreaks in hospitals General measures applied to all hospitals and health-care facilities To contain the SARS outbreaks in the hospitals, a number of key public health measures were implemented in all hospitals and other health-care facilities. Control measures were added and made more stringent as the outbreak developed and as more information on SARS became available, as summarized in Table 16.3. The principal elements were prevention, early detection and rapid response.
Prevention 1 Triage at the point of first patient contact with the hospital (i.e. at emergency departments, outpatient clinics and direct ward admissions) to separate out febrile patients. Patients who required admission were transferred to TTSH if SARS was suspected or isolated if it was not. 2 Use of personal protective equipment (PPE) for staff and strict infection control measures. When TTSH was designated as the SARS hospital, all its staff used full PPE in all clinical areas. For other hospitals, healthcare workers initially used PPE in higher risk clinical areas only but from April onwards following the outbreak in SGH, full PPE was used in all clinical areas. 3 Limitations were imposed on visitors, in-patient transfers and re-admissions; and restrictions of practising rights of healthcare workers to single institutions.
Early detection of cases Each hospital centrally monitored the temperatures (twice to thrice daily) and sick leave of all staff. Hospitals with SARS transmission would refer any febrile staff to TTSH for isolation. All hospitals would look out
Table 16.3 Key measures introduced to contain the SARS outbreaks in hospitals, in Phase 2 of the Singapore epidemic Date
Guidelines and directives to hospitals and health-care facilities
16.3.03
Triage at Emergency Departments to separate out patients with fever and travel history; suspicious cases to be transferred to TTSH; strict infection control including N95 mask for handling cases Strict infection control measures in all high risk clinical areas Hospitals to monitor their staff centrally. Sick staff to be seen in hospital’s staff clinic only Detailed infection control guidelines for health-care workers issued
22.3.03
4.4.03
Other measures
TTSH designated as the SARS hospital MOH hotline for hospitals to check if a person is a contact Hospitals provided with updated lists of all contacts of suspect/ probable SARS cases
Public Health Response: A View from Singapore
151
Table 16.3 continued Date
Guidelines and directives to hospitals and health-care facilities
Other measures
8.4.03*
Directive enforceable under the PHMC Act requiring CEOs/managers of hospitals/nursing homes to ensure effective implementation of detailed procedures on triage and isolation, PPE** use and infection control in all clinical areas, handling of suspected SARS cases & their contacts and establishing an organizational structure to prevent and control SARS. Mandatory twice daily temperature monitoring of all health-care workers No transfers of patients from TTSH to nursing homes. Transfers from other hospitals only if patient is afebrile for >72 hours and has no contact history. Transferred patients required to be kept in isolation in nursing homes for further 10 days. Detailed procedures for handling bodies of patients who died of SARS. Requirement for double bagging and cremation within 24 hours (immediate burial for Muslims) SARS patients who have recovered put on 14 days home quarantine order on discharge
No visitors for suspect or probable SARS patients Visitors to non-SARS patients restricted to 2/day Contact details of all visitors recorded
8.4.03
15.4.03
17.4.03
20.4.03 22.4.03
29.4.03 1.5.03
12.5.03
Web-based system for hospitals/all doctors listing all contacts of SARS cases, persons on home quarantine orders, discharge history of patients in preceding month
Home quarantine orders issued for 10 days to non-SARS patients with chronic co-morbid conditions on discharge from SGH and TTSH. Other discharged patients without chronic conditions placed on phone surveillance for 14 days Precautionary measures and procedures for haemodialysis centres All doctors and health-care workers restricted to working in only 1 hospital Inter-hospital transfer of patients disallowed. Patients must be re-admitted to same hospital only, within 21 days of discharge. These rules do not apply in emergency situations or if special waiver is allowed by MOH No visitors allowed in all public hospitals Fever clinics with enhanced infection control measures set up at 4 public primary care polyclinics Expanded lab provision for dengue PCR testing to facilitate clinical differentiation of dengue fever from SARS
*From 8 April 2003 onwards, all directives were issued under the Private Hospitals & Medical Clinics Act. **PPE: personal protective equipment, i.e. test-fitted N95 mask, gowns, gloves; goggles if dealing with suspicious cases; powered air purified respirators (PAPR) for high risk procedures such as intubation.
152
Severe Acute Respiratory Syndrome
for clusters of more than two febrile staff who worked in the same clinical area.
Rapid response to possible SARS cases MOH issued standard operating procedures for handling suspected SARS cases and contacts. From 8 April 2003 onwards, all directives and guidelines from MOH were issued under the Private Hospitals and Medical Clinics Act or Infectious Diseases Act, so that they were legally enforceable. Frequent audits were carried out by MOH staff throughout the outbreak to ensure that directives were being implemented on the ground.
Designation of TTSH as the SARS hospital In addition to the general measures above, an initiative which proved very useful was the designation of TTSH as Singapore’s SARS hospital on 22 March. TTSH ceased all its non-SARS-related clinical work including that of its Emergency Department, which was the busiest in Singapore. Discharges of existing non-SARS patients were allowed if they had no known exposure to SARS but daily telephone surveillance was carried out for 14 days post-discharge. All cases in Singapore which could possibly be SARS were referred to TTSH for assessment and isolation either in TTSH itself or in the adjacent Communicable Diseases Centre. All health-care workers in TTSH adopted the use of N95 masks, gowns and gloves. A strict regime of thrice daily, monitored temperature surveillance of all staff was instituted to identify affected staff as early as possible and to isolate them immediately. The decision to designate TTSH as the SARS hospital was made for the following reasons: 1 To facilitate rapid containment of the outbreak which affected a number of wards in TTSH. 2 To allow rapid accumulation of clinical
experience and expertise in the diagnosis and clinical management of patients. In addition, the practices introduced and tested within TTSH such as the stringent temperature monitoring regime and the requirement that all staff must be test-fitted for N95 masks, provided the empirical evidence for these measures to be implemented nationally. 3 Transferring all possible SARS cases to TTSH would reduce the risk of the infection inadvertently taking root in other hospitals and health-care facilities, which could then better carry on with their normal operations. 4 To improve the accuracy and timeliness of clinical and public health information collation and flow. In particular, it would facilitate interviews of patients for the purpose of contact tracing and epidemiological investigation. The institution of control measures successfully contained the intra-hospital transmission of SARS in TTSH, with the date of onset of the last such case being 12 April 2003. With the benefit of hindsight, however, the spread of the Singapore epidemic may have been further reduced if more aggressive measures had been introduced in TTSH on 15 March, specifically if all discharges from the hospital were stopped and no visitors to the hospital allowed.9
Containment of the SGH outbreak The SGH outbreak was started by Patient D, who was an in-patient in TTSH from 5 March till 20 March 2003, where he had contact with a SARS patient. Patient D was re-admitted to an open ward (Ward 57) in SGH on 24 March 2003 for steroid-induced gastrointestinal bleeding, on a background history of chronic renal failure and diabetes. He developed fever on 26 March 2003 and E. coli was cultured from his blood. On 29 March 2003, he was transferred to an adjacent open ward (Ward 58) until 2 April 2003. He only developed chest X-ray changes of pneumonia and was recognized
Public Health Response: A View from Singapore
to have SARS on 5 April 2003. The SGH outbreak was identified on 4 April 2003 when a cluster of 13 febrile health-care workers from Wards 57 and 58 was detected. The containment strategy was the removal in toto of all potentially exposed patients and health-care workers to TTSH on 5 and 6 April, even though none outside of Wards 57 and 58 had symptoms suggestive of SARS at the time. The ‘hot period’ was set as 24 March, when Patient D was admitted, to 5 April when the cluster was detected. The patients who were transferred to TTSH were 80 in-patients in Wards 57 and 58 on 5 April, and 135 patients who were in wards 57 and 58 during the ‘hot’ period but who had subsequently been transferred to 3 other wards in SGH. These three wards were ‘closed’ to all new admissions and discharges for 2 weeks. A total of 386 patients who had been in Wards 57 and 58 in the ‘hot’ period and who had been discharged were reviewed and found to be afebrile. All these patients were then put on thrice daily phone surveillance for 10 days. All health-care workers who worked in Wards 57 and 58 were transferred to TTSH. A further 236 SGH health-care workers who had been in contact with exposed staff from the two wards were put on 10-day quarantine. Contact tracing and home quarantine of visitors to Wards 57and 58 was also carried out. Stringent thrice daily temperature monitoring regime of all SGH staff was carried out and any staff with fever was referred to TTSH for assessment. This containment strategy resulted in control of the SGH outbreak within 10 days, without the need to close SGH, the largest public hospital in Singapore, down. Of the potentially exposed patients and health-care workers transferred to TTSH, eight persons later developed probable SARS. However, all the eight were detected early in TTSH before any secondary transmission from them occurred. Two important lessons were drawn from the early phase of the SGH outbreak. Firstly, the risk posed by atypical SARS cases was
153
again underlined. In view of this, MOH imposed a requirement under the Infectious Diseases Act for all in-patients with chronic medical conditions in TTSH and SGH to be placed on home quarantine orders for 10 days on discharge (Table 16.3). Secondly, the fact that Patient D had recently been discharged from TTSH had not been appreciated by the SGH staff who did not therefore elicit the history of contact with SARS. In response, MOH set up a web-based system, SARS-Web, which allowed all hospitals and doctors 24-hour access to daily updated lists of suspect and probable SARS cases, contacts of SARS patients and individuals on home quarantine orders, as well as the hospitalization record of patients in the preceding 4 weeks. MOH also imposed a requirement for all patients to be readmitted to the same hospital within 21 days of discharge (Table 16.3).
Containment of the Pasir Panjang wholesale market (PWM) outbreak The PPWM outbreak represented the first significant threat of widespread community spread of SARS. Patient E who was responsible for the NUH outbreak, worked in the PPWM for a few hours each day on 5, 7 and 8 April. On 19 April, two SARS patients linked to PPWM were diagnosed — a worker in PPWM and a taxi-driver who had ferried Patient E to work. In view of this, a major containment effort was launched. The contact tracing centre in MOH, augmented by 200 additional personnel from the People’s Association, traced a total of 1917 persons who worked in or frequented PPWM between 5 and 19 April. Home quarantine orders were served on these potential contacts. PPWM was also closed for 15 days, during which time the premises were thoroughly cleaned. As there was a possibility that workers or buyers at PPWM may have visited other wet markets in Singapore, the National Environment Agency coordinated a programme of daily tempera-
154
Severe Acute Respiratory Syndrome
ture monitoring of workers in wet markets throughout Singapore. In the end, the extent of spread of infection in PPWM was fortunately very limited. Twelve SARS patients were identified, of whom seven belonged to one family.
ongoing challenge, therefore, is how to be more specific and discriminatory in identifying contacts who are at risk, so that the numbers that need to be quarantined can be markedly reduced, without loss of effectiveness of the measure.
Effectiveness of isolation and quarantine measures and screening at the airport and seaports
Preventing the importation (and exportation) of SARS through temperature screening at the airport and seaports
Isolation of suspected SARS cases and quarantine of contacts Table 16.4 summarizes Singapore’s experience with isolation of suspected SARS cases and quarantine of contacts. The very ‘wide-net’ approach to surveillance and isolation of suspected cases resulted in progressively earlier isolation of probable SARS cases as the outbreak progressed. As shown in Table 16.5, the average duration between onset of symptoms to isolation for probable SARS patients was decreased from 6.8 days on week 2 of the outbreak, to 1.3 days at week 9. In parallel with this, the percentage of probable SARS cases who had been previously identified to be suspect cases progressively increased, being more than 80% for weeks 7, 8 and 9 of the outbreak. The percentage of probable SARS cases that had previously been on either home quarantine orders or on telephone surveillance also increased steadily, being 45%, 50% and 100% on weeks 7, 8 and 9 respectively. However, this degree of ‘effectiveness’ of early detection and isolation of probable SARS cases was achieved at the cost of quarantining very large numbers of individuals who eventually turned out not to have SARS. During the outbreak, 7863 contacts were served home quarantine orders while a further 4331 were put on daily telephone surveillance for 10 days. In total, 58 of the 206 probable SARS cases had been on either home quarantine orders or on telephone surveillance prior to their diagnosis. The
A major public health objective was to prevent outbreaks due to importation of new cases of SARS. On 14 March, MOH had already issued advice to the Singaporean public to avoid travel to SARS-affected countries unless absolutely necessary. On 30 March, with the worsening SARS situation in other parts of the world, Health Alert Notices were issued at the airport to inbound air passengers from affected areas. The Health Alert Notice highlighted the common symptoms of SARS and advised the passenger to seek immediate medical attention at TTSH should they have fever. On 31 March, nurses were stationed at the airport to screen passengers from all inbound flights from SARS-affected areas into Singapore. On 2 April, this was extended to Singapore’s seaports. On 9 April, passengers of all inbound flights were required to complete a Health Declaration Card, providing information on whether they had symptoms of SARS and the areas to which they had travelled in the preceding 10 days. Thermal scanners developed by Singapore’s Defence Science and Technology Agency were deployed at the airport. The thermal scanners allowed rapid temperature screening of all passengers on inbound flights. Those screened who had a temperature greater than 37.5°C had their temperatures rechecked by nurses, using aural thermometers. Those whose temperatures were elevated on retesting, were screened by the airport doctor and those who had suspicious symptoms were sent to TTSH by a dedicated
Table 16.4 Initial classification of possible SARS cases and the public health containment responses to be considered Clinical features
Observation for SARS (4) Deaths
As per WHO guidelines Atypical pneumonia
Nil
Fever (>38°C)
Positive travel history in preceding 10 days* Cluster of >2 cases in same work area in a health-care facility, home, same work area
Unexplained fever (>38°C)
Unexplained fever where some clinical suspicion of SARS Death on arrival to hospital due to pneumonia without an identifiable cause OR post-mortem findings of respiratory distress syndrome
Nil Nil
*Positive travel history: travel in the previous 10 days to countries listed by WHO with local SARS transmission. **HQO = home quarantine orders.
Public health action To initiate contact tracing as soon as notified HQO** to be served on all contacts To initiate contact tracing as soon as notified HQO to be served on all contacts To initiate contact tracing as soon as notified Daily phone surveillance for 10 days Health-care facility to initiate contact tracing and increase temperature surveillance over staff in area To notify MOH for decisions for further public health response To initiate contact tracing if case upgraded to suspect or probable SARS or observation (1) Attending doctor and pathologist to notify MOH To initiate contact tracing as soon as notified. MOH to decide on HQO
Public Health Response: A View from Singapore
Probable SARS Suspect SARS Observation for SARS (1) Observation for SARS (2) Observation for SARS (3)
Travel/contact history
155
156
Week of outbreak
Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 8 Week 9 Week 11
Period
24/2/03–2/3/03 3/3–9/3 10/3–16/3 17/3–23/3 24/3–30/3 31/3–6/4 7/4–13/4 14/4–20/4 21/4–27/4 5/5–11/5
Number of probable cases (n = 205)
2 15 39 33 18 44 31 14 8 1
Cases who were previously diagnosed as suspect SARS
Duration between onset of SARS symptoms and isolation in TTSH (days)
Number
%
Average
Minimum
Maximum
0 0 16 16 13 28 25 12 7 1
0.0 0.0 41.0 48.5 72.2 63.6 80.6 85.7 87.5 100.0
4.3 6.8 4.5 3.5 4.1* 2.9 2.2** 2.7 1.3 7.0
2 5 0 0 0 0 0 0 0
9 9 11 8 20 10 18 9 4
Number of cases previously on HQO
No. of cases previously on daily phone surveillance
Total no. of cases on HQO or surveillance
0 0 0 0 0 4 2 3 2 0
0 0 0 5 7 13 12 4 6 0
0 0 0 5 7 17 14 7 8 0
*Excludes one case which was retrospectively classified as probable SARS case after death. **Excludes two cases who died before admission to hospital and one SGH case transferred to TTSH on 7 April who developed onset of SARS symptoms on 12 April.
Severe Acute Respiratory Syndrome
Table 16.5 Probable SARS cases (n = 205 excluding index case) by week of outbreak, showing the numbers (%) who were previously diagnosed as suspect SARS cases, the duration between time of onset of SARS symptoms to isolation, and the numbers (%) who had previously been on home quarantine orders (HQO) or daily telephone surveillance
Public Health Response: A View from Singapore
ambulance service. On 23 April, the use of thermal scanners was extended to the screening of all departing passengers from Singapore’s airport, to prevent the possibility of inadvertent ‘exportation’ of SARS. Singapore has two major, and busy, land links with its neighbour Malaysia. On 23 April thermal imaging scanners were introduced at these land checkpoints and their use was progressively scaled up thereafter. These measures represent a key component of the overall SARS prevention strategy. However, the yield from these measures has been very low. As at 21 September 2003, 4044 travellers were found to have elevated temperatures (>37.5°C) through temperature screening at the airport and sea terminals, of whom 327 were referred to TTSH for assessment. Of these, 39 were admitted to TTSH but none turned out to have SARS. This low yield has also been reported by the authorities in Hong Kong, Singapore had six imported SARS cases, excluding the initial index case. Not one of these cases was picked up by screening at the airport but had presented to hospital subsequently when fever developed.
Key challenges in controlling the epidemic and how these were managed Leadership, command, control and co-ordination It is not easy to convey the atmosphere in which decisions had to be made during the SARS epidemic — we were breaking totally new ground, the stakes were high but information was often limited and it was hard to know how to interpret it. In addition, a very large number of agencies and personnel were involved in implementing control measures and in various supporting roles. Consequently, there was always a big risk of confusion and disorganized execution of public health containment operations. A very major challenge was ensuring that there was good communications
157
throughout — up and down the chain of command, with those personnel working on the ground, as well as the general public. Under these trying circumstances, strong leadership and effective command, control and co-ordination systems were critical for the containment of the SARS epidemic. In Singapore, the command, control and co-ordination systems were rapidly upgraded in response to the developing outbreak. A SARS Taskforce was set up by MOH on 15 March. The Taskforce, chaired by the Director of Medical Services, includes the clinical leadership and Chief Executive Officers of all hospitals, and infectious disease physicians and other experts. The focus of the Taskforce is on containment and prevention of SARS in the health-care sector. As such, it considers public health containment policies and helps drive their implementation in health-care facilities. The Taskforce also provides a regular forum to disseminate information, co-ordinate and resolve cross-hospital issues such as patient referrals, staff redeployment and logistics, and for health-care facilities to provide feedback to MOH. On 7 April, a Ministerial Committee on SARS was established to oversee the formulation and implementation of operational response plans for various scenarios that could arise, resolve cross-Ministry policy issues and give political guidance to handle the impact of SARS on the society and economy. The Committee is chaired by the Minister for Home Affairs and includes Ministers from Health, Education, National Development and Manpower. The Ministerial Committee provided strong leadership throughout the crisis and was the forum where strategic decisions were made and major initiatives and control measures were approved. The Executive Group chaired by the Permanent Secretary of the Ministry of Home Affairs and comprising the Permanent Secretaries of key Ministries provided overall guidance and co-ordination in the implementation of measures. An InterMinistry SARS Operations Committee
158
Severe Acute Respiratory Syndrome
chaired by MOH was responsible for working-level co-ordination of crossMinistry SARS issues and operations. A Ministerial SARS Combat Unit which included three Ministers of State who were medical doctors was set up on 20 April. The Ministerial Unit worked very closely with hospitals and other health-care facilities to contain existing outbreaks and to prevent fresh clusters of infection from arising. The appointment of this unit underlined the government’s commitment to put all necessary resources into the fight against SARS. The Ministerial Committee and SARS Combat Unit provided strong political leadership and rapid, strategic, decision making. In matters relating to SARS, MOH exercised direct ‘control’ over all healthcare facilities, through the use of the Infectious Diseases Act and Private Hospitals and Medical Clinics Act. This helped ensure that control measures in all health-care facilities were effectively implemented and verified through intensive audit. For issues and operations outside of the health-care setting particularly those involving multiple agencies, the Executive Group provided strong policy guidance and co-ordination.
Rapid, accurate information collation and flow for decision making To manage a complex and fast-moving epidemic, timely and accurate information is vital for informed decision making. This was a major challenge particularly in the earlier part of the outbreak. Typically, public health containment measures are triggered when patients are assessed by their attending clinicians to be possible SARS cases. However, because of the non-specific clinical presentation of SARS particularly early on in the course of illness and the lack of a good diagnostic test, different doctors could assess the same patient differently and the categorization of the patient may therefore change during the hospital admission. For example, TTSH clinicians developed their own system of
subclassifying possible SARS cases as ‘low suspect’ and ‘high suspect’ based on the clinical level of suspicion. Consequently, the public health responses triggered were sometimes inconsistent and prone to change. Another major challenge was that the flow of data on possible cases to MOH was slow and incomplete, and significant time and effort had to be spent to obtain and verify data. Information on contacts of SARS patients was also held in a separate database so integration of data was difficult. To address these issues, MOH implemented a uniform system of initial patient categorization for the purpose of triggering public health measures (Table 16.4). TTSH instituted a daily clinical meeting of senior clinicians which discussed all possible SARS cases which had been admitted and confirmed their initial clinical categorization. To facilitate epidemiological investigation and contact tracing, MOH posted a team of eight nurses led by a senior public health physician to TTSH, to interview suspected SARS patients. The clinical and epidemiological data were entered into a clinical database from which reports were generated for MOH’s daily epidemiology meeting. This was the forum where decisions on public health responses were made, and implementation of measures initiated (Fig. 16.7). The meeting also decided on the final categorization of patients as probable and suspect SARS cases when more clinical and laboratory data were available.
Meeting surge requirements A key problem during the SARS epidemic was the sudden huge surge in human resources requirements it created within and outside the health-care setting. In Singapore, the political leadership responded by committing additional personnel to assist MOH directly and by ensuring that other agencies were rapidly co-opted to manage situations outside the health-care sector.
Public Health Response: A View from Singapore
159
11 am review & categorization by senior physicians MOH team in TTSH interviews all patients
All suspicious cases admitted to TTSH
Fever clusters in hospitals, nursing homes; situational updates
Notifications by drs
MOH DAILY EPIDEMIOLOGICAL MEETING
MOH Contact Tracing Centre
Issuance of Home Quarantine Orders Phone surveillance
MOH Field Response Team
Contact tracing team in hospitals/schools/ army camps/etc.
Trace in 24 hours HQO same day
Figure 16.7 Daily MOH review of all suspicious cases and situations, and triggering of public health responses.
Within MOH, in March and early April, there was rapid and large-scale redeployment of staff, particularly public healthtrained doctors, to carry out SARS work. An additional 50 trained staff were loaned from the National Environment Agency to assist in contact tracing. More than 100 nurses from the Health Promotion Board and from government polyclinics were deployed to screen travellers at the airport and sea terminals, and to counsel contacts who were being served home quarantine orders. A commercial security company, CISCO, was contracted to serve home quarantine orders. However, the scale of contact tracing for the PPWM outbreak required the shortterm co-opting of a further 200 staff from the People’s Association. Following this experience, 250 personnel from the Singapore Armed Forces were attached to MOH
on 23 April 2003 for 2 months. These personnel greatly enhanced the contact tracing capacity and strengthened the operational capability of MOH to deal with a major quarantine operations. The Armed Forces team also included relevant experts that were able to expedite the development of the IT systems to support the contact tracing and epidemiological investigation work. In hospitals, infectious disease, respiratory and ICU physicians were very stretched. While the attendances and bed occupancies of hospitals nose-dived during the SARS outbreak, there was an increase in internal medicine-type clinical work, while workload in surgical disciplines fell drastically. At the height of the outbreak, hospital isolation rooms were in severe shortage. There was also a shortage of medical epidemiologists and staff to carry out contact
160
Severe Acute Respiratory Syndrome
tracing and maintain telephone surveillance of discharged patients. In response, the hospital CEOs and clinical leadership redeployed their staff and resources internally. Through the temporary conversion of wards, the number of isolation rooms in public hospitals was increased to 535, and an additional 120 beds were created through the use of customized containers, which in TTSH were put up in a 2-week timeframe. The MOH SARS Taskforce also co-ordinated the redeployment of staff across hospitals to meet specific needs, as well as the shifting of patient load. Most importantly, however, the huge and uneven surge in demand across hospitals and disciplines was largely met because of the dedication and professionalism of healthcare workers, who were prepared to work much harder and to help each other out.
Engaging the community and public communications One of the most striking features of the SARS epidemic was the tremendous fear which it created amongst the public throughout the world including Singapore. Some of the more visible manifestations of this fear were empty shopping malls and restaurants, dramatically reduced passenger traffic, and the widespread use of masks in public places in some countries. Managing this fear and turning it into a positive force for community bonding and action, was one of the most important aspects of the Singapore SARS experience. There were several key elements. Firstly, updated information on SARS and the local SARS situation was provided to the public in a timely and transparent manner. Secondly, the government’s strategies to contain the outbreak and to keep the general public safe through ‘ring-fencing’ of all possible SARS cases, were clearly explained. Thirdly, the importance of social responsibility and the steps which members of the public should take to reduce the risk of spreading SARS were emphasized. Fourthly, the full range of
governmental and voluntary community bodies was mobilized to assist in the fight against SARS. Fifthly, the political leadership highlighted the dedication and fearlessness of front-line health-care workers. This found resonance with the general public, and the huge wave of public support and goodwill was a critical factor in bolstering the morale of front-line medical staff. The bottom-line message was that SARS is a national problem and that all must play a part in combating it. The government’s containment policies were designed to minimize the risk that members of the community would be exposed to SARS in public places by ‘ring-fencing’ all cases where there was a suspicion of SARS. This ring-fencing involved the screening of travellers at the ports of entry; a dedicated ambulance transport system to ferry any suspicious cases to TTSH from the airport, sea terminals, medical clinics and homes of persons on home quarantine orders; a designated SARS hospital where all possible cases of SARS were transferred and isolated; and temperature screening at schools, work places and public buildings. Members of the public were urged to play their part in this ‘ring-fencing’ strategy and to be socially responsible — to comply with home quarantine orders if they were contacts; to stay at home if they had fever or were unwell; to adopt good personal hygiene habits such as not spitting in public. If everyone played their part, the rest of the general community could carry on with life and work safely and normally. Singapore’s Prime Minister and his Ministers took the lead by addressing and elaborating on these issues and strategies clearly and frequently. Specific information on SARS was provided by the very frequent press conferences chaired by the Minister of Health and through MOH’s SARS web-site and daily press releases. MOH also actively engaged the media which played a critical role in providing accurate information and educating the public about SARS and the SARS situation in Singapore. Posters and
Public Health Response: A View from Singapore
161
Table 16.6 Definitions of Yellow, Orange and Red Alert in the three-level response system to SARS Alert level
Yellow alert
Orange alert
Red alert
Definition
No local cases or sporadic imported cases only, without further local transmission
Local transmission but confined to close contacts in health-care settings or households
Local transmission no longer confined to close contacts; community outbreak
brochures in Singapore’s four main languages were put out to educate the public on the essential facts about SARS and the main symptoms. The first TV channel dedicated just to SARS was started. A SARS toll-free 24 hours hotline provided an important avenue for the public to get quick updates and accurate information, instead of relying on rumours. MOH also made an intensive effort to keep community leaders, industry and public agencies informed about SARS through numerous briefings and talks. Singaporeans in general had a high degree of confidence in the ability of the government to contain the epidemic. A survey carried out by the Health Promotion Board of 1086 respondents from 30 April 2003 to 13 May 2003, found that about 93% were satisfied or very satisfied with the Government’s response to SARS. The public also responded cohesively and responsibly, with numerous spontaneous community initiatives to assist those affected by home quarantine orders and convalescing from SARS, to show support to health-care workers and to assist in educating others about SARS.
Ensuring better preparedness for SARS and other infectious diseases WHO removed Singapore from its list of areas with recent local transmission of SARS on 31 May 2003. However, the SARS outbreak has dramatically underlined the need for maintaining a higher level of preparedness against infectious disease outbreaks. In
terms of enhancing the preparedness of the public health system, the steps detailed below have been taken or are in progress.
Establishment of a three-level response system A three-level response system which corresponds to the prevailing level of local transmission of SARS and severity of threat to public health has been established (Table 16.6). The range of control measures in health-care facilities and in the general community that will be implemented at Yellow-Orange-Red Alert levels has been formalized. This response system serves as a framework for preparedness planning and a platform for co-ordinating the response measures for the various agencies. At Yellow Alert, the main focus is to prevent imported cases and detect SARS cases early. Active surveillance and enhanced infection control measures and use of personal protective equipment in high risk areas in health-care facilities are maintained. Hospitals adopt workflow changes to separate febrile and non-febrile patients as much as possible. Temperature screening of inbound visitors are be implemented at all entry points. Public health containment measures are initiated if necessary. At Orange and Red Alert levels, additional measures are implemented to contain an outbreak, based on the extent of the outbreak and the public health threat. Infection control measures in health-care institutions and tracing and quarantine of contacts are be enhanced. Community sur-
162
Severe Acute Respiratory Syndrome
veillance through daily temperature taking at workplaces and schools is also be instituted. At Red Alert levels, closure of schools and other public places is considered.
out because the attending physician was alerted by the history that the patient worked in a laboratory which worked with live SARS CoV.
Enhancement of surveillance systems for infectious disease outbreaks
Strengthening public health operations capabilities and capacity
The rapid sharing of information, experience and best practices, particularly through the strong leadership and facilitation of WHO, played a major role in controlling the global SARS epidemic. SARS has also highlighted the crucial importance of early intelligence and the sharing of surveillance data. MOH continues to actively strengthen its links with its public health counterparts in other countries and with WHO and US CDC, to contribute to and to tap into early information on possible SARS cases in other areas as well as other potential infectious disease threats. At the same time, MOH is strengthening its local surveillance system for SARS and other infectious diseases. For SARS, surveillance for suspected cases includes the following: 1 In-patients (>16 years old) with atypical pneumonia. 2 In-patients with unexplained fever >72 hours and travel history in the preceding 10 days to previously SARS-affected areas/ countries. 3 Deaths due to unexplained acute respiratory illness. 4 Clusters of three or more health-care staff in the same work area with fever >38°C within 48 hours; and clusters of in-patients that the hospitals have assessed to be cause for concern. Cases that fall into these categories are isolated and notified to MOH. Coronavirus testing is carried out if the patient does not respond to treatment, if another cause for the illness cannot be found or if there are features in the history which raise the suspicion of SARS. A good example is the case of the doctoral student who was admitted to SGH on 3 September. The patient was isolated and coronavirus testing was carried
MOH has re-organized and strengthened its operational arm and systems to enhance its capacity for outbreak response management, contact tracing and quarantine operations. A new IT infrastructure has been put in place to support the surveillance and management of SARS and other infectious diseases. Essential epidemiological, clinical and laboratory data are captured in this system, known as the Infectious Disease Alert and Clinical Database. These data will be available online to MOH and to authorized users and can be used to facilitate rapid decision making and directing response activities. A Contact Tracing Centre has been set up in MOH to manage contact tracing in the community and co-ordinate and assist hospitals and other agencies undertaking other contact tracing activities. The target is to complete tracing of all contacts within 24 hours of the decision to do so. Training of designated staff within MOH and its related agencies, would meet the surge requirements up to Orange Alert level. Provisions have been made for additional manpower support to be provided by agencies outside of MOH to meet additional staffing needs. A major exercise was carried out in October 2003 which validated and assessed the operational effectiveness of the enhanced MOH systems and processes as well as that of participating hospitals.
Maintaining a high level of hospital preparedness All hospitals are maintaining a high level of preparedness. Triage of patients and the use of PPE in higher risk clinical areas continues. Monitoring of health-care workers and
Public Health Response: A View from Singapore
for suspicious clusters of febrile staff or patients is being carried out. MOH has conducted a series of exercises with all hospitals to assess their level of preparedness. Public hospitals are building additional isolation rooms and a new isolation centre, Communicable Disease Centre 2, with 39 isolation beds and 18 ICU beds, has been built next to TTSH. MOH is also working with the public hospitals to adopt practical arrangements so that patients who are potentially infectious can, as far as possible, be managed separately from other patients. The focus is on workflow practices as well as areas where patients from different wards and clinics congregate such as X-ray departments and operating theatres. For these areas, separate waiting areas and staggered appointment times may be necessary to manage febrile patients separately from non-febrile patients. For the febrile patients, additional precautions would be put in place such as cleaning equipment and trolleys after use associated with each patient. Vaccination against influenza has been carried out for residents of the Institute of Mental Health and nursing homes, and is being offered to health-care workers in public hospitals and health-care facilities.
Stockpiling of critical supplies MOH is maintaining a 6-month stockpile of critical supplies such as N-95 masks, gowns and gloves.
163
Continuation of temperature screening at the airport, seaport and land-links Temperature screening through thermal imaging scanners is being continued at the airport, seaport and land-links. This was discontinued in April 2004 but provisions have been made which allow for temperature screening to be re-instituted at short notice.
Conclusion The SARS outbreak in Singapore put a large number of patients in hospital, resulted in the death of 33 patients, disrupted the lives of countless Singaporeans and damaged the economy. On the positive side, however, it highlighted the importance of strong, effective political leadership and a cohesive professional and community response, in resolving the crisis rapidly. Moving forward, the lessons learnt from SARS will result in Singapore having a much higher level of preparedness for infectious diseases outbreaks. However, the model of close community participation and partnership in tackling public health problems is one that can and should be applied to pressing noncommunicable diseases such as obesity, diabetes, heart disease and cancer. In this respect, we can learn much more from SARS, than just the control of infectious diseases.
Acknowledgements Provision of information to doctors and medical professionals An enhanced web-based system has been developed and rolled out. It will give access to doctors and hospitals to contact and patient hospitalization information previously captured on SARS-Web, and will also provide customized advisories and guidelines. A series of training seminars has been launched to provide health-care professionals with the most updated information on SARS.
I am grateful to Dr Stefan Ma and Ms Gowri Gopalakrishna for collation and verification of data. The dedication and enormous contributions of health-care workers in Singapore were critical to the containment of the SARS epidemic.
References 1 Centres for Disease Control and Prevention. Update: Outbreak of Severe Acute Respiratory Syndrome — Worldwide, 2003. MMWR 2003, 52: 241–5.
164
Severe Acute Respiratory Syndrome
2 WHO issues emergency travel advisory, 15 March 2003. 3 WHO Update 44: Situation in China, revised casedefinition, support to National Labs, 1 May 2003 4 Ministry of Health, Singapore: Report of the Review Panel on new SARS case and biosafety. Biosafety and SARS incident in Singapore, September 2003. 5 Centres for Disease Control and Prevention. Severe acute respiratory syndrome — Singapore, 2003. MMWR 2003;52: 405. 6 Spurgeon D. Toronto succumbs to SARS a second time. BMJ 2003;326: 1162. 7 Lipsitch M, Cohen T, Cooper B et al. Transmis-
sion dynamics and control of severe acute respiratory syndrome. Science 2003,300: 1966–70. 8 World Health Organization. Influenza A (H5N1), Hong Kong Special Administrative Region of China — update. Wkly Epidemiol Rec 2003;78: 57–8. 9 Gopalakrishna G, Choo P, Leo YS et al. SARS transmission and hospital containment. Emerg Infect Dis 2004;10: 395–400.
Chapter 17
Public Health Response: A View from Hong Kong T Tsang
Intelligence gathering and surveillance systems With over 1700 persons diagnosed with SARS and almost 300 deaths attributed to the disease, Hong Kong was one of the regions worst hit by the epidemic.1 In addition to the immediate morbidity and mortality of the disease, the long-term health consequences among survivors are just beginning to emerge, such as residual pulmonary fibrosis2 and avascular necrosis of hip.3 SARS adversely affects the mental health of our people,4 and above all, continues to inflict psychological trauma to survivors and their families. The government recently earmarked HK$130 million (US$17 million) to help the families of SARS victims. The economic impact of SARS has been dramatic. In Hong Kong, public health interventions have evolved in keeping with knowledge about the disease. By and large, our public health response to SARS is no different from that of other affected places. The key measures include disease surveillance, port health screening, contact tracing and quarantine, public education and risk communication, infection control, suspension of schools, community mobilization, and international liaison. Nonetheless, some peculiar features of Hong Kong’s SARS outbreak give special meaning to our public health response and the lessons that we have learnt. Disease intelligence gathering in neighbouring areas proved critical to Hong Kong due to our geographical proximity to areas †
where SARS first appeared. Unfortunately, the mechanisms for outbreak information exchange between Guangdong and the rest of the world were not well established at that time. Things have changed as a result of SARS. On 17 April 2003, Hong Kong set up a meeting with Guangdong officials to exchange information on SARS prevention and control. Since May 2003, regular expert group meetings have been held between officials in Guangdong, Hong Kong, and Macao. Participants agreed to update each party with weekly/monthly statistics on SARS and other notifiable diseases, notify each other as soon as possible of any unusual outbreaks whether or not the aetiology is known, establish point-to-point contacts in disease communication, and co-operate in scientific research. Even before SARS hit Hong Kong, disease surveillance was stepped up. Hong Kong had already in place a sensitive influenza surveillance system that detected human cases of A(H5N1) in 1997 and 2003, and A(H9N2) in 1999. The day after Guangdong announced an outbreak of atypical pneumonia on 10 February 2003, surveillance for severe community-acquired pneumonia (SCAP)† was initiated. Contacts of every SCAP patient were traced and put on medical surveillance. Infection control guidelines on SCAP were provided in public hospitals. Between February 1 and 26 2003, the system detected 39 cases of SCAP.5 Despite heightened surveillance, the presence of a novel lurking agent proved difficult to
SCAP was defined as community-acquired pneumonia that required intubation or ICU care.
165
166
Severe Acute Respiratory Syndrome
uncover without additional information from across the border. As SARS developed in Hong Kong, the sheer number of cases rapidly overwhelmed routine ‘pencil and paper’ epidemiology. In early April 2003, a real-time electronic platform for SARS reporting was set up linking public hospitals and the Department of Health. A contact tracing application was soon added to combine data on SARS cases and their contacts. Furthermore, a police computer system was adopted to generate alerts on geographical clusters, and common exposures among cases. Together, these systems vastly improved the speed and efficiency of data gathering, management, and analysis. These systems are indispensable for the public health management of SARS and other infections that progress rapidly, generate a high caseload, and require frequent updates of status. Further development of web-based, real-time surveillance systems for infectious diseases is now under way in Hong Kong, building on the SARS experience.
Spread of the disease The international spread of SARS from Hong Kong to Canada, Vietnam, Singapore and other places (resulting in over 400 cases in total) through a single SARS patient staying at a hotel in February 2003 is one of the most intriguing epidemiological stories.5 This event highlights Hong Kong’s strategic position as an international hub of travel, which unfortunately, applies to both people and disease. To prevent the spread of SARS through international travel, Hong Kong required all arriving, departing, and transit passengers at airport, seaport, and land border control points to undergo temperature screening and fill in health declaration forms. Persons having close contact with SARS patients were barred from leaving Hong Kong during their quarantine period. Guidelines were issued to tourists and tourist agencies on the prevention and management of SARS and febrile respiratory illnesses during travel. During March 2003, two SARS cases were
picked up through their health declaration forms. Since April 2003, Hong Kong has not exported any new SARS cases via air travel. When the WHO removed Hong Kong from its list of SARS-affected areas on 22 June 2003, we had screened approximately 1.5 million passengers at the airport and 10 million passengers at land and sea border control points. Despite the disappearance of SARS worldwide, we are continuing the same port health measures and deploying additional staff to cope with increasing number of travellers. Quarantine and medical surveillance of SARS contacts underwent changes as the outbreak went on. Since 31 March 2003, close contacts were required to attend one of four designated medical centres daily for 10 days after the last contact with a SARS case. At these centres, all attendees were screened for body temperature, and chest X-ray were carried out on those with symptoms of fever, cough, or shortness of breath. Persons with positive chest X-ray findings were hospitalized. Approximately 15 800 attendances were recorded at the DMC, and 39 persons were found to have SARS.
Community control From 10 April 2003, household contacts of probable SARS cases were required to undergo home confinement (i.e. quarantine) for 10 days. Fifteen teams of visiting nurses conducted regular medical monitoring on the quarantined households. The police made spot compliance checks. With effect from 25 April 2003, household contacts of suspect cases were also put under home confinement. A total of 1262 persons have been quarantined, and 34 developed SARS. Because Hong Kong has not imposed quarantine for decades, societal acceptance was critical in ensuring that the fear of quarantine would not drive household contacts into hiding, leading to a paradoxical increase in transmission risk. A survey conducted during early April 2003 found that almost 90% of adults aged 18 and
Public Health Response: A View from Hong Kong
above agreed to be quarantined for 10 days if they were close contacts of SARS cases.6 Generally, compliance to the home confinement scheme was satisfactory. Only 45 warnings were issued for non-compliance and 3 cases were referred to the police. Quarantine of close contacts has been instrumental in retarding the SARS outbreak worldwide. It is interesting to note that the reproduction number of the outbreak in Hong Kong had been decaying even before home confinement came into effect on 10 April 2003. The reproduction number was 2.7 in the initial phase, falling to 0.9 on 26 March 2003, and to 0.14 on 10 April 2003.7 This may be related to societal awareness of the disease and the personal hygiene measures that people take to protect themselves. A notable feature of Hong Kong’s SARS outbreak is the relative high proportion of cases (52%) that occurred outside the hospital setting. This is largely due to the occurrence of some large community outbreaks, such as that in Amoy Gardens (329 cases) and several other housing estates. The transmission of SARS via a contaminated sewer system in Amoy Gardens was particularly striking.8,9 The Amoy Gardens outbreak led to a territory-wide campaign to maintain U-traps and sewer systems properly. It also prompted the formation of specialized multi-disciplinary response teams to inspect every building inhabited by SARS cases and carry out thorough disinfection in those buildings. Another rather unique experience concerns elderly homes. The SARS outbreak affected 72 elderly home residents (mean age 81 years) causing 57 deaths among them. Elderly home residents are a vulnerable target due to frequent visits to hospitals. Quite often they do not show the typical clinical presentation of SARS. To prevent residents discharged from SARS-affected hospitals from spreading the disease in homes for the elderly, hospitals adopted ‘step down isolation’ for contacts of SARS patients for 10 days before discharge. Some elderly homes also isolated recently discharged elders for 10 days. For those homes with inadequate isolation facilities, medical
167
social workers would work out alternative placements of discharged residents. Other important measures include medical surveillance, guidelines and briefing sessions on infection control for operators, providing homes with adequate protective gear, and monitoring home performance. Public education and risk communication were not easy during the initial phase of the SARS outbreak. Faced with uncertainties in risk assessment about the novel agent, definitive health advice is hard to give. Regular updates and changes are necessary as we come to know more about the disease. Professional assessment and public perception are not uncommonly at odds with each other. The lesson is to keep an open mind, and be honest enough to say there are things that we do not know. Public education and risk communication take more time than is generally recognized. Besides local and international media, other key players include politicians and community leaders, professional medical groups, the private medical sector, academic institutions, community institutions such as schools and homes for the aged, government departments, businessmen and tourism bodies, consular corps, and international health authorities. Particularly useful channels of public communication include daily press briefings, TV and radio announcements, the internet, sector-specific guidelines, and hotlines for public enquiry. The close partnership between Hong Kong, the WHO, and health authorities in other countries was instrumental in the global control of SARS. There was global concern when Hong Kong reported two human cases of influenza A(H5N1). Alarm bells were set ringing when Hong Kong reported two human cases of influenza A(H5N1) to the WHO on 19 February 2003. When Hong Kong reported the outbreak at the Prince of Wales Hospital to the WHO on 11 March 2003, the WHO issued a Global Alert on 12 March 2003. This Global Alert prompted countries to increase surveillance, and ultimately helped to discover the international spread of SARS arising from a hotel in Hong Kong.
168
Severe Acute Respiratory Syndrome
Preparing for the future In October 2003, a SARS Expert Review has come up with 46 far-reaching recommendations to enhance Hong Kong’s capability to respond to possible resurgence of SARS and other emerging infections. Many of these recommendations are now being undertaken. On the surveillance front, Hong Kong has fully adopted the protocol recommended by the WHO.10 Hospitals and elderly homes are monitored for respiratory outbreaks by laboratory analysis. To ensure accuracy in reporting of SARS laboratory results, the government is providing a free public health laboratory diagnostic and consultation service to the private sector. To minimize diagnostic confusion between SARS and influenza, the government provided free influenza vaccination to its health-care staff, long-stay residents of elderly homes and homes for the physically and mentally disabled, and elders in the community with chronic heart and lung disease who are receiving comprehensive social security allowance. This helps minimize the opportunity diagnostic confusion between influenza and SARS. Contingency plans for SARS have been drawn up. The government has published a checklist of measures against SARS, which categorized our SARS response into three levels: alert level, level 1 and level 2. The checklist not only involves the public medical and health sector but also the private sector and some 13 government departments. A great deal of emphasis is now being put on building surge capacity. Like other places in the world, the SARS outbreak has stretched our resources to the limit. It exposed the lack of sufficient manpower in some critical areas, including data management, infection control, and field epidemiological research. We are addressing the issue of surge capacity through recruitment of overseas experts, in-house staff training programmes, mobilization of the private medical sector and voluntary organizations, and
embedding public health personnel to work in hospitals and vice versa. Meanwhile, public hospitals in Hong Kong are busy expanding isolation facilities (1300 extra isolation beds), organizing training in infection control, improving ward design, ventilation, and bed spacing, and stockpiling personal protective equipment. A Centre for Health Protection is now being set up in Hong Kong. Its functions will include comprehensive public health surveillance on communicable diseases, partnerships with health-care professions, community, academics, government departments, national and international authorities, contingency plans for disease outbreaks, building capacity and professional expertise.
References 1 Asian Development Bank. Asian Development Outlook 2003. Hong Kong, China: Oxford University Press for the Asian Development Bank, 2003. 2 Antonio GE, Wong KT, Hui DS et al. Thinsection CT in patients with severe acute respiratory syndrome following hospital discharge: preliminary experience. Radiology 2003;228: 810–15. 3 Tan EL. Bone disease worry for former SARS patients. Medlineplus 2003; 10 October. 4 The Chinese University of Hong Kong. A survey of mood disorders after the SARS outbreak in Hong Kong, 2003; Press release 18 May. 5 SARS Expert Committee. Report; SARS in Hong Kong: from Experience to Action. October 2003. 6 Cheng C. Report on the public responses to the SARS outbreak in Hong Kong. 2003: Survey Research Center, Hong Kong University of Science and Technology. 7 Riley S, Fraser C, Donnelly CA et al. Transmission dynamics of the etiological agent of SARS in Hong Kong: impact of public health interventions. Science 2003;300: 1961–6. 8 Yu IT, Li Y, Wong TW et al. Evidence of airborne transmission of the severe acute respiratory syndrome virus. N Engl J Med 2004;350: 1731–9. 9 WHO. WHO environmental health team reports on Amoy Gardens, 2003. 10 WHO. Alert, verification and public health management of SARS in the post-outbreak period, 14 August 2003.
Chapter 18
Public Health Response: A View from a Region with a Low Incidence of SARS James W LeDuc
The spread of the epidemic The epidemic of severe acute respiratory syndrome (SARS) that occurred during the first half of 2003 was in many ways unlike any outbreak ever experienced by humankind. While past global epidemics, such as those caused by influenza and other infectious diseases, may have infected far more people or caused greater economic impact, the SARS outbreak was unique in several respects. SARS was caused by a novel, heretofore unrecognized virus; the route of virus transmission was not well known, but it was clear from even the earliest reports that health-care workers were among those most affected; the disease rapidly spread internationally through modern air transport so that all countries justifiably ascertained that their residents were at risk of infection; and there was no recognized treatment. These unknowns, coupled with heightened concerns about terrorism following the tragic attacks on the World Trade Centre and the deliberate use of Bacillus anthracis as a weapon of terror in late 2001, set the stage for massive global demand for information about this new disease and genuine fear in all sectors of both affected and unaffected nations. The World Health Organization (WHO) quickly recognized the need for global leadership in co-ordinating response efforts, and through its efforts, unprecedented international collaborations were mounted in the acute outbreak response. Places such as
China, Canada, Singapore, Vietnam, Taiwan, and Hong Kong experienced the full brunt of epidemic SARS, and routine life in these affected areas was seriously disrupted; however, the impact of SARS was not limited to these most hard-hit areas. Nations fortunate enough not to have large numbers of SARS cases were also adversely affected by the outbreak. The following is a brief summary of the SARS public health response mounted by the USA, where only eight serologically confirmed cases were documented, yet a massive response effort was nonetheless undertaken. The USA was extremely fortunate to have had only a handful of serologically confirmed SARS cases recorded during the outbreak, and no evidence of significant person-to-person spread domestically. All confirmed cases were among travellers returning from known areas of active transmission. While some of these patients became seriously ill, resulting in the need for intensive care and respiratory support, none died from their infections, and with only one possible exception, there is no evidence to suggest that these individuals passed their infections on to others. The exception is a husband and wife who had travelled together from Hong Kong and may have transmitted the infection from one to the other, or they may have both been exposed to a common source. Notably, the few cases seen in the USA do not correspond to the tremendous efforts made by federal, state, and local health authorities to protect
169
170
Severe Acute Respiratory Syndrome
the nation from epidemic SARS. A review of this US response, in which the Centers for Disease Control and Prevention (CDC) served as the lead federal agency, may be instructive as to the magnitude and intensity of domestic efforts undertaken to meet this global challenge. Rumours of epidemic respiratory disease in southern China began circulating in early 2003, but it was not until late February and early March 2003 that disease occurred outside of mainland China and the SARS outbreak came to the attention of the international health community. The situation was confounded by early reports of highly pathogenic strains of H5N1 influenza transmitted among birds, and the isolation of the virus from residents of Hong Kong on their return from Fugian Province, China.1 These events had already stimulated communications among members of the WHO influenza network and deployment of WHO and CDC scientists in February to Beijing to attempt an investigation. At this time Chinese officials were not receptive to offers of assistance; while these discussions were under way, word came of outbreaks of what
was later recognized as SARS in both Hong Kong and Hanoi. The small two-person team in Beijing elected to split up, with one person going to Hong Kong and the other to Hanoi to investigate these emergent clusters of cases. These outbreaks were soon recognized as the first indications of SARS outside of China. Over the course of the outbreak, a total of 86 CDC scientists were deployed on 92 separate overseas missions to 11 countries, as summarized in Table 18.1. These individuals included medical epidemiologists, laboratory scientists, infection control experts, pathologists, environmental scientists, and others; collectively, they dedicated approximately 7.8 work–years of effort to the response.
Laboratory response to SARS The first challenge facing those investigating the growing outbreaks in Hong Kong and Hanoi was to determine the cause of the illnesses seen. Local scientists in both countries rapidly ruled out known causes of disease, and the WHO, realizing that some-
Table 18.1 CDC staff deployed internationally in response to requests for assistance in managing the SARS outbreak, 2003 Country
Number deployed
Total days
Expertise*
Cambodia Canada China Hong Kong, SAR Laos Philippines Singapore Switzerland Taiwan Thailand Vietnam Totals
1 9 17 6 2 4 5 4 30 4 10 92 deployments**
15 103 498 88 5 98 137 33 696 60 226 1959 days (=7.8 work-years)
Med/Epi Med/Epi; IH; Media Med/Epi; Path/lab Med/Epi; IC; IT Med/Epi; PHA Med/Epi; IC Med/Epi; Path/lab; IT Med/Epi; Path/lab Med/Epi; Path/lab; IC; PHA Med/Epi; PHA Med/Epi; Path/lab
*Med/Epi, medical officer, epidemiologist; Path/lab, pathologist, laboratory scientist; IC, infection control expert; IH, industrial hygienist; IT, information technologist, data manager; PHA, public health advisor and administrator; Media, communications expert. **86 personnel deployed; 6 deployed to 2 or more countries.
Public Health Response: A View from a Region with a Low Incidence of SARS
thing unusual was taking place, organized a group of 11 international centres of excellence to determine collaboratively the aetiology of the outbreaks.2 Through the leadership of the WHO, their effective use of daily conference calls and a secure website for immediate posting of results, and a remarkable openness and sharing of findings among the collaborating scientists, the discovery of a novel coronavirus as the cause of the outbreak was made.3,4 The isolation and identification of the virus was quickly followed by the determination of its complete genomic sequence and the recognition that the virus responsible for SARS was new to science, almost certainly the result of a naturally occurring jump from a wild animal reservoir to humans, and not an intentional act of terrorism.5 As the global epidemic progressed, CDC scientists worked closely with partners in affected nations, the WHO, and state and local facilities to develop rapidly and validate diagnostic reagents for the new SARS-associated coronavirus (SARS CoV), examine a growing number of specimens from suspect and probable SARS cases seen in the USA and internationally, and provide training to domestic and international partners in the use of newly developed diagnostic tests for SARS. Laboratory scientists were deployed to several countries to assist affected nations through training and technology transfer, and to help conduct diagnostic testing on suspect and confirmed SARS cases. SARS serological diagnostic test reagents were produced and provided to health departments of virtually every state in the USA, while well-validated primers for real-time polymerase chain reaction (PCR) analysis of specimens were prepared for the national Laboratory Response Network (LRN), established to help combat bioterrorism. In all, over 120 CDC laboratory scientists worked on some aspect of the laboratory response to SARS, either at CDC, in state or local health departments, or on site in other nations. Furthermore, more than 130 shipments of SARS CoV, viral
171
nucleic acids, or SARS CoV-specific antibodies were provided to domestic and international partners in academic centres, government agencies, or the commercial sector.
Surveillance for SARS An aggressive surveillance campaign was initiated concurrently with the laboratorybased investigations. The goal of this campaign was to identify all SARS cases as quickly as possible and to determine if secondary transmission to health-care workers, family members, or other contacts might have occurred. To do this, CDC scientists worked closely with a multitude of domestic partners, including state and local public health departments and professional organizations, and global partners, including WHO. A broad, non-specific case definition was used that was based on WHO’s definition and included current or recent past history of an acute febrile illness with respiratory involvement and a history of travel to or transit through an area where active SARS transmission was under way, or close contact with ill persons who had travelled to such areas or who had suspect or probable SARS. This use of a broad-based case definition maximized the opportunity to ensure that no case was missed; however, it also meant that literally hundreds of suspect cases had to be tracked and investigated, histories taken, clinical specimens drawn and tested, and potential contacts identified. Persons meeting the basic case definition were categorized as suspect cases; those individuals with the additional documented evidence of pneumonia were classified as probable cases. Nasal pharyngeal swabs, aspirates, sputum, or other specimens taken from the respiratory tract were tested for the presence of SARS CoV by PCR, and acute- and convalescent-phase serum samples were tested for the presence of IgM and IgG antibodies to the SARS CoV, as well as for evidence of infections with other respiratory pathogens.
172
Severe Acute Respiratory Syndrome
Communications One of the most complex challenges faced during the SARS response was in meeting the constant demand for current, accurate information about the outbreak. Several different groups required dedicated communications efforts. The travelling public wanted to know the risk of SARS should they go to an area where SARS was being actively transmitted. The general public wanted to know about the latest case counts, the risk in their daily lives, and the evolving scientific discoveries and outbreak control efforts. The health-care delivery community needed to know how to recognize SARS cases, how to care safely for those infected, where diagnostic testing was available, and how health-care professionals could protect themselves from infection with SARS CoV. State and local health departments required information and assistance; the WHO requested daily case counts and updates; federal agencies needed to be briefed, as did many others, including multinational businesses and other countries. To meet these enormous communications demands, a number of dedicated information update and exchange activities were begun. Within CDC, daily briefings were prepared each morning for the Director and other top agency officials. This detailed information was summarized and provided, also daily, to the Secretary, Department of Health and Human Services, and other senior staff within the Department and its agencies, including the National Institutes of Health and the Food and Drug Administration. The actual outbreak response efforts were led by scientists from CDC’s National Center for Infectious Diseases (NCID), but the staff members needed to handle these various duties were drawn from programmes throughout CDC. Several internal investigative teams were formed (e.g. laboratory, clinical, quarantine, international collaborations, communications), and team leaders met twice a day throughout the course of the outbreak to discuss
progress and co-ordinate the next steps. Internal conference calls were scheduled two to three times a week to inform CDC programmes outside NCID of progress made and resources needed. Similar calls were frequently made to all state epidemiologists and their professional organizations. All the while, individual teams from CDC worked hand-in-hand with state and local health department officials to track suspect cases, obtain diagnostic specimens, and determine if close contacts had developed a similar illness. In all, over 800 CDC staff members worked on some aspect of the SARS response efforts. As shown in Table 18.2, numerous SARS interim guidance documents were posted on the CDC web-site. These documents provided accurate, timely information to clinicians, travellers, patients, Americans living abroad, and many other specialized groups. Information regarding specific topics was also posted, such as guidance for transport of patients, collection and transport of specimens for laboratory testing, infection control, isolation and quarantine, reporting mechanisms and others. Information was often made available on the web in multiple languages as well. It was estimated that more than 17 million page views were made to the SARS web-sites during the outbreak, with 3.8 million for the week of 20–26 April 2003 alone during the height of the outbreak. Health-care professionals had special communications needs, and these were met through many avenues for information dissemination, including the creation of a dedicated 24-hour-a-day clinician information hotline that responded to more than 2000 telephone calls from physicians. A similar hotline for the general public answered more than 34 000 phone calls. Three special video presentations that dealt with patient management, diagnostic testing, and related health-care issues were broadcast by satellite to the medical community, reaching an estimated 1.9 million participants. In addition, 30 health-care provider
Public Health Response: A View from a Region with a Low Incidence of SARS
173
Table 18.2 SARS Web-based Guidance Documents posted on CDC website (HTTP://www.CDC.gov/ncidod/sars) as of October 2003 Clinical and hospital • Guidelines for Collection of Specimens from Potential Cases of SARS • Fact Sheet for Clinicians: Interpreting SARS Test Results from CDC and Other Public Health Laboratories • Interim Domestic Guidance for Management of Exposures to SARS for Healthcare and Other Institutional Settings • Interim Guidance on Infection Control Precautions for Patients with Suspected SARS and Close Contacts in Households • Interim Domestic Guidance on Infection Control for Persons Who Have Laboratory Evidence of SARS Coronavirus but Who Have Either No Symptoms or Mild Symptoms That Do Not Meet the Clinical Case Definition for SARS • Updated Interim Domestic Guidelines for Triage and Disposition of Patients Who May Have SARS • Updated Interim Guidance: Pre-Hospital Emergency Medical Care and Ground Transport of Suspected SARS Patients • Interim Domestic Guidance on the Use of Respirators to Prevent Transmission of SARS • Interim Domestic Infection Control Precautions for Aerosol-Generating Procedures on Patients with Severe Acute Respiratory Syndrome (SARS) • Interim Domestic Guidance on Persons Who May Have Been Exposed to Patients with Suspected SARS • Interim Guidance: Air Medical Transport for Severe Acute Respiratory Syndrome (SARS) Patients • Interim Recommendations for Cleaning and Disinfection of the SARS Patient Environment • Updated Interim Domestic Infection Control Guidance in the Health-Care and Community Setting for Patients with Suspected SARS • Safe Handling of Human Remains of Severe Acute Respiratory Syndrome (SARS) Patients: Interim Domestic Guidance Laboratory • Interim Guidelines for Laboratory Diagnosis of SARS-CoV Infection • Interim Laboratory Biosafety Guidelines for Handling and Processing Specimens Associated with SARS • Packing Diagnostic Specimens for Transport: Summary Instructions • Instructions for Collecting and Shipping Internationally Originated Laboratory Specimens Associated with SARS with Corresponding Epidemiologic and Clinical Information General Information • Interim Guidelines about SARS for Persons in the General Workplace Environment • Guidance about SARS for Americans Living Abroad • Interim Guidelines about Severe Acute Respiratory Syndrome (SARS) for Persons Traveling to Areas with SARS • Interim Guidance for Businesses and Other Organizations with Employees Returning to the United States from Areas with SARS • Interim Guidance for Institutions or Organizations Hosting Persons Arriving in the United States from Areas with Severe Acute Respiratory Syndrome (SARS) • Fact Sheet: Isolation and Quarantine • Fact Sheet on Legal Authorities for Isolation/Quarantine • Interim Domestic Guidance for Health Departments in the Management of School Students Exposed to Severe Acute Respiratory Syndrome (SARS) • Interim Guidelines for Personnel Interacting with Passengers Arriving from Areas with SARS • Interim Guidelines about SARS for Airline Flight Crew Members • Interim Guidance for Cleaning of Commercial Passenger Aircraft Following a Flight with a Passenger with Suspected Severe Acute Respiratory Syndrome (SARS) • Interim Guidelines for Personnel Boarding Maritime Vessels from Areas with SARS • Interim Guidelines about Severe Acute Respiratory Syndrome (SARS) for Cruise Ship Passengers and Crew Members • Interim Guidelines and Recommendations: Prevention, Identification and Management of Suspect & Probable Cases of SARS on Cruise Ships • Interim Guidelines for Workers Handling Cargo or Other Packages • Interim Guidelines about SARS for International Adoptees and Their Families
174
Severe Acute Respiratory Syndrome
conference calls were held to discuss SARS and patient management issues. The traditional CDC publication, Morbidity and Mortality Weekly Report, carried weekly updates throughout the epidemic, often with dedicated in-depth summaries from the mostaffected countries. And finally, CDC staff answered literally thousands of email inquiries from colleagues, friends, and various others interested in the evolving SARS outbreak. To meet the informational needs of the general public, CDC hosted 21 live telebriefings and news conferences, issued 12 news releases, handled more than 10 000 news media calls, and provided many indepth interviews with news media organizations. A dedicated historian was retained to compile extensive meeting notes and to record events as they unfolded. To inform the travelling public returning from areas where SARS transmission was under way that they may have been exposed, more than 2.7 million health alert notices (Fig.
18.1) were passed out to arriving passengers at airports and docks as they disembarked. These notices, which were available in several different languages, described the signs and symptoms of SARS and recommended that persons who thought they might have been exposed should contact their local health-care provider in advance of their visit to seek medical care, and to inform their doctor of their relevant travel history.
Conclusions While the USA had only a few confirmed cases of SARS, the outbreak nonetheless had a significant impact on national, state, and local health officials. Thousands of public health officials, health-care providers, and other professional and support staff dedicated countless hours responding to enquiries and addressing scientific and technical issues raised by the SARS global outbreak. The fact that any traveller could conceivably be a source of infection made
Figure 18.1 Health alert notice passed out to more than 2.7 million passengers arriving either directly or indirectly from areas where SARS transmission was occurring. Notices were published in eight different languages (English, French, Japanese, Spanish, Korean, Traditional Chinese, Simplified Chinese, Vietnamese).
Public Health Response: A View from a Region with a Low Incidence of SARS
every nation take notice, whether they actually had a case or not. The SARS epidemic provided definitive evidence that emerging diseases are a true global threat, that they do not recognize international borders, and that we are all at some level of risk. The outbreak also gave witness to the effectiveness of international co-operation and collaboration, and proved beyond a doubt the value of accurate, timely surveillance and effective response to outbreaks of infectious diseases.
Acknowledgement I wish to thank Dr Anne Pflieger for technical support in compiling the information presented in this chapter. Significant portions of this summary also appeared in: LeDuc JW, Pflieger A. The centers for disease control and prevention’s role in international coordination and collaboration in response to the SARS outbreak. In: Learning from SARS: Preparing for the Next Disease Outbreak (Workshop Sum-
175
mary). Knobler S, Mahmoud A, Lemon S, Ma ck A, Sivitz L, Oberholtzer K (eds.), Forum on Microbial Threats, Board on Global Health, Institute of Medicine; 2004:50–6.
References 1 Influenza A (H5N1), Hong Kong, Special Administrative Region of China. Wkly Epidemiol Rec 2003;78: 49–50. 2 Stohr K and World Health Organization Multicentre Collaborative Network for SARS Diagnosis. A multicentre collaboration to investigate the cause of severe acute respiratory syndrome. Lancet 2003;361; 1730–3. 3 Peiris JSM, Lai ST, Poon LLM et al. Coronavirus as a possible cause of severe acute respiratory syndrome. Lancet 2003;361: 1319–25 [epub ahead of print 8 April 2003]. 4 Ksiazek TG, Erdman D, Goldsmith C et al. A novel coronavirus associated with severe acute respiratory syndrome. N Engl J Med 2003: 348: 1953–66 [epub ahead of print 10 April 2003]. 5 Rota PA, Oberste MS, Monroe SS et al. Characterization of a novel coronavirus associated with severe acute respiratory syndrome. Science 2003;300; 1394–9 [epub ahead of print 1 May 2003].
Chapter 19
Infection Control for SARS: Causes of Success and Failure WH Seto, PTY Ching and PL Ho
Introduction Hospitals proved to be a major amplifier in the spread of SARS.1 Significant numbers of health-care workers and others acquired the infection within a hospital setting. Overall, over 1700 health-care workers (HCW) are reported to have contacted the disease. The percentage of SARS accounted for by HCW in localities with large outbreaks are 57% in Vietnam, 43% in Canada, 41% in Singapore, 22% in Hong Kong, 19% in Mainland China and 20% in Taiwan.2 This relationship of SARS with nosocomial outbreaks provides the underlying rationale for the WHO post-SARS surveillance strategy. Member nations are recommended to implement a ‘SARS Alert’ programme, which among other things, emphasizes the need to maintain vigilance for outbreaks in the hospitals.3 The presence of a nosocomial outbreak is therefore used as an indicator that the disease might have re-emerged in a locality. Thus, good infection control practices underpin all strategies for the control of SARS. Some of the early observations on possible modes of disease transmission within hospitals that guided early recommendations on infection control arose from the careful observations made by the late Carlo Urbani and others in Hanoi, Vietnam who coped with the SARS outbreak there. In this chapter, we review some of the lessons learned in Hong Kong and outline the infection control programme in place during the
176
SARS outbreak in Queen Mary Hospital, Hong Kong. This hospital cared for patients with SARS but only contributed to 2 of the 386 infected HCW in Hong Kong.1
Evidence for the efficacy of infection control measures in preventing SARS Case control study for mode of transmission in the hospital This study was conducted in mid-March, immediately after the WHO global alert in Hong Kong and has been reported elsewhere.4 It compared the infection control precautions taken by 241 non-infected and 13 infected staff that had provided direct care to 11 confirmed SARS patients. Four specific measures were especially studied, which were the washing of hands and the wearing of masks, gowns and gloves. The results showed that adherence to proper ‘droplets and contact precautions’ as recommended in the guidelines of the Centers for Disease Control Guideline5 significantly protected HCW from SARS. None of the 69 staff reporting the practice of all four measures were infected. In contrast, all 13 infected staff omitted at least one of the measures (p<0.0224 Fisher’s two-tailed).4 It was also notable that all infected staff were either from the general medical wards (12 persons) or the emergency department (1 person). In contrast, the majority of the non-infected staff (135 of 225) were either
Infection Control for SARS: Causes of Success and Failure
from the isolation ward or the intensive care unit (p<0.0001 Fisher’s two-tailed). It was reasonable to assume that infection control measures, especially in the early stage of the SARS outbreak, were much better in the isolation wards and intensive care units as compared to the general medical wards. The effect of ‘recall bias’ inherent in such studies has to be kept in mind. However, the study was completed in the relatively early stage of the global outbreak and before the formal announcement by the Hong Kong Government on 27 March 2003 that SARS had spread into the community. The disease at that stage was still relatively new to most of the subjects surveyed and the effect of recall bias is likely to be minimal. Furthermore, the associations demonstrated were clear and information requested was on simple concrete behaviour regarding recent events
Study on infection control practices of the general medical wards in ten hospitals Towards the end of the SARS outbreak in June 2004, another study was conducted in ten Hong Kong hospitals to evaluate infection control practices (ICPs) in the general medical wards. Two general medical wards in each of the ten hospitals to which SARS patients had been admitted were randomly selected for study. The problem of ‘recall bias’ must obviously be addressed at this stage of the outbreak and therefore two different methods were used to evaluate the ICPs in these hospitals. One was a staff survey of the practices, rather similar to the survey conducted in the first study reported above. In addition, direct unobtrusive observations on the ICPs were also conducted in each of the wards. The observers were nurses from the study team who obtained permission from the hospital administration to enter these wards because they needed data on the physical environment, including the size of the different ward
177
areas, number of windows and the space between beds for future resource planning. They were able to observe the practices in each of these wards for one full working day. Data on eight specific ICPs were collected in both studies. At the same time, the number of SARS patients in these wards, their duration of stay in the ward and the number of HCW infected were obtained from the clinical records. The staff survey was administered to the staff on duty the day after the direction observation was conducted. In total 331 staff were surveyed. In the direct observational study, 844 ICPs carried out by 397 subjects were evaluated. Correlation statistics by Spearman were carried out to explore significant relationships. The results are shown in Table 19.1. There were 26 SARS patients admitted to these 20 wards and 34 staff were infected. The results from the two studies were highly convergent with a significant Spearman’s correlation of 0.87 (p<0.001), indicating that there was substantial convergence validity in the data. The level of ICPs compliance was impressively high. Nearly everyone was wearing a mask when rendering care to the patients. The use of other relevant personal protective equipment (PPE), with the exception of shoe covers, was also high. The use of goggles and face shields were alternative options but some staff were wearing both PPEs. Handwashing was reported to be 97% in the survey, but was only 78% in direct observation. Nevertheless this is one of the highest handwashing compliance rates reported in an observational study.6 None of the ICPs correlated significantly with the number of staff infected in these wards (Table 19.1). The only factor found to correlate with the number of staff infected, was the number of SARS patient days in the ward. This result is not altogether unexpected. The survey was carried out towards the end of the outbreak period and the staff were already highly primed for infection control. It was therefore reasonable to observe such a high level of compliance in ICPs. When ICPs reach a certain level of
178
Severe Acute Respiratory Syndrome
Table 19.1 Infection control practices for patients in general medical wards in ten Hong Kong hospitals (a) Direct observations Practices
Average (%) (n = 884)
1. Mask N95 41 Surgical Both 2. Glove 3. Gown 4. Faceshield 5. Goggles 6. Cap 7. Shoe-covers 8. Handwash (before) 9. Handwash (after)
100 0.11 20 39 91 99 69 46 92 7 65 78
Correlation with* number of staff infected in ward
0.63 0.10 0.25 0.29 0.15 0.12 0.13 0.27 0.22 0.00 0.03
p
0.66 0.30 0.22 0.53 0.62 0.60 0.24 0.35 0.99 0.90
(b) Staff survey Practices
Average (%) (n = 331)
1. Mask N95 55 Surgical Both 2. Glove 3. Gown 4. Faceshield 5. Goggles 6. Cap 7. Shoe-covers 8. Handwash
99 0.23 25 19 90 81 61 46 76 15 97
Correlation with* number of staff infected in ward
p
0.15 0.36 0.06 0.04 0.48 0.05 0.09 0.18 0.20 0.02 0.09
0.53 0.80 0.88 0.85 0.85 0.72 0.47 0.43 0.92 0.74
(c) SARS patient’s days in wards Average number of days** 13.3
Correlation with* number of staff infected in ward
p
0.56
0.10
* Correlation by Spearman R (total of 34 staff infected in the 20 wards). ** Total of patients admitted (total of 26 patients admitted to the 20 wards).
compliance, one would expect protection to peak at a certain level. This was the probable reason why the levels of ICPs compliance noted have no correlation with the number of staff infected. However, it would be unrealistic to expect no lapses in ICP
practice from the staff. If one of these lapses occurred when the infectious virus was present, it could still result in infection. Thus, the duration of exposure becomes relevant because it would increase the likelihood of one of these lapses leading to contamina-
Infection Control for SARS: Causes of Success and Failure
tion. Thus, the duration of hospital stay for SARS patients is a measure of the ‘quantity’ of exposure and there was a significant correlation of this factor with the number of staff infected in the wards. These studies indicated that with good infection control practices, it was possible to reduce the risk of HCWs being infected. However, it also suggests that the occasional lapse in practice led to infection in HCW.
Infection control programme in Queen Mary Hospital Queen Mary Hospital (QMH) is a 1400-bed teaching hospital of the Medical Faculty of the University of Hong Kong. Like all other acute hospitals in the Hong Kong SAR, it had its share in the care of patients with SARS during the outbreak. A total of 710 suspected SARS cases were admitted to the isolation wards, of which 52 were subsequently virologically confirmed to be SARS. Six of these patients were intubated and five of them had a fatal outcome. There was a 28day period in which there were more than 25 cases of SARS in the hospital, indicating the extent of potential exposure. However, only 2 of the 386 SARS- infected HCWs in Hong Kong were from QMH, viz. 0.5% of the total, by far the lowest reported infection rate for HCWs in Hong Kong. Furthermore no nosocomial SARS was reported among patients from QMH during the outbreak. It was relevant therefore, to review the infection control programme of the hospital during the outbreak period. The salient points of the programme are summarized in Table 19.2. It is pertinent to note that the hospital already had a strong ongoing infection control programme with emphasis on the ‘basics’ of infection control (Table 19.3). For example, surveillance for nosocomial endemic infections had been in place for over 10 years and included surveillance for surgical wound infections with feedback to the surgeons7 and device-related infections in the ICU. The infection control nurse (ICN):
179
bed ratio was close to 1:250, and QMH is probably one of the few hospitals in Hong Kong to have such a high ratio of infection control practitioners. There was a strong emphasis on altering inappropriate patient care practices, some examples being listed in Table 19.2 (point 2). In addition, the concept of an ‘infection control link nurse’ had been initiated and implemented for over 10 years and has been reported previously.8 This was a programme where nurses in individual wards were selected to represent the Infection Control Unit ‘at the ward level’. They were provided with 4 days of training every year and were given the title ‘Infection Control Link Nurse’. There were 60 such nurses in the hospital at the beginning of the SARS outbreak. These nurses became an informal network providing valuable information to the Infection Control Team during the months of the outbreak. Five aspects of the QMH infection control programme during the SARS outbreak are discussed below.
Strong administrative leadership Overall leadership was obviously critical for the success. A task force was immediately assembled but the key was the willingness of the various departments to work together. This was evident by the rapid decanting of workload that occurred, so that the hospital could concentrate in the management of SARS patients. The number of admissions from March to May in 2003 was 16 374 as compared to 18 902 for the same period the previous year, a drop of over 13%. The expertise in the hospital was appropriately mobilized and utilized. The leaders in each subspeciality were mobilized for management of SARS. A senior respiratory consultant was appointed to lead the clinical team, a certified intensivist led the intensive care management of SARS patients, a virologist was responsible for the viral diagnostics tests for SARS and the Infection Control Officer was a microbiologist with extensive
180
Severe Acute Respiratory Syndrome
Table 19.2 Salient features of infection control programme in Queen Mary Hospital 1 Strong administrative leadership Formation of SARS task force for the hospital Facilitating professional expertise Using only senior staff for direct patient care No deploying of new staff unfamiliar with the hospital environment Rapid decanting of non-SARS workload Rapid provision of adequate manpower for clinical care and infection control 2 Strong infection control emphasis, especially on the basics Strong infection control culture already existed in the hospital Focus on the basics like handwashing and wearing mask as obligatory Standard system of triage and isolation of possible SARS patients Eliminate common errors: neglecting handwash after de-gloving, gloving all the time instead of for specific procedures, double gloving without washing hands, use of unnecessary multiple layers of PPEs, contamination of personal items (e.g. name tags), wearing of used PPEs outside patient care areas, careless de-gowning procedures at the end of shift 3 Intensive surveillance programme Daily follow-up of all hospitalized SARS cases until 10 days after discharge Contact tracing of all SARS cases contacts for follow-up by Department of Health Follow up of all patients discharged from suspected SARS wards for 10 days Active surveillance to identify SARS patients in the general medical wards Immediate investigation of even one staff with suspected SARS Immediate investigation of outbreaks of SARS in the hospital Staff survey after all high-risk procedures involving SARS Informal communication by the 60 infection control link nurses Maintain database 4 Education and communication Face to face education for all staff Daily report to task force core members of new cases and progress of existing cases Daily newsletter to all hospital staff Hot-line for staff advice and counseling Trouble-shooting sessions with specific departments Update guidelines available in the hospital website 5 Logistics and staff welfare Ensure sufficient supply of PPEs for all staff Provision of shower facilities of staff Provision of living quarters for staff requiring lodging in the hospital Provision of quarantine facilities for staff
training on Infection Control in the United States in the early 1980s. Only medical staff with over 6 years postgraduate experience were recruited to care for SARS patients. While it was common in Hong Kong at the time to recruit staff from other hospitals to assist in managing the SARS outbreak, this was not the case in QMH. Only staff who were familiar with the hospital environ-
ment were deployed. It was felt that staff working in an unfamiliar environment were more likely to be more prone to lapses in infection control.
Emphasize the ‘basics’ of infection control During the SARS outbreak, the two infec-
Infection Control for SARS: Causes of Success and Failure
Table 19.3 Basics in infection control 1 Ongoing surveillance programme for nosocomial infections with feedback. 2 Inappropriate patient-care practices are responsible for most nosocomial infections. Eliminate these and replace them with good practices (handwashing most important). 3 Ensure good environmental hygiene. 4 Chemical disinfection or sterilization to be considered only after proper physical cleaning. 5 Ensure protection against infection is effective in the staff health programme. 6 Isolation precautions must be effectively implemented for infected patients. 7 All clusters to be identified and dealt with by the appropriate response. 8 Effective education that will result in staff compliance to infection control policies. 9 Deploy sufficient full-time infection control nurses for the programme. 10 Established the appropriate infrastructure including supervision by an infection control doctor.
tion control practices that were strongly promoted and ‘non-negotiable’ under any circumstance were handwashing and the proper wearing of masks. This was enshrined in the hospital with the slogan ‘Wash hands, wear masks, control SARS’. Wearing a clean surgical mask when approaching patient is the minimum. The N95 must be used in high-risk areas such as the SARS isolation wards and for high-risk procedures for SARS patients, such as intubation and aspirations, with the fit-check being carried out each time the mask is worn. A mask must be assumed contaminated through wearing and must be thrown away after patient care or after a high-risk procedure. It should not be kept in a paper bag or other retainer for future use. The hospital infection control policy was not to use the more sophisticated respirators such as N100 or the Powered Air Purifying Respirator (PAPR). Presently the N95 is regarded as acceptable protection for SARS patients in the most current guidelines from both the WHO9 and CDC.10 All patients in the hospi-
181
tal were also advised to wear a surgical mask at all times. A system for admission, triage and isolation was put into effect in the hospital.11 All patients who were suspected to have SARS were admitted into the Triage Ward. To reduce the risk of cross-infection, the number of beds in each bay was reduced and the minimum bed-to-bed distance was set at 2 m. Furthermore, whenever possible a bed was left unoccupied in between two consecutive patients. The patients in the Triage Ward would be evaluated intensely by CXR, computer tomography and laboratory tests. When the diagnosis of SARS was likely, the patient would be transferred to the SARS isolation ward. On the other hand, if the diagnosis of SARS was excluded, they would be transferred to a step-down ward for 10 days before being discharged home. The triage system of the hospital has been reported in greater detail elsewhere.11 A total of 710 patients were admitted into these wards during the SARS outbreak and 52 were confirmed to be SARS, but none of these patients were reported to have acquired the infection nosocomially in the hospital.
Intensive surveillance programme At the onset of the SARS outbreak, the hospital deployed eight additional staff in the Infection Control Unit. Surveillance on all vital aspects of the outbreak was conducted. These included surveillance on SARS cases and their contacts. All new SARS admissions were interviewed and their contact information given to the Department of Health so that they could initiate further surveillance or quarantine where needed. Suspected SARS cases were similarly interviewed and both the confirmed and suspected cases would be followed up personally or by phone for up to 10 days after discharge from the isolation area. In the months of the outbreak, the Infection Control Unit made over 3000 phone calls. All suspected clusters of infected staff were immediately investigated. The epidemiologi-
182
Severe Acute Respiratory Syndrome
cal investigation of the one confirmed cluster of two cases in the hospital was completed within 48 hours. It was found that the HCW infection was related to the use of 100% oxygen therapy. Technically, a single case does not constitute an outbreak and would not entail an investigation. However, because SARS was seen to have such dire consequences, it was decided that even a single case would be investigated. A predesigned protocol for investigation was used (Table 19.4). There was also an active surveillance programme to identify possible SARS patients in the general medical wards who might be inadvertently admitted. All patients who developed fever would be evaluated with the support of reviews by senior clinical staff where needed. All patients with suspected SARS would be then transferred to the Triage Ward. During the SARS epidemic, a total of 24 patients were transferred to the Triage Ward of whom four were later confirmed to be SARS.11
Education and communication Education has always been a key component of infection control. In QMH, research had shown that effective education that would actually alter behaviour must be carried out through face-to-face communication.12 So the strategy was to conduct an educational campaign that would provide face-to-face interaction with all staff. A total of 63 lectures was given during the outbreak period, with a host of small group sessions to discuss particular problems in specific departments. Other education aids such as CD-Roms and posters were utilized, but the face-to-face coverage was close to 90% of all staff. There were also specific communications to leadership in the task force on the status of all SARS cases and a daily newsletter to all staff. This newsletter was written, not just as a news article, but passionately, as a rallying call for all to band together for the battle against SARS in QMH. The
Table 19.4 Protocol for investigating a single suspected case of SARs in staff for QMH Introduction In an investigation in which there is only one case, two possibilities should be entertained 1 There is an event or a series of processes, which resulted in a cluster. The present case is just the first one identified. In such a scenario, we explore for some events or procedures that may involve several staff. 2 The present staff member is the only one infected. The factor that had resulted in him being infected, somehow has not affected the others. Steps in investigation 1 Collect detailed history of the particular staff member 2 Collect details on possible contact history outside the hospital and with other staff. These may require quarantine when the diagnosis of SARS is confirmed 3 Review all activities of the staff in the hospital in the last 10 days before presentation. 4 Identify possible major catastrophic events in the ward, especially high-risk procedures. 5 For such procedures in which suspected case is involved, generate a list of staff similarly implicated. Surveillance on these staff should be carried out. 6 Review with the suspected staff member his infection control practices in the last 10 days. 7 If lapses are identified, unobtrusive observation in the ward will be made to ensure that this practice is not widespread. 8 Conduct overview scan with ward leaders of possible infection control deficiencies in the ward. 9 A hospital-wide scan must be conducted for all deficiencies identified to ensure that these are addressed. Note: In step 7, if deficiency is found to be mainly personal and not practice by others in the ward, there would be a lower chance of a possible cluster emerging.
Infection Control for SARS: Causes of Success and Failure
web-site was utilized to the full, with the latest guidelines available for all to download when needed.
2
Logistics and staff welfare Staff were provided with the material required to implement the mandated infection control policy of the hospital, for example, the recommended PPEs that were available in sufficient quantity throughout the SARS outbreak period. The policy of the hospital was to encourage staff to shower freely if they believed they had been contaminated. The slogan was ‘In infection control lapse — wash, wash, wash’. Thus shower facilities were made available to enable this to be practised by the staff. Living quarters were also provided for staff who wanted to stay in the hospital overnight.
3
Conclusion
8
The outbreak of SARS had affected many hospitals across Hong Kong, the region and the world. While this disease proved to be a major threat to the health-care profession, the greatest threat was its potential to spread within the hospital setting. We believe that nosocomial spread can be prevented by simple good infection control practices. However they must be rigorously implemented and practised by all staff. SARS is a virus that is unforgiving of lapses in infection control.
References 1 WHO. WHO issues global alert about cases of atypical pneumonia. 12 March 2003. Geneva,
4
5
6
7
9 10
11
12
183
Switzerland. www.who.int/mediacentre/ releases/2003/pr22/en/print.html WHO. Summary of probable SARS cases with onset of illness from 1 November 2002 to 31 July 2003. 21 April 2004. Available from URL: http://www.who.int/csr/sars/country/table 2004_04_21/en/ WHO. Alert, verification and public health management of SARS post-outbreak period. 14 August 2003. Available from URL http://www.who. int/csr/sars/postoutbreak/en/ Seto WH, Tsang D, Yung RWH et al. Effectiveness of ‘droplets’ and ‘contact precautions’ in preventing nosocomial transmission of severe acute respiratory syndrome (SARS). Lancet 2003;361: 1519–20. Centers for Disease Control: Guideline for isolation precautions in hospitals. Am J Infect Control 1996:24: 24–52. Larson E, Kretzer EK. Compliance with handwashing and barrier precautions. J Hosp Infect 1995:30 Suppl: 88–106. Pritchett CJ, Seto WH. The surgeon-specific wound infection rate — a worthwhile strategy for Asian hospitals. Asian J Surg 1990;13: 121–4. Ching TY, Seto WH. The efficacy of the infection control liaison nurse. J Adv Nurs 1990;15: 1128–31. WHO. Practical Guidelines for Infection Control in Health Care Facilities, 8 December 2003. Centers for Disease Control: Severe Acute Respiratory Syndrome, Supplement I: Infection Control in Healthcare, Home and Community Settings. 8 January 2004. http://www.cdc.gov/ncidod/ sars/guidance/I/index.htm Ho PL, Tang X, Seto WH. SARS hospital infection control and admission strategies. Respirology 2003;8: S41–S45. Seto WH, Ching TY, Fung JPM et al. The role of communication in the alteration of patient care practices. J of Hospital Infection. 1989;14: 29–38.
Chapter 20
Antiviral Agents for SARS Frederick G Hayden and Mark R Denison
Introduction An urgent need exists for the identification and development of effective antiviral drugs for the prevention and treatment of SARS coronavirus illness. During the first worldwide outbreak of this novel viral pneumonia, a wide range of therapeutic interventions were used in illness management in efforts to prevent disease progression and salvage desperately ill patients. These included antivirals such as ribavirin, interferons, and HIV protease inhibitors, host immunomodulatory agents, particularly systemic corticosteroids, and biologicals including convalescent plasma. However, the uncontrolled nature of these interventions, the uncertain natural history of untreated SARS in different patient groups, and the potential for interactions among different modalities have resulted in a circumstance that no drug interventions of proven therapeutic or prophylactic value have been established to date. We have incorporated selected examples from studies of other respiratory viral infections, particularly for influenza for which notable successes have been achieved in the development of both vaccines and antivirals. However, the limited data available regarding the virological course of SARS and the pathogenesis of acute lung injury make comparisons to influenza and RSV illness speculative and potentially erroneous. It is hoped that advances in understanding of the disease pathogenesis and management
184
for each of these infections will provide insights for improving management of SARS. The converse is also true, in that a better understanding of SARS illness might open new possibilities for treatment of influenza and other respiratory viral illnesses.
General considerations Several agents with documented in vitro antiviral activity against SARS coronavirus are already commercially available therapeutic agents for other purposes and could be studied immediately for prevention and early treatment. In addition, multiple laboratories are engaging in screening programs for novel antivirals, so that it is likely that several alternative agents with activity in vitro and in relevant animal models will be available for testing within the next few years. The availability of multiple agents would enable the study of drug combinations in preclinical systems and ultimately in humans in efforts to enhance antiviral activity, reduce toxicity, and diminish the risk of antiviral resistance emergence.
Preclinical antiviral testing Considerable progress has already been made in developing the preclinical tools for SARS coronavirus that fostered the development of antivirals and vaccines for influenza and to a lesser extent RSV. In vitro screening for antiviral agents, based on
Antiviral Agents for SARS
inhibition of virus cytopathic effect or replication in cell culture, is being employed by multiple laboratories already. In addition, the development and testing of candidate antivirals and vaccines against SARS CoV will require the validation of animal models capable of reproducing viral replication, immune response, and disease patterns similar to human SARS. This is especially important since the SARS epidemic has remitted and it is therefore not possible to test the impact of vaccines and therapeutics directly in humans. Laboratory infection of cynomolgous macaques, ferrets, cats, and mice has been reported.1–3 Some pulmonary pathology and disease have been described for macaques and ferrets, but this finding has yet to be reproducibly demonstrated in multiple laboratories. Intranasal infection of mice has resulted in short-lived but significant virus replication, as well as humoral and cell-mediated immune responses in the absence of disease. Macaques have been used to show inhibition of SARS disease in pneumocytes by pegylated interferon-alpha,4 as well as by immune response to adenovirus vectors expressing the S and N proteins. The mouse replication model has been used to show immune responses to virus infection, inactivated virus vaccination, DNA-S protein vaccination, and DNA-N protein vaccination, as well as protection from SARS CoV challenge at the level of virus replication of the challenge virus for inactivated virus and DNA-S protein vaccination and antibody immunoprophylaxis.3,5,6 It is likely that approaches such as adaptation of the virus to animal hosts, the use of immunodeficient animals, different mechanisms for inoculation, and development of chimeric viruses, will result in models that are both informative and useful in understanding determinants of infection and for testing strategies for prevention and treatment of SARS CoV, should it re-emerge as a worldwide threat. Appropriate animal models will be essential in assessing putative antivirals, candidate vaccines, and
185
interventions directed against deleterious host immune response.
Clinical development Rapid clinical development of a therapeutic intervention is possible when knowledge of its epidemiology, clinical presentation, and natural history is available and studies can be performed in resource-rich medical environments. Predictive animal models, and in the case of several respiratory viral infections, the availability of experimentally induced human models of infection can speed the process. The clinical testing of antiviral agents and other interventions in SARS presents logistical problems, some that are shared with other respiratory viral pathogens and others that are unique to SARS. The relatively rapid development of the neuraminidase inhibitors for influenza depended heavily on prompt access to large numbers of ambulatory patients with characteristic illness. This was fostered by the typical seasonality of influenza, prospective virological surveillance, predictive clinical case definitions, the availability of rapid diagnostic assays, and the occurrence of sufficient influenza activity during the study periods. The occurrence of SARS, on the other hand, is unpredictable, both in location and time, making placement and conduct of clinical trials challenging. Because of the difficulties in trying to initiate studies in the context of an ongoing epidemic, trial capability would need to be prepositioned in areas with anticipated SARS activity and with sufficient, trained staff. SARS is non-specific in its initial manifestations, so that clinical case definitions may not be predictive unless there is a strong epidemiological link or substantial outbreak. According to current diagnostic criteria that include lower respiratory tract symptoms, the specificity of clinical diagnosis is high but sensitivity low in non-hospitalized persons during an outbreak.7 However, if one waits for characteristic lower respiratory tract manifestations, the opportunity to
186
Severe Acute Respiratory Syndrome
intervene early with an antiviral treatment may be lost. Furthermore, currently available SARS diagnostics lack sensitivity early in disease and are not applicable at the point-of-care. The sample size needed to test rigorously an intervention depends heavily on the endpoint selected and relatively infrequent events (e.g. death or intubation) would require large sample sizes to show clinically important differences. Frequently occurring events that are clinically relevant and accessible in both high- and low-technology settings (e.g. frequency and duration of serious hypoxaemia) are needed as endpoints for trials. Furthermore, once an outbreak is recognized, the use of appropriate infection control measures would probably reduce the number of affected persons available for study. Unfortunately, the current lack of a co-ordinated investigative network with adequate financial support, internationally agreed protocols and disease endpoints, mechanisms for rapid ethics approval, and in some countries, the reluctance to participate in controlled clinical trials or in studies utilizing placebos are serious limitations to implementing multicentre studies.8 On the other hand, the protracted course of viral replication in SARS patients, the evidence for increased viral load in the respiratory tract during the second week of illness, the substantial period between prodrome and lower respiratory phase, and the high frequency of severe and objective clinical outcomes (hypoxaemia, pulmonary infiltrates, respiratory failure, and death) should facilitate clinical studies of antivirals. In situations where exposure can be defined, the relatively prolonged incubation period of SARS provides an opportunity to test the strategy of postexposure prophylaxis. Such prophylaxis studies should provide early and clear evidence for an antiviral effect, since generally prevention with antivirals is more readily accomplished than treatment of established infection. Practical challenges to a post-exposure prophylaxis study would include identification of a target population
with sufficient risk of transmission despite appropriate infection control measures (e.g. health-care workers, household contacts). Treatment protocols need to consider the stage of disease, since early intervention with an antiviral is more likely to be beneficial. For example, it might be possible to use pre-emptive treatment during the prodromal phase to prevent progression to serious pneumonic involvement and its sequelae. In the absence of placebo-controlled trials, an international registry of SARS patients that gathers data on risk factors for severe disease, treatments, and outcomes would help to develop the foundation for protocol design.
Disease pathogenesis For effective clinical application of antiviral agents, an understanding of the relationship between viral replication and disease pathogenesis is essential. One consistent observation is that the earlier antiviral treatment is initiated in acute respiratory viral infections like influenza, the greater the likelihood of clinical benefit. A temporal relationship exists between influenza virus replication, cytokine and chemokine elaboration, and symptoms, such that early antiviral treatment blocks this sequence of events.9 The virological course of SARS is incompletely defined with respect to viral replication in different cell types and body sites over time. Prolonged detection of viral RNA in the upper respiratory tract, stool, and less often urine has been documented10 and high viral load in the respiratory tract and serum correlates with adverse prognosis.11,12 Such observations support the hypothesis that ongoing viral replication is probably contributing to lung injury and ultimately death.13 Histopathologic studies from patients dying of virologically confirmed SARS CoV infection have documented a range of findings that depended on the duration of illness and probably use of ventilatory support (Chapter 10).14–16 Diffuse alveolar
Antiviral Agents for SARS
damage, a finding in other viral pneumonias and many other conditions, has been seen in all cases including those dying early. It has been more prominent in cases of longer duration, and an organizing pattern of DAD occurs in illnesses of more than 10 days duration.15 SARS CoV-specific antibody titres generally start to rise about 14 days after onset of symptoms, and the mean time to seroconversion is 20 days in patients receiving ribavirin/corticosteroid regimens.10 The development of this adaptive immune response appears to be temporally linked to decreases in viral RNA loads in the upper respiratory tract, but it has been postulated to contribute to severe inflammatory damage in the setting of high viral burdens in the lower respiratory tract. However, convalescent plasma has been used in management with claims of benefit, and passive transfer studies in experimental murine SARS CoV infection indicate that convalescent plasma is highly protective.3 Neutralizing human monoclonal antibodies have been developed.17 One concern about antibody treatment is the possibility of antibody-mediated enhancement of disease. This concern highlighted by recent studies suggesting possible increased hepatic inflammation in ferrets immunized with recombinant modified vaccinia ankara expressing the SARS CoV spike and challenged with SARS CoV.17a It is not known if this occurs or if other host immune responses contribute to the disease process. The possibility that host immunopathological responses are contributing to pulmonary injury is an additional argument for early antiviral intervention in hopes that reducing antigenic load would decrease the risk of immune-mediated disease. The possible contribution of coinfections to disease pathogenesis in SARS remains incompletely resolved. Some reports have described high frequencies of co-infection with human metapneumovirus (hMPV).18 and in one instance reovirus.19 In the macaque model of SARS CoV infection, concurrent infection with
187
hMPV did not appear to increase disease severity,1 but a possible interaction remains uncertain in human infection and could be a confounder in studies of antiviral agents.
Routes of administration The pathogenesis of SARS and sites of replication in different stages of the illness are also critical factors in determining the optimal routes of antiviral drug delivery. For example, controlled studies in influenza showed that inhaled, but not intranasal, zanamivir was effective for both prophylaxis and treatment,20,21 observations consistent with initial acquisition and subsequent replication of influenza virus in the pharynx and/or lower airways. In contrast with SARS, other epidemiological observations indicate that influenza is transmitted commonly by small particle aerosols, so that drug delivery to the lower airways is essential. In contrast, rhinovirus infections can be prevented by intranasal administration of antiviral agents to the nasal passages,22 consistent with acquisition at that site through large droplets or hand contamination via fomites and subsequent selfinoculation. The mechanisms of SARS CoV transmission and disease pathogenesis will determine whether topical application of antivirals to the respiratory tract might be protective. In the event that multiple sites (pharynx, tracheobronchial tree, gastrointestinal tract, conjunctiva) serve as portals of entry, protection of the nasopharynx by intranasal administration would probably be ineffective for treatment or prophylaxis. Because SARS CoV infection is associated with viraemia and replication in multiple sites including both upper and lower respiratory and gastrointestinal tracts, systemic antiviral drug administration will presumably be necessary once illness develops. In this regard the postulated dissemination of virus from mucosal sites to the pulmonary parenchyma during the prodromal phase of illness should enable early antiviral
188
Severe Acute Respiratory Syndrome
treatment to reduce the risk of developing severe pneumonic disease.
Immunomodulators Glucocorticoids have been used frequently in varying regimens for managing SARS patients and acutely appear to reduce fever, pulmonary infiltrates, and respiratory distress. In addition, one report from Guangzhou found that the frequency of death tended to be higher in cohorts of SARS patients not given corticosteroids during the first 14 days after symptom onset (4/70) compared to a cohort given high-dose methylprednisolone 160–1000 mg/day if they experienced persistent fever or increasing infiltrates (0/60), but lower than a cohort given methylprednisolone 80–160 mg/day for 2–3 days (7/60).23 However, inhibition of host responses with glucocorticoids or other non-selective immunomodulators could have deleterious consequences on viral replication and possibly long-term clinical outcomes. Use may foster increased viral replication and, as the taper occurs, also allow expression of pathological host immune responses. The uncontrolled use of such agents could certainly confound the study of candidate antiviral agents. In one case series from Hong Kong, relatively high doses of hydrocortisone were administered in conjunction with systemic ribavirin for 10 days, followed by a tapering regimen of glucocorticoids.10 The patients, who were admitted to hospital an average of 3–4 days after symptom onset, promptly defervesced but 85% developed recurrent fever on average of 8–9 days after symptom onset in association with approximate 100-fold increases in nasopharyngeal viral RNA loads. Many patients developed new pulmonary infiltrates, worsening gas exchange, and in a fraction acute respiratory distress syndrome. Whether these findings reflect the natural history of infection or possibly adverse consequences of early glucocorticoid treatment remains to be determined. One small trial found that early
administration of hydrocortisone (<7 days from illness onset) was associated with higher plasma SARS CoV RNA levels in the second and third week of illness compared to placebo.23a Glucocorticoid use could also predispose patients to nosocomial complications and secondary infections including aspergillosis,24 as well as late complications including prolonged weakness and avascular necrosis. One study found that the use of pulsed corticosteroids in SARS was highly associated (odds ratio 26.0, 95% CI 4.4 to 155) with 30-day mortality,11 although a causal relationship could not be ascertained in this retrospective analysis. Another retrospective study reported that pulsed corticosteroids were more effective than non-pulsed in SARS management (see Table 20.2).25 Consequently, controlled studies are needed to determine not only the shortterm effects of corticosteroid use but also the long-term outcomes in SARS. Other studies have demonstrated a deleterious effect of corticosteroids on viral replication in animals with experimentally induced coronavirus infections,26,27 and in several human respiratory viral infections. Systemic glucocorticoids delay viral clearance in the resolution of RSV bronchiolitis28 and have been associated with impaired viral clearance and increased mortality in animal models of influenza and murine pneumovirus infection.29 Intranasal steroids are associated with prolonged viral recovery in rhinovirus colds in adults and with increased risk of acute otitis media in children. Whether potent immunosuppressive agents such as glucocorticoids will provide clinical benefit, particularly in conjunction with an effective antiviral agent, needs controlled clinical testing in SARS. On the other hand, targeted immune response modifiers that down-regulate cellular responses evoked by viral infection can provide direct antiviral effects and therapeutic benefit. For example, up-regulation of the mitogen-activated protein kinase (MAPK) p38 appears to be central in the
Antiviral Agents for SARS
replication of the CoV murine hepatitis virus in certain cells.30 Under in vitro conditions a selective p38 MAPK inhibitor (designated SB 203580) can diminish activation, reduce associated increases in IL-6 elaboration, and modestly decrease productive virus yield.30 Whether SARS CoV utilizes the p38 pathway needs to be determined as other inhibitors of this enzyme are available. Of note, the antioxidant N-acetylcysteine, which is commercially available, has been reported to inhibit IL-1b activation of p38 MAPK and diminish production of reactive oxygen species in vitro.31 Immunomodulatory agents with other mechanisms of action also warrant study. One T-cell modulator termed PR-879–317A appeared effective in single doses in increasing survival and reducing liver viral titres and damage in murine hepatitis virus infection.32 Severe avian flu has been associated with tumour necrosis factor (TNF) dysregulation.16,33 Therapies directed against TNF have also been proposed for SARS, in part because of the availability of specific blockers (adalimumab, etanercept, and infliximab) and the findings that antiTNF antibody reduces illness and lung pathology in animal models of influenza and RSV without adverse effects on viral clearance.34
Antiviral resistance Although no specific antivirals for SARS CoV are available at present, it is important to recognize that antiviral resistance can emerge rapidly and have important clinical and epidemiologic consequences. When a selective antiviral agent is available for SARS, monitoring of resistance development and its clinical and epidemiologic consequences will be essential.
Antiviral targets Theoretically, all stages of the coronavirus life-cycle are targets for therapeutic intervention. Studies over the past 20 years into
189
coronavirus replication and cell biology have defined specific stages in the coronavirus life-cycle that may be particularly attractive for interference. In addition, the rapid sequencing, characterization, and emerging structural and replication studies of SARS CoV have defined additional targets for possible antiviral interference and have accelerated the pace of research in coronavirus biology, replication and pathogenesis.
Coronavirus life-cycle Coronavirus replication and pathogenesis have been best studied in the model virus mouse hepatitis virus (MHV),35 but the principles are probably directly applicable to SARS CoV and will be emphasized where data on SARS CoV have been generated. The coronavirus life-cycle may vary among viruses and in different cell types, but the complete MHV replication cycle is 6–12 hours in naturally permissive cells and probably less than 24 hours for all coronaviruses in culture. Recent studies of SARS CoV replicase proteins and growth have shown detection of replication complexes by 4 hours p.i. and new virus by 8–12 hours.35a,35b Coronavirus infection is initiated by binding of the spike glycoprotein (S) to specific receptors and for SARS CoV a functional receptor has been identified to be angiotensin-converting enzyme 2 (ACE-2).36 SARS CoV virus entry may involve both pH-dependent (endocytic) and pH-independent (direct fusion) mechanisms for entry into cells (reviewed in REF).36c Peptidic fusion inhibitors active in the micromolar range have been described for SARS CoV.36a,36b Following virus entry, coronavirus RNA transcription and replication occur on membrane-associated replication complexes in the cell cytoplasm.37,38 Proteins encoded by the replicase gene (gene 1) are thought to direct membrane recruitment, replication complex formation, and all stages RNA synthesis, including negative-strand RNA synthesis, subgenomic
190
Severe Acute Respiratory Syndrome
mRNA transcription, and genome replication. The formation of complexes involves virus induction of cellular autophagy, a process resulting in double membrane vesicles used by MHV and SARS CoV in the formation of replication complexes.38a,38b,39 Replication products are delivered to sites of assembly by an unknown mechanism that may be mediated by viral components of the replication complexes40 followed by assembly and budding of progeny virions in the ER-Golgi-intermediate compartment (ERGIC)41 by a process that involves interactions of genome RNA, and the nucleocapsid (N), membrane (M), and envelope (E) proteins.41a–41d Release of MHV from cells occurs by fusion of secretory vesicles with the plasma membrane, a process that does not require cell lysis.42 Expression of the S protein on the cell surface or cell-bound virus mediates syncytia formation with adjacent cells in culture for coronaviruses that contain a fusion activity.
Target selection SARS coronavirus genome organization and gene products have been summarized
in Chapter 8. Based on the understanding of coronavirus biology and protein functions as described above, there are clear targets for the development of potential antivirals, the most obvious of which involve well-defined stages in the virus life-cycle and proteins of known functions (Table 20.1). The spike protein has been targeted for development of both antivirals and vaccines. Antibodies to spike have been shown to be protective in a mouse replication model when induced either by inactivated virus, SARS CoV infection, or DNA plasmid expression.3,5 Thus, it is possible that peptides that bind to S may be directly inhibitory to virus binding and entry. The spike glycoprotein also contains a fusion activity that may be critical for virus entry and genome uncoating, as well as possible roles in virus pathogenesis. Inhibitors of the fusion peptide or fusion activity might allow for virus binding but prevent productive replication in cells.36b,36c In addition, such inhibitors might be developed to limit any pathology caused by S protein-mediated cell–cell fusion. The identification of ACE2 as a functional receptor also provides a wellcharacterized target for possible antivirals.
Table 20.1 Targets for inhibition of SARS CoV replication and pathogenesis Stage of replication
Targets
Potential inhibitors
Virus binding, entry and uncoating
Spike (S) Receptor (ACE-2)
Replicase protein processing
Papain-like proteinase (PLP) 3C-like proteinase (3CLpro) Cleavage sites Inhibitors of autophagy or membrane synthesis
Antibodies Peptides ACE inhibitors Fusion inhibitors Proteinase inhibitors Cleavage site analogues
Replication complex formation
RNA synthesis
Expression of structural and accessory proteins Virus assembly and release
RNA-dependent RNA polymerase Helicase RNA processing enzymes Genes 2–9, spike, membranes, envelope, nucleocapsid, accessory proteins ER trafficking Microtubule trafficking Secretory pathway
Kinase inhibitors Cholesterol synthesis inhibitors RNAi Nucleoside analogues RNAi RNAi Antibodies Microtubule inhibitors Antibodies
Antiviral Agents for SARS
Continuous expression and processing of replicase polyproteins are required for ongoing viral RNA synthesis and virus growth. This is a vulnerable stage of coronavirus replication and thus proteinase inhibitors have the potential to be highly effective antivirals. Inhibitors of cysteine proteinases have been shown to abort replication of SARS CoV and other coronaviruses,43,44 and uncontrolled case studies suggest a possible effect of aspartyl proteinase inhibitors.45 The availability of crystal structures of several coronavirus proteinases should help direct rational design of effective inhibitors of these proteins. Other replicase gene proteins including the RNA-dependent RNA polymerase (pol) and helicase have been experimentally confirmed, and a bioinformatics prediction of pol structure should allow testing of potential nucleoside and non-nucleoside inhibitors in culture and experimental animals.46 These replicase gene proteins represent only a fraction of the proteins processed from the replicase polyprotein; the majority remain to be determined. Consequently, it is likely that additional targets will be identified in the immediate future. Coronavirus replication requires the induction of double membrane vesicles in the cytoplasm and involves pathways of cellular autophagy. While targeting of constitutive or induced cell functions is potentially fraught with obstacles, it is important to define steps of viral interaction with the host cell. It is possible that transient inhibition of cell pathways such as autophagy could limit replication without undue toxicity and allow for development of immune responses and clearance of virus. The development of inhibitors of cellular autophagy might complement approaches targeting direct viral functions.
Specific antiviral agents A number of compounds that inhibit replication of SARS CoV or other CoVs in vitro have been reported (see Appendix). Few of these agents have been administered
191
to SARS patients. One in vitro study of >10,000 compounds, including approximately 200 FDA-approved drugs and 500 protease inhibitors, found several (valinomycin, aescin, resperpine) that inhibited SARS in Vero cells with varying degrees of selectivity, whereas other agents were inactive at 10 mM concentrations.46a Both the antimalarial chloroquine and the antihelminthic niclosamide are active at micromolar concentrations (Table 20.2) but evidence for in vivo activity for these agents is currently unavailable. The following sections highlight the status of selected examples of candidate antiviral agents for SARS management that have received clinical use.
Ribavirin The nucleoside analogue ribavirin has a broad spectrum of antiviral activity in vitro encompassing many RNA viruses including some CoVs47,48 and hMPV.49 One in vitro study found that ribavirin concentrations of 2.5 mg/ml inhibited growth of feline infectious peritonitis virus in feline kidney cells, although selectivity was limited in that sevenfold higher concentrations inhibited cell proliferation.50 A synergistic interaction between ribavirin and rHuIFN-a was found for this CoV. Ribavirin was also found to be inhibitory in vivo in experimental murine hepatitis virus infections.48,51 However, cell culture-based assays have found no evidence for a selective antiviral activity of ribavirin against SARS CoV in vitro (see Appendix p. 255).52–54 Concentrations as high as 2000 mg/ml lack inhibitory activity for SARS CoV in Vero cells,55 whereas much lower concentrations are inhibitory for several haemorrhagic fever viruses in the same cell type. However, ribavirin has modest antiviral effects in FRhK-4 cells,56 although a recent study found inhibitory effects (EC50s of 2.2–9.4 mg/ml) against SARS CoV in other cell types (MA104, PK-15, CaCo2, CL14, human kidney), as well as synergism between ribavirin and interferonb.56a Screening of combinatorial libraries
192
Severe Acute Respiratory Syndrome
has identified inhibitors of SARS CoV protease, helicase, and S-mediated cell entry.57 Modelling studies of the catalytic domain of the SARS coronavirus RNA polymerase have predicted that resistance to ribavirin may relate to alterations in the conserved motif A that controls rNTP binding and the fidelity of polymerization.58 Oral and intravenous ribavirin were used extensively in multiple centres for treating SARS patients. Unfortunately, multiple different dose regimens were employed, and none of the published reports to date have provided placebo-controlled evidence that treatment of SARS with ribavirin is beneficial or at least not harmful (Table 20.2).58 One case series of 31 hospitalized patients with probable SARS reported that a regimen of systemic ribavirin for 10–14 days and of high-dose glucocorticoids in a tapering course over 21 days was associated with no instances of mechanical ventilation or death.59 Pulse methylprednisolone (500 mg twice daily for 2 days) was used in 11 of 30 treated patients for severe disease. One retrospective report from Hong Kong concluded that a delayed initiation of combined treatment with systemic corticosteroids and ribavirin increased the likelihood of severe disease requiring ICU admission or mechanical ventilation.60 In contrast, use of ribavirin without corticosteroids has not been associated with apparent clinical benefit. A group of 40 patients randomized to intravenous ribavirin in Guangzhou had persistent fever and increasing infiltrates,23 and an uncontrolled study from Toronto found that high-dose intravenous ribavirin treatment tended to be associated with poor outcomes.61 Ribavirin use in 110 Canadian patients was associated with dose-related haemolytic anaemia in 61%, of whom 28% required transfusion support.62 Hypomagnesaemia and hypocalcaemia were also common in ribavirin recipients. Little information is available regarding in vivo antiviral effects in SARS patients. One regimen of intravenous ribavirin (8
mg/kg/8 h for 14 days) and glucocorticoids (hydrocortisone 600 mg/day for 10 days followed by oral prednisolone taper to day 21) was associated with prompt defervescence within 48 hours in 75 treated patients, but fever recurred in 85% at an average of 9 days after onset of symptoms in temporal association with 100-fold increases in nasopharyngeal viral RNA levels at day 10 and with radiographic worsening in 80% of patients.10 Progression to ARDS occurred in 20%, and five patients (7%) died. Clearly ribavirin did not prevent the increases in viral load, nor did the combination of ribavirin and corticosteroids appear to prevent serious lung injury. Furthermore, in patients dying with SARS despite ribavirin treatment, high viral RNA levels have been observed in post-mortem lung tissues.13 SARS CoV RNA loads ranged from 7 ¥ 106 to 4 ¥ 109 copies per gram of lung tissue in those dying within 15 days of illness onset and from 3 ¥ 104 to 3 ¥ 107 in those succumbing later. Such findings indicate that ribavirin did not exert substantial inhibitory effects against SARS CoV in treated patients. Further animal model studies would be helpful in assessing whether ribavirin has antiviral action against SARS CoV in vivo before controlled human trials might be warranted.
Interferons Type I interferons (IFNs) inhibit a wide range of RNA and DNA viruses63 including human respiratory coronaviruses64 and SARS coronavirus in vitro (see Appendix). In vitro testing in both monkey kidney (Vero) and intestinal epithelial (CaCo2) cell lines found inhibition of SARS coronavirus replication with recombinant IFN-a and -b.65 IFN-b was more potent in these studies than IFN-a-2. Another in vitro study confirmed the greater potency of IFN-b-1b and also found that IFN-a-n1 and -n3 were inhibitory.54 These differences could relate to differences in IFN-cell interaction or possibly differential up-regulation of intracellular
Table 20.2 Selected case series of outcomes in SARS treated with ribavirin and/or corticosteroids Location
No. patients
Comorbid conditions
CXR infiltrates (bilateral)
Regimen
Outcomes
So et al. 200358
Hong Kong, PYN Eastern Hospital
31
4 (13%)
100% (71%)
Ribavirin 1200 mg/d IV for ≥ 3 days, PO 2400 mg/d for 10–14 days (97%) + MP 3 mg/kg/d IV for 5 days + 1 mg/kg/d for 5 days + PO prednisolone taper for 21 days (total) (97%) + pulse MP prn (35%)
Rapid, sustained response 55% Supplemental O2 52%, NIPPV 13%, Intubation 0%, Death 0%
Peiris et al. 200310
Hong Kong, United Christian Hospital
75
13 (17%)
71% (21%)
Ribavirin 8 mg/kg/8 hrs IV for 14 days + HC 600 mg/d IV for 10 days + oral prednisolone taper for 21 days (total) + pulse MP 500 mg for 2–3 doses prn
98% afebrile with 48 hrs but fever recurrence in 85% (mean ± SD 9 ± 3 days), Radiologic worsening 80% (7 ± 2 days), ARDS 20%, 44% SaO2 < 90% (RA) at 9 ± 4 days Deaths 7%
Tsang et al. 200311
Hong Kong, Princess Margaret Hospital
40 ± 14 (62 PCR -)
10 (5%)
96% (21–42%)
As above
30% ICU (vs 10%) 24% Intubation (vs 8%) 13% Death (vs 3%)
Booth et al. 200361
Toronto
147
11% DM 8% Cardiac 6% Cancer 2% Other
75% (29%)
Ribavirin IV 2 gm load + 4 gm/d for 4 days + 1.5 gm/d for 3 days (88%) ± HC 20– 50 mg/d for 10 days (40%)
Death 6.5%, ICU admit 20% Ventilation 14%, Supplemental oxygen 34%, Radiologic progression 31%
Zhao et al. 200323
PRC, Guangzhou
40
NS
49% overall
IV Ribavirin 0.4–0.6 gm/day (no steroids for 14 days)
30
NS
Death 2 (5%), mechanical ventilati on 3 (7.5%), time to respiratory improvement 11 ± 7 days Death 2 (7%), mechanical ventilation 2 (7%), time to respiratory improvement 10± 5 days Continued p.194
193
MP, methylprednisolone; NIPPV, non-invasive, positive-pressure ventilation.
Interferon-a 3.0 M/day (no steroids for 14 days)
Antiviral Agents for SARS
Study
194
Study
Location
No. patients
Comorbid conditions
60
NS
CXR infiltrates (bilateral)
Regimen
Outcomes
Interferon-a 3.0 M/day (75% of patients) plus MP 160–1000 mg/day for 5–14 days if lack of response
Death 0, mechanical ventilation 0, time to respiratory improvement 6 ± 3 days
Sung et al. 200487
Hong Kong, Prince of Wales Hospital
138
19 (14%)
78% (45%)
PO Ribavirin 3600 mg/d + Prednisolone 0.5–1 mg/kg/d (n = 94) or IV ribavirin 1200 mg/d + hydrocortisone 300 mg/d (n = 44) + pulse MP 500 mg/d for 2–3 days prn (n = 107)
Failure to improve on ribavarin + low-dose corticosteroids 82% ICU admit 27%, mechanical ventilation 15%, death 11%
Ho et al. 200325
Hong Kong, Queen Mary and Queen Elizabeth Hospitals
72
1.5%
Score 2 (IQR, 1–4)
Ribavirin IV 8 mg/kg tid for 7 days, then PO 1.2 gm tid for 10–14 days plus HC 2 mg/kg qid or 4 mg/kg tid by PO prednisolone 2 mg/kg in taper or MP IV 500 mg/day for 5–7 days or 1000 mg/day for 3 days followed by PO prednisolone 50 mg bid tapering to 20–30 mg/day by day 21+ pulse MP prn
ICU admit 17%, mechanical ventilation 8%, death 6% overall. Trends towards less frequent ICU use and less O2 use in those receiving pulsed steroids
Choi et al. 200388
Hong Kong, Princess Margaret Hospital
267 (85% confirmed)
15 (7%)
96% (34% multilobar)
Ribavirin 24 mg/kg/d IV for 14 days + HC 10 mg/kg/d IV + pulse MP 500–1,000 mg/ day for 2–3 doses prn
ICU admit 26%, mechanical ventilation 21%, death 12%
Severe Acute Respiratory Syndrome
Table 20.2 continued
Antiviral Agents for SARS antiviral proteins.65a Compared to activity against a control virus (VSV), the relative inhibitory effects of IFN-a-2 and IFN-b were similar in Vero cells but IFN-b was about 30fold more active in CaCo2 cells.65 IFN-a has also been shown to play a role in controlling coronavirus disease in animals. In particular, type I IFN inhibits infectious bronchitis virus replication and associated respiratory illness in chickens66 and coronavirus hepatitis in mice.67,68 Intranasal dosing of the IFN inducer 7-thia8-oxoguanosine protected against lethal CoV infection in rats.69 More recently, prophylactic intramuscular injection of pegylated IFN-a 2b 3 days before SARS coronavirus infection reduced lung viral titres by approximately 1000-fold and histopathological changes by 80% in experimentally infected cynomolgus macaques.4 Administration at 1 and 3 days post infection was associated with intermediate effects. In humans, intranasal IFN-a-2 partially protected against experimental human respiratory CoV colds caused by type 229E.70,71 Protection appeared to be dose-dependent, such that high doses (4 MU 3 times daily beginning 1 day before viral inoculation) prevented both infection (55% reduction relative to placebo) and illness (85% reduction), whereas a lower dose (2 MU once daily beginning 7 days before challenge) did not reduce infection but diminished illness (44% reduction). However, it remains to be established that intranasal administration of IFN can protect against natural human respiratory CoV infection,72 let alone SARS CoV, or that the mucosa of the nasal passages and nasopharynx is the principal initial site of SARS CoV acquisition. IFN-a may modulate some inflammatory effects by decreasing cell proliferation and increasing circulating levels of TNF-a receptor.63,73 Thus IFN-a may have both an antiviral and anti-inflammatory role in the treatment of SARS-related CoV infection. On the other hand, some strains of animal CoVs induce high levels of IFN-a. It
195
remains to be determined whether CoVs inducing high-level IFN-a are more pathogenic and whether high-level IFN-a production is a protective host response to contain CoV infection, or whether it plays a role in disease production. A number of viruses, including HCV and influenza, have been shown to antagonize IFN production or action as a mechanism to evade innate immunity. Various systemically administered IFN-as have been used for a number of therapeutic indications including chronic hepatitis C virus, chronic hepatitis B virus, and various malignancies. Although tolerable in most patients during months of dosing, severe adverse effects including cytopenias, autoimmune phenomena, and depression with suicidal ideation may occur. In addition, very rare cases of pneumonitis, similar to bronchiolitis obliterans organizing pneumonia (BOOP), have been reported during the use of IFN-a.74,75 It is unknown whether the pulmonary disease associated with SARS will predispose to these possible pulmonary complications of IFN-a. Given the relatively short IFN courses anticipated for use in SARS and the presence of preexisting leukopenia and thrombocytopenia, the greatest concern would be cytopenias. A report from Guangzhou indicated that mortality was low in a treatment regimen incorporating intramuscular rIFN-a 3.0 MU/ day and restricted use of steroids during the first 14 days of illness (2 deaths among 30 treated) and that mortality was absent (0/60) in a regimen combining IFN-a (used in 75% of patients) and high-dose corticosteroids if progression.23 Preliminary observational clinical experience with subcutaneously administered IFN-alfacon-1 for 8–13 days (doses of 9 mg/day for minimum of 2 days followed by dose increase to 15 mg/day for 8 days, if no apparent clinical response observed) in combination with systemic corticosteroids found acceptable tolerability and suggestion of improved clinical outcomes compared historically to systemic corticosteroids alone.76 Treatment
196
Severe Acute Respiratory Syndrome
was initiated at a median of 8 days post symptom onset (range 3–10 days). Among nine IFN-alfacon-1 and corticosteroidtreated patients, adverse events were noted in only three (two fever, one neutropenia that resolved post drug cessation) and times to 50% resolution of lung X-ray abnormalities (median 4.0 vs 11.5 days) and need for supplemental oxygen resolved faster (median 10 vs 16 days) compared to 13 historical controls receiving steroids alone. For six critically ill patients treated for late-stage disease, four (67%) died despite combination therapy, an observation that emphasizes the need for early treatment. Further prospective, controlled trials of IFNs are needed in early-stage SARS. The appropriate therapeutic use of exogenous IFNs will depend on a more complete understanding of the endogenous cytokine and chemokine responses in SARS patients, particularly in the lower respiratory tract and during different stages of illness. For example, uncomplicated influenza in adults is associated with brisk IFN responses in the upper respiratory tract and blood. In contrast, a deficient IFN response was described in one older study of victims dying of influenza viral pneumonia.77 IFN responses appear to be down-regulated during RSV infection, but no controlled studies of exogenous IFNs have been reported in treatment of either RSV or influenza viral lower respiratory tract disease.
Protease Inhibitors SARS CoV encodes a single 33.8-kDa main protease (Mpro)78,79 that is similar to the 3C proteinase in picornaviruses and hence often designated 3C-like (3CLpro). This enzyme has been described as a chymotrypsin or papain-like protease; it possesses a Cys-His catalytic dyad and cleaves the CoV polyprotein in at least 11 conserved sites.80 A three-dimensional model, based on the crystal structure of human respiratory CoV 229E and porcine transmissible gastroenteritis virus (TGEV), found that
the substrate binding site of the SARS CoVencoded 3CLpro is very similar to that of 229E.80 In addition, expressed SARS CoV 3CLpro mediated the cleavage of a TGEV protease substrate. More recently, solution of the crystal structure of the SARS CoV 3CLpro found that it exists as a dimer and possesses pH-dependent enzymatic activity.79 These advances will facilitate the rational design and selection of protease inhibitors. Initial modelling suggested that the picornavirus 3C proteinase inhibitor ruprintrivir (formerly AG7088), which is inhibitory for rhinovirus replication in vitro81 and in experimentally infected volunteers after intranasal delivery,82 would be a reasonable starting molecule for the design of inhibitors.80 Most available human immunodeficiency virus (HIV) inhibitors have been found to lack in vitro inhibitory activity against SARS CoV in Vero cell systems54 (C Laughlin, personal communication). In contrast, one report indicated that nelfinavir, but not other HIV protease inhibitors, was active against SARS CoV in Vero cells.82a One group has reported inhibition of SARS cytopathic effect with the HIV aspartate protease inhibitor lopinavir/ritonavir in fetal rhesus monkey kidney cells (see Appendix).45 Other antiretrovirals (zidovudine, stavudine, nevirapine, and abacavir) were inactive under these test conditons. Combinations of lopinavir 1 mg/ml and ribavirin 6.25 mg/ml were reported to inhibit plaque formation by low but not higher viral inocula in this cell system.45 In an openlabel observational study, the addition of lopinavir/ritonavir 400/100 mg every 12 hours to a standard regimen of ribavirin and corticosteroids appeared to reduce the likelihood of severe hypoxaemia and death at 21 days (2.4% of 41 SARS patients) compared historically to ribavirin and corticosteroids (28.8% of 111 patients).45 Compared to the historical controls, the lopinavir/ribavirin/corticosteroid-treated patients had significantly lower frequencies of diarrhoea 24% versus 62%), recurrent
Antiviral Agents for SARS
fever (39% versus 61%), and worsening of pulmonary infiltrates (51% versus 81%). In patients treated early before use of pulse methylprednisolone, the lopinavir/ ribavirin/corticosteroid groups showed reductions in both pulse methylprednisolone use and in nasopharyngeal titres of SARS CoV RNA. Tolerance of the combination was acceptable; mild adverse reactions developed in 27% and included gastrointestinal upset, liver dysfunction, headache, and blurred vision. An earlier report from this group found that the addition of lopinavir/ritonavir to ribavirin/corticosteroids was associated with reduced mortality (2.3% versus 15.6%) as initial therapy but not when used as salvage treatment (12.9% versus 14.0%) compared to historical controls.83 The biological basis for lopinavir’s antiviral action against SARS CoV is uncertain, and further animal model studies and controlled clinical trials of this intervention are needed.
Neutralizing antibodies Several groups have developed monoclonal antibodies directed against the major surface (S) glycoprotein that neutralize SARS CoV infectivity and have the potential to be effective for passive immunoprophylaxis and possibly therapy.17,88a,89,90 Prophylactic intraperitoneal administration of one anti-S human monoclonal antibody (designated CR3014) at a dose of 10mg/kg resulted in over 1000-fold reductions in lung SARS CoV titers, prevented pharyngeal shedding, and completed protected against histopathologic changes in ferrets.88 Two other human neutralizing antibodies (designated 68 and 201) directed at different S epitopes at a dose of 40 mg/kg completely protected against SARS CoV replication in mice (over 106-fold reduction in titers), and doses as low as 1.6 mg/kg resulted in over 100-fold reductions in lung titers.89 As expected, smaller reductions in nasal viral levels were observed. One human monoclonal (designated S3.1) has neutralizing activity
197
up to 300-fold greater than that of convalescent serum and showed dose-related protection against pulmonary viral replication in mice.90 Furthermore, passive transfer of murine hyperimmune serum with neutralizing activity prevented SARS CoV replication in the lower respiratory tract and to a lesser extent in the nares of mice.3 Uncontrolled clinical experience suggested that infusions of convalescent plasma from recovered SARS patients may have provided clinical benefit, particularly if used early in disease.91 Patients deteriorating on ribavirin and corticosteroid therapy were given convalescent plasma (4–5ml/kg) of uncertain neutralizing titer. Overall 36% of 92 had a good outcome, as defined by hospital discharge by day 22, and 12.5% died. The 48 given plasma before day 14 had a higher frequency of good outcome (58%) than those given it later (16%), as well as a trend towards lower mortality (6% versus 22%). PCR positivity and SARS seronegativity at the time of infusion were associated with better outcomes, although no virologic data were included in this report.91 These findings are encouraging with respect to the potential value of the human anti-S monoclonals, but concerns regarding potency for a range of SARS CoV isolates and the potential for immune enhancement need to be addressed.
Nitric oxide Nitric oxide (NO) has a wide range of biologic activities including vasodilation, modulation of innate immune responses and inflammatory processes leading to tissue injury, and direct antiviral effects. Glycyrrhizin, which has anti-SARS in vitro, induces nitrous oxide synthase in Vero cells.53 SARS CoV replication was also inhibited when the NO donor DETA NONOate was used in vitro.53 Similarly, another NO donor S-nitroso-N-acetyl-penicillamine (SNAP) resulted in concentration-dependent inhibition of SARS CoV in Vero cells.92 Upregulation of inducible NO synthase by
198
Severe Acute Respiratory Syndrome
interferon exposure was associated with a modest (<1 log10) reduction in virus yield. The levels of NO expression in SARS patients are uncertain. Inhalation of NO has been used in patients with acute respiratory distress syndrome. One open-label cohort study of 14 SARS patients with severe lung injury (mean 29 days from diagnosis) reported sustained improvements in arterial oxygenation and lung infiltrates in 6 patients receiving NO inhalation in a tapering regimen over 3–7 days compared to 8 patients without NO treatment.93 No virologic data were provided in this report, but further studies in animal models and more severely affected patients appear warranted.
References 1
2
3
4
5
6
7
8
9
Fouchier RA, Kuiken T, Schutten M et al. Aetiology: Koch’s postulates fulfilled for SARS virus. Nature 2003;423: 240. Martina BEE, Haagmans BL, Kuiken T et al. Virology: SARS virus infection of cats and ferrets. Nature 2003;425: 915. Subbarao K, McAuliffe J, Vogel L et al. Prior infection and passive transfer of neutralizing antibody prevent replication of severe acute respiratory syndrome coronavirus in the respiratory tract of mice. J Virol 2004; 78: 3572–7. Haagmans BL, Kuiken T, Martina BE et al. Pegylated interferon-alpha protects type 1 pneumocytes against SARS coronavirus infection in macaques. Nature Medicine 2004;10: 290–3. Yang ZY, Kong WP, Huang Y et al. A DNA vaccine induces SARS coronavirus neutralization and protective immunity in mice. Nature 2004;428: 561–4. Kim TW, Lee JH, Hung CF et al. Generation and characterization of DNA vaccines targeting the nucleocapsid protein of severe acute respiratory syndrome coronavirus. J Virol 2004;78: 4638–45. Rainer TH, Cameron PA, Smit D et al. Evaluation of WHO criteria for identifying patients with severe acute respiratory syndrome out of hospital: prospective observational study. BMJ 2003;326: 1354–8. Muller MP, McGeer A, Straus SE et al. Clinical trials and novel pathogens: lessons learned from SARS. Emerg Infect Dis 2004;10: 389– 94. Hayden FG, Treanor JJ, Fritz RS et al. Use of the oral neuraminidase inhibitor oseltamivir
in experimental human influenza. JAMA 1999;282: 1240–6. 10 Peiris JS, Chu CM, Cheng VC et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;361: 1767–72. 11 Tsang OTY, Chau TN, Choi KW et al. Coronavirus-positive nasopharyngeal aspirate as predictor for severe acute respiratory syndrome mortality. Emerg Infect Dis 2003;9: 1381–7. 12 Ng EKO, Hui DS, Chan KCA et al. Quantitative analysis and prognostic implication of SARS coronavirus RNA in the plasma and serum of patients with severe acute respiratory syndrome. Clin Chem 2003;49: 1976–80. 13 Mazzulli T, Farcas GA, Poutanen SM et al. Severe acute respiratory syndrome-associated coronavirus in lung tissue. Emerg Infect Dis 2004; (Online) 10. 14 Nicholls JM, Poon LL, Lee KC et al. Lung pathology of fatal severe acute respiratory syndrome. Lancet 2003;361: 1773–8. 15 Franks TJ, Chong PY, Chui P et al. Lung pathology of severe acute respiratory syndrome (SARS): a study of 8 autopsy cases from Singapore. Hum Pathol 2003;34: 743–8. 16 To KF, Chan PK, Chan KF et al. Pathology of fatal human infection associated with avian influenza A H5N1 virus. J Med Virol 2001;63: 242–6. 17 Sui J, Li W, Murakami A et al. Potent neutralization of severe acute respiratory syndrome (SARS) coronavirus by a human mAb to S1 protein that blocks receptor association. PNAS 2004;101: 2536–41. 17a Czub M, Weingartl H, Czub S, He R, Cao J. Evaluation of modified vaccinia virus Ankara based recombinant SARS vaccine in ferrets. Vaccine 2005;23: 2273–9. 18 Kuiken T, Fouchier RAM, Schutten M et al. Newly discovered coronavirus as the primary cause of severe acute respiratory syndrome. Lancet 2003;362: 263–70. 19 The Chinese SARS Molecular Epidemiology Consortium. Molecular evolution of the SARS coronavirus during the course of the SARS epidemic in China. Science 2004;303: 1666–9. 20 Hayden FG, Gubareva LV, Monto AS et al. Inhaled zanamivir for preventing influenza in families. N Engl J Med 2000;343: 1282–9. 21 Hayden FG, Osterhaus ADME, Treanor JJ et al. Efficacy and safety of the neuraminidase inhibitor zanamivir in the treatment of influenza virus infections. N Engl J Med 1997;337: 874–9. 22 Hayden FG, Albrecht JK, Kaiser DL et al. Prevention of natural colds by contact prophy-
Antiviral Agents for SARS
laxis with intranasal alpha 2-interferon. N Engl J Med 1986;314: 71–5. 23 Zhao Z, Zhang F, Xu M et al. Description and clinical treatment of an early outbreak of severe acute respiratory syndrome (SARS) in Guangzhou, PR China. J Med Microbiol 2003;52(Pt 8): 715–20. 23a Lee N, Allen C, Hui DS et al. Effects of early corticosteroid treatment on plasma SARSassociated Coronavirus RNA concentrations in adult patients. J Clin Virol 2004;31: 304–9. 24 Wang H, Ding Y, Li X et al. Fatal aspergillosis in a patient with SARS who was treated with corticosteroids. N Engl J Med 2003;349: 507–8. 25 Ho JC, Ooi GC, Mok TY et al. High dose pulse versus non-pulse corticosteroid regimens in severe acute respiratory syndrome. Am J Respir Crit Care Med 2003;168: 1449–56. 26 Shimizu M, Shimizu Y. Effects of ambient temperatures on clinical and immune responses of pigs infected with transmissible gastroenteritis virus. Vet Microbiol 1979; 4: 109–16. 27 Tsunemitsu H, Smith DR, Saif LJ. Experimental inoculation of adult dairy cows with bovine coronavirus and detection of coronavirus in feces by RT-PCR. Archi Virol 1999;144: 167–75. 28 Buckingham SC, Jafri HS, Bush AJ et al. A randomized, double-blind, placebocontrolled trial of dexamethasone in severe respiratory syncytial virus (RSV) infection: effects on RSV quantity and clinical outcome. J Infect Dis 2002;185: 1222–8. 29 Domachowske JB, Bonville CA, Ali-Ahmad D et al. Glucocorticoid administration accelerates mortality of pneumovirus-infected mice. J Infect Dis 2001;184: 1518–23. 30 Banerjee S, Narayanan K, Mizutani T et al. Murine coronavirus replication-induced p38 mitogen-activated protein kinase activation promotes interleukin-6 production and virus replication in cultured cells. J Virol 2002;76: 5937–48. 31 Wuyts WA, Vanaudenaerde BM, Dupont LJ et al. N-acetylcysteine reduces chemokine release via inhibition of P38 MAPK in human airway smooth muscle cells. Eur Respir J 2003;22: 43–9. 32 Sidwell RW, Huffman JH, Call EW et al. Inhibition of murine hepatitis virus infections by the immunomodulator 2,3,5,6,7,8-hexahydro-2-phenyl-8,8dimethoxy-imidazo[1,2a]pyridine (PR879–317A). Antimicrob Agents Chemother 1987;31: 1130–4. 33 Cheung CY, Poon LL, Lau AS et al. Induction of proinflammatory cytokines in human
34
35.
35a
35b
36
36a
36b
36c
37
38
38a
38b
199
macrophages by influenza A (H5N1) viruses: a mechanism for the unusual severity of human disease? [see comment]. Lancet 2002;360: 1831–7. Hussell T, Pennycook A, Openshaw PJM. Inhibition of tumor necrosis factor reduces the severity of virus-specific lung immunopathology. Eur J Immunol 2001;31: 2566– 73. Holmes K, Lai MMC. Coronaviridae: the viruses and their replication. In: Fields BN, Knipe DM, Howley PM, eds. Virology. Philadelphia: Lippincott-Raven, 1996: 1075– 93. Prentice E, McAuliffe J, Lu X, Subbarao K, Denison MR. Identification and characterization of severe acute respiratory syndrome coronavirus replicase proteins. J Virol 2004; 78: 9977–86. Harcourt BH, Jukneliene D, Kanjanahaluethai A et al. Identification of severe acute respiratory syndrome corona-virus replicase products and characterization of papain-like protease activity. J Virol 2004;78: 13600–12. Li W, Moore MJ, Vasilieva N et al. Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus. Nature 2003;426: 450–4. Liu S, Xiao G, Chen Y et al. Interaction between heptad repeat 1 and 2 regions in spike protein of SARS-associated coronavirus: implications for virus fusogenic mechanism and identification of fusion inhibitors. The Lancet 2004;363: 938–47. Bosch BJ, Martina BE, Van D et al. Severe acute respiratory syndrome coronavirus (SARS-CoV) infection inhibition using spike protein heptad repeat-derived peptides. Proc Natl Acad Sci USA 2004;101: 8455–60. Hofmann H, Pohlmann S. Cellular entry of the SARS coronavirus. Trends Microbiol 2004; 12: 466–72. van der Meer Y, Snijder EJ, Dobbe JC et al. Localization of mouse hepatitis virus nonstructural proteins and RNA synthesis indicates a role for late endosomes in viral replication. J Virol 1999;73: 7641–57. Gosert R, Kanjanahaluethai A, Egger D. RNA replication of mouse hepatitis virus takes place at double-membrane vesicles. J Virol 2002;76: 3697–708. Goldsmith CS, Tatti KM, Ksiazek TG et al. Ultrastructural characterization of SARS coronavirus. Emerg Infect Dis 2004;10: 320–26. Prentice E, McAuliffe J, Lu X, Subbarao K, Denison MR. Identification and characterization of severe acute respiratory syndrome coronavirus replicase proteins. J Virol 2004; 78: 9977–86.
200
39
40
41
41a
41b
41c
41d
42
43
44
45
46
46a
47
Severe Acute Respiratory Syndrome
Prentice E, Jerome WG, Yoshimori T et al. Coronavirus replication complex formation utilizes components of cellular autophagy. J Biol Chem 2004;279: 10136–41. Bost AG, Prentice E, Denison MR. Mouse hepatitis virus replicase protein complexes are translocated to sites of M protein accumulation in the ERGIC at late times of infection. Virology 2001;285: 21–9. Rottier PJ, Rose JK. Coronavirus E1 glycoprotein expressed from cloned cDNA localizes in the Golgi region. J Virol 1987;61: 2042–5. Vennema H, Godeke GJ, Rossen JW et al. Nucleocapsid-independent assembly of coronavirus-like particles by coexpression of viral envelope protein genes. Embo J 1996;15: 2020–28. Narayanan K, Maeda A, Maeda J, Makino S. Characterization of the coronavirus M protein and nucleocapsid interaction in infected cells. J Virol 2000;74: 8127–34. Narayanan K, and Makino S. Cooperation of an RNA packaging signal and a viral envelope protein in coronavirus RNA packaging. J Virol 2001;75: 9059–67. He R, Leeson A, Ballantine M et al. Characterization of protein-protein interactions between the nucleocapsid protein and membrane protein of the SARS coronavirus. Virus Res 2004;105: 121–5. Salanueva IJ, Carrascosa JL, Risco C. Structural maturation of the transmissible gastroenteritis coronavirus. J Virol 1999;73: 7952–64. Kim JC, Spence RA, Currier PF et al. Coronavirus protein processing and RNA synthesis is inhibited by the cysteine proteinase inhibitor E64d. Virology 1995;208: 1–8. Yount B, Curtis KM, Fritz EA et al. Reverse genetics with a full-length infectious cDNA of severe acute respiratory syndrome coronavirus. Proc Nat Acad Sci USA 2003; 100: 12995–3000. Chu CM, Cheng VC, Hung IF et al. Role of lopinavir/ritonavir in the treatment of SARS: initial virological and clinical findings. Thorax 2004;59: 252–6. Xu X, Liu Y, Weiss S, Arnold E et al. Molecular model of SARS coronavirus polymerase: implications for biochemical functions and drug design. Nucleic Acids Res 2003;31: 7117–30. Wu CY, Jan JT, Ma SH et al. Small molecules targeting severe acute respiratory syndrome human coronavirus. Proc Natl Acad Sci USA 2004; 101: 10012–17. Sidwell RW. Ribavirin: in vitro antiviral activity. In: Smith RA, Kirkpatrick W, eds. Ribavirin: A Broad Spectrum Antiviral
Agent. New York: Academic Press, 1980: 23–42. 48 Sidwell RW, Huffman JH, Campbell N et al. Effect of ribavirin on viral hepatitis in laboratory animals. Ann NY Acad Sci 1977;284: 239–46. 49 Wyde PR, Chetty SN, Jewell AM et al. Comparison of the inhibition of human metapneumovirus and respiratory syncytial virus by ribavirin and immune serum globulin in vitro. Antiviral Res 2003;60: 51–9. 50 Weiss RC, Oostrom-Ram T. Inhibitory effects of ribavirin alone or combined with human alpha interferon on feline infectious peritonitis virus replication in vitro. Vet Microbiol 1989;20: 255–65. 51 Allen LB. Review of in vivo efficacy of ribavirin. In: Smith RA, Kirkpatrick W, eds. Ribavirin: A Broad Spectrum Antiviral Agent. New York: Academic Press, 1980: 43–58. 52 Centers for Disease Control. Severe acute respiratory syndrome (SARS) and coronavirus testing — United States, 2003. MMWR— Morbid Mortal Wkly Rep 2003;52: 297– 302. 53 Cinatl J, Morgenstern B, Bauer G et al. Glycyrrhizin, an active component of liquorice roots, and replication of SARS-associated coronavirus. Lancet 2003;361: 2045–6. 54 Tan ELC, Ooi EE, Lin CY et al. Inhibition of SARS coronavirus infection in vitro with clinically approved antiviral drugs. Emerg Infect Dis 2004;10: 581–6. 55 Ströher U, DiCaro A, Li Y et al. Severe acute respiratory syndrome related coronavirus is inhibited by interferon-aplha. J Infect Dis 2004;189: 1164–7. 56 Chen F, Chan KH, Jiang Y et al. In vitro susceptibility of ten clinical isolates of SARS coronavirus to antivirals and pure compounds extracted from traditional Chinese medicinal herbs. J Clin Virol 2004;31: 69–75. 56a Morgenstern B, Michaelis M, Baer PC, Doerr HW, Cinatl J. Ribavirin and interferon-b synergistically inhibit SARS-associated coronavirus replication in animal and human cell lines. Biochem Biophy Res Commun 2005;326: 905–8. 57 Kao RY, Tsui WH, Lee TS et al. Identification of novel small molecule inhibitors of severe acute respiratory syndrome associated coronavirus by chemical genetics. Chem Biol 2004;11: 1293–9. 58 Fujii T, Nakamura T, Iwamoto A. Current concepts in SARS treatment. J Infect Chemother 2004;10: 1–7. 59 So LK, Lau AC, Yam LY et al. Development of a standard treatment protocol for severe acute respiratory syndrome. Lancet 2003; 361: 1615–17.
Antiviral Agents for SARS
60
Peiris JSM, Lai ST, Poon LLM et al. Coronavirus as a possible cause of severe acute respiratory syndrome. Lancet 2003; 361: 1319–25. 61 Booth CM, Matukas LM, Tomlinson GA et al. Clinical features and short-term outcomes of 144 patients with SARS in the greater Toronto area. JAMA 2003;290: 334. 62 Knowles SR, Phillips EJ, Dresser L. Common adverse events associated with the use of ribavirin for severe acute respiratory syndrome in Canada. Clin Infect Dis 2003;37: 1139–42. 63 Ahmed R, Biron C. Immunity to viruses. In: Paul WE, ed. Fundamental Immunology. Philadelphia: Lippincott-Raven, 1999: 1295. 64 Sperber SJ, Hayden FG. Comparative susceptibility of respiratory viruses to recombinant interferons-alpha 2b and -beta. J Interferon Res 1989;9: 285–93. 65 Cinatl J, Morgenstern B, Bauer G et al. Treatment of SARS with human interferons. Lancet 2003;362: 293–4. 65a Der SD, Zhou A, Williams BR, Silverman RH. Identification of genes differentially regulated by interferon alpha, beta, or gamma using oligonucleotide arrays. Proc Natl Acad Sci USA 1998;95: 15623–8. 66 Pei J, Sekellick MJ, Marcus PI et al. Chicken interferon type I inhibits infectious bronchitis virus replication and associated respiratory illness. J Interferon Cytokine Res 2001;21: 1071–7. 67 Aurisicchio L, Delmastro P, Salucci V et al. Liver-specific alpha 2 interferon gene expression results in protection from induced hepatitis. J Virol 2000;74: 4816–23. 68 Aurisicchio L, Bujard H, Hillen W et al. Regulated and prolonged expression of mIFN(alpha) in immunocompetent mice mediated by a helper-dependent adenovirus vector. Gene Therapy 2001;8: 1817–25. 69 Smee DF, Alaghamandan HA, Bartlett ML et al. Intranasal treatment of picornavirus and coronavirus respiratory infections in rodents using 7-thia-8-oxoguanosine. Antiviral Chemistry & Chemotherapy 1990;1: 47–52. 70 Higgins PG, Phillpotts RJ, Scott GM et al. Intranasal interferon as protection against experimental respiratory coronavirus infection in volunteers. Antimicrob Agents Chemother 1983;24: 713–5. 71 Turner RB, Felton A, Kosak K et al. Prevention of experimental coronavirus colds with intranasal alpha-2b interferon. J Infect Dis 1986;154: 443–7. 72 Douglas RM, Moore BW, Miles HB et al. Prophylactic efficacy of intranasal alpha 2interferon against rhinovirus infections in the family setting. N Engl J Med 1986;314: 65–70.
73
201
Tilg H, Vogel W, Dinarello CA. Interferonalpha induces circulating tumor necrosis factor receptor p55 in humans. Blood 1995;85: 433–5. 74 Patel M, Ezzat W, Pauw KL et al. Bronchiolitis obliterans organizing pneumonia in a patient with chronic myelogenous leukemia developing after initiation of interferon and cytosine arabinoside. Eur J Haematol 2001; 67: 318–21. 75 Ogata K, Koga T, Yagawa K. Interferonrelated bronchiolitis obliterans organizing pneumonia. Chest 1994;106: 612–3. 76 Loutfy MR, Blatt LM, Siminovitich KA et al. Interferon alfacon-1 plus corticosteroids in severe acute respiratory syndrome: a preliminary study. JAMA 2003;290: 3222–8. 77 Baron S, Isaacs A. Absence of interferon in lungs from fatal cases of influenza. BMJ 1962; 1:18–20. 78 Holmes KV, Enjuanes L. Virology. The SARS coronavirus: a postgenomic era. Science 2003;300: 1377–8. 79 Yang H, Yang M, Ding Y et al. The crystal structures of severe acute respiratory syndrome virus main protease and its complex with an inhibitor. Proc Nat Acad Sci USA 2003;100: 13190–5. 80 Anand K, Ziebuhr J, Wadhwani P et al. Coronavirus main proteinase (3CLpro) structure: basis for design of anti-SARS drugs. Science 2003;300: 1763–7. 81 Patick AK, Binford SL, Brothers MA et al. In vitro antiviral activity of AG7088, a potent inhibitor of human rhinovirus 3C protease. Antimicrob Agents Chemother 1999;43: 2444–50. 82 Hayden FG, Turner RB, Gwaltney JM et al. Phase II, randomized, double-blind, placebo-controlled studies of ruprintrivir nasal spray 2-percent suspension for prevention and treatment of experimentally induced rhinovirus colds in healthy volunteers. Antimicrob Agents Chemother 2003;47: 3907–916. 82a Yamamoto N, Yang R, Yoshinaka Y et al. HIV protease inhibitor nelfinavir inhibits replication of SARS-associated coronavirus. Biochem Biophys Res Commun 2004;318: 719–25. 83 Chan KS, Lai ST, Chu CM et al. Treatment of severe acute respiratory syndrome with lopinavir/ritonavir: a multicentre retrospective matched cohort study. Hong Kong Med J 2003;9: 399–406. 84 Leibowitz JL, Reneker SJ. The effect of amantadine on mouse hepatitis virus replication. Adv Exp Med Biol 1993;342: 117–22. 84a Keyaerts E, Vijgen L, Maes P, Neyts J, Ranst MV. In vitro inhibition of severe acute respiratory syndrome coronavirus by chloroquine.
202
Severe Acute Respiratory Syndrome
Biochem Biophys Res Commun 2004;323: 264–8. 84b Wu CJ, Jan JT, Chen CM et al. Inhibition of severe acute respiratory syndrome coronavirus replication by niclosamide. Antimicrob Agents Chemother 2004;48: 2693–6. 84c Barnard DL, Hubbard VD, Burton J et al. Inhibition of severe acute respiratory syndrome-associated coronavirus (SARSCoV) by calpain inhibitors and beta-D-N4hydroxycytidine. Antivir Chem Chemother 2004;15: 15–22. 85 Hensley LE, Fritz EA, Jahrling PB et al. Interferon-beta 1a and SARS coronavirus replication. Emerg Infect Dis 2004;10: 317–19. 86 He ML, Zheng B, Peng Y et al. Inhibition of SARS-associated coronavirus infection and replication by RNA interference. JAMA 2003;290: 2665–6. 87 Sung JJY, Wu A, Joynt GM et al. Severe acute respiratory syndrome: report of treatment and outcome after a major outbreak. Thorax 2004;59: 414–20. 88 Choi KW, Chau TN, Tsang O et al. Outcomes and prognostic factors in 267 patients with severe acute respiratory syndrome in Hong Kong. Ann Intern Med 2003;139: 715–23. 88a ter Meulen J, Bakker ABH, van den Brink
89
90
91
92
93
EN et al. Human monoclonal antibody as prophylaxis for SARS coronavirus infection in ferrets. Lancet 2004;363: 2139–141. Greenough TC, Babcock GJ, Roberts A et al. Development and characterization of a severe acute respiratory syndrome-associated coronavirus-neutralizing human monoclonal antibody that provides effective immunoprophylaxis in mice. J Infect Dis 2005;191: 507–14. Traggiai E, Becker S, Subbarao K et al. An efficient method to make human monoclonal antibodies from memory B cells: potent neutralization of SARS coronavirus. Nat Med advanced online publication, 2005 Cheng Y, Wong R, Soo OY et al. Use of convalescent plasma therapy in SARS patients in Hong Kong. Eur J Clin Microbiol Infect Dis 2005;24: 44–6. Akerstrom S, Mousavi-Jazi M, Klingstrom J, Leijon M, Lundkvist A, Mirazimi A. Nitric oxide inhibits the replication cycle of severe acute respiratory syndrome coronavirus. J Virol 2005;79: 1966–9. Chen L, Liu P, Gao H et al. Inhalation of nitric oxide in the treatment of severe acute respiratory syndrome: a rescue trial in Beijing. Clin Infect Dis 2004;39: 1531–5.
Chapter 21
Vaccines Kanta Subbarao
Introduction Severe acute respiratory syndrome (SARS) is a severe respiratory illness caused by a newly identified virus, the SARS coronavirus (SARS CoV).1–4 As discussed in other chapters, the syndrome associated with this virus is characterized by fever, chills or rigors, headache and non-specific symptoms such as malaise and myalgias, followed by cough, and dyspnoea.4–6 Respiratory tract disease progresses to acute respiratory distress syndrome requiring intensive care and mechanical ventilation in more than 20% of patients.4–8 Prolonged hospitalizations associated with complications5,7 have been reported in several case series. The agespecific mortality rate in persons 60 years of age and older approached 50%.9 The severe morbidity and mortality associated with SARS make it imperative that effective means to prevent and treat the disease be developed and evaluated based on principles that apply to animal coronaviruses as well as other viral pathogens. Because SARS presents primarily as a respiratory illness, one can assume that some of the general principles established during the development of vaccines against respiratory syncytial virus, parainfluenza viruses and influenza viruses can be applied towards the development of SARS vaccines; these principles include the need for vaccines that generate a strong serum neutralizing antibody response and the desirability of vaccines that induce a specific mucosal
antibody response. More importantly however, SARS is different from illnesses caused by other respiratory viruses in many ways and the pathogenesis of SARS and the basis of protective immunity and virus clearance in SARS will help design appropriate vaccines. Basic pathogenesis questions relevant to vaccine development include the following: 1 How much of the disease associated with SARS is caused by the cytopathic effect of the virus and how much by immunopathology? In many instances, deaths following SARS infection occurred in association with the development of acute respiratory distress syndrome (ARDS) late in the course of hospitalization5,7 resulting in questions regarding the relative contribution of primary viral infection and immunopathology. This is an important issue because it can influence the choice of vaccine strategy between one that is directed towards the development of a neutralizing antibody response or towards the induction of specific cellular immune responses. 2 What is the mechanism of viral clearance from the respiratory tract and from the gastrointestinal (GI) tract? It has been suggested that SARS CoV falls into the category of a pneumo-enteric virus because it has been isolated from both sites and transmission may result from virus shed from either site.10–12 Serum neutralizing antibodies are better able to protect the lower respiratory tract than the upper respiratory tract or the GI tract.13 Mucosal antibodies may also play
203
204
Severe Acute Respiratory Syndrome
a major role in protection of the respiratory and enteric tracts. 3 Is infectious SARS CoV persistently shed from the gastrointestinal tract? If so, what is the potential for such spread from a live attenuated vaccine? Presence of SARS CoV genome has been detected by reverse transcriptase polymerase chain reaction (RTPCR) in the faeces of patients,14 including one report of detection as late as 73 days after the onset of illness.15 The presence of coronavirus in the small intestines has been demonstrated by electron microscopy15 and virus has been isolated from faeces in the first 3 weeks of illness but not subsequently.11 Contaminated faeces have been suspected to be the source of transmission in a major community outbreak of the disease viz. that at Amoy gardens in Hong Kong.16 However, there is no evidence of persistent shedding of infectious virus from the GI tract or evidence of transmission during late convalescence. Although the risk of faecal spread following administration of a live attenuated SARS CoV vaccine may be low because the virus is attenuated, the risk of transmission must be considered. 4 Is there evidence of central nervous system involvement in SARS? Although central nervous system involvement was not reported in case series during the outbreak, follow-up studies of survivors may help clarify whether there was unrecognized involvement of the central nervous system. The potential for spread of SARS CoV to the central nervous system must be established because of a reported association between human coronavirus infection and multiple sclerosis17 and because another prototype coronavirus, mouse hepatitis virus, is neurovirulent.18 This is an important issue in vaccine development because central nervous system involvement will have to be evaluated if live attenuated vaccines are developed. 5 Are convalescent patients with detectable neutralizing antibodies protected from subsequent infection? This is a critical issue because it will define the correlates of protective immunity. At a minimum, the
roles of neutralizing serum antibody, T cells and mucosal antibodies in protection must be established so that these specific arms of the immune system can be targeted. Unlike other respiratory viral pathogens such as influenza that have a very short incubation period, clinical data from SARS outbreaks suggest that respiratory illness begins several days after the onset of fever and viral shedding from the respiratory tract peaks 10 days from the onset of illness.14 This longer interval between exposure and peak viral replication may be sufficient to permit an expansion of memory T cells in SARS which is not possible in influenza, so that protection that relies on cellular immune responses may be effective in addition to vaccine strategies that elicit serum antibodies. 6 What is the significance of the limited genetic variability seen among SARS CoV isolates? A large number of SARS CoV isolates have been sequenced;19–24 and although deletions have been reported in open reading frame 8,20 the spike protein appears to be quite conserved suggesting that protective antibodies elicited by one virus will probably protect against others. Deletions in open reading frame (ORF) 8 such as the 29-nucleotide deletion reported in most human isolates compared with the viruses isolated from civet cats and raccoon dogs,20,25 may alter virulence but are unlikely to alter the antigenicity of the spike protein. Therefore immunity generated against a virus without the deletion would probably protect against subsequent challenge with a virus in which the deletion is not seen.20,25 7 Animal models will be needed to understand better the enhanced morbidity and mortality seen in the elderly. Although children were infected with SARS CoV, illness in children was generally mild26 and mortality associated with SARS was observed in the elderly.9 An improved understanding of this unusual epidemiology, with severe morbidity and mortality in the elderly and lack of severe illness and low attack rates in children, can have important implications for vaccine development.
Vaccines
Lessons from animal coronavirus vaccines Vaccines have been developed against a number of animal coronaviruses, with varying degrees of success. Live attenuated vaccines, inactivated vaccines, subunit and vectored vaccines have been evaluated in different instances (reviewed in Enjuanes et al.27), but there are few licensed veterinary coronavirus vaccines on the market. Vaccines are available for feline infectious peritonitis virus (FIPV), bovine coronaviruses, transmissible gastroenteritis virus (TGEV), and infectious bronchitis virus (IBV).28 Porcine respiratory coronavirus (PRCV) is a non-enteropathogenic virus derived from TGEV, with a 224 to 227 amino acid deletion in the spike protein that results in the deletion of some antigenic sites while others are conserved; this deletion may account for differences in tropism of the two viruses. PRCV seems to act as a natural vaccine against TGEV;29 the spread of PRCV in the swine population has been accompanied by a reduction in TGEV outbreaks because lactogenic antibodies induced by PRCV protect pigs against TGEV.30 If it is established that SARS-like coronaviruses with deletions in the genome are less virulent in humans, as has been suggested from molecular epidemiology studies,20 these features can be incorporated into the development of a vaccine. Although cost can be a significant issue for veterinary vaccines, it has not been the only deterrent to the development of animal coronavirus vaccines. Other significant obstacles have been the inability of the vaccine strain to protect against a range of field isolates, strain variation among avian coronaviruses31 and lack of reproducibility of efficacy data with certain veterinary vaccines.32,33 For example, a temperaturesensitive vaccine against FIPV was more efficacious in one clinical trial than in subsequent trials.32,33 Enhanced and accelerated disease similar to antibody-mediated immune enhancement seen in dengue haemorrhagic fever
205
has been reported when cats that are seropositive due to prior infection, passive transfer of antibody or vaccination are infected with FIPV.34–39 Although the phenomenon is not fully understood, enhanced uptake and spread of the virus are believed to result from binding of virusantibody immune complexes to Fc receptors on the surface of macrophages,38 and neutralizing antibodies directed against the spike protein are mainly responsible for the phenomenon.40,41 This antibodydependent enhancement appears to be limited to FIPV among coronaviruses, but concerns about a similar association have been raised with regard to SARS CoV. An important biological distinction between FIPV and SARS CoV is that FIPV infects macrophages while autopsy studies of fatal SARS cases.42,43 and pathology in animal models indicate that SARS CoV primarily infects epithelial cells44,45 (Fig. 21.1). Several systems have been developed independently to generate and rescue cDNAs encoding the genome of coronaviruses; these techniques include rescue of TGEV by stable cloning of the full-length genome into bacterial artificial chromosomes,46 IBV in vaccinia virus47 and rescue of mouse hepatitis virus and SARS CoV by in vitro ligation from several contiguous cDNA clones.48,49 These techniques will enable researchers to engineer specific attenuating mutations into the genome of the virus to develop live, attenuated vaccines.48,50 The ability to introduce specific mutations into a viral genome is a very powerful technique that has been applied to the development of other vaccines.51,52
Approach to the development of vaccines against SARS We have established that intranasally administered SARS CoV replicates to high titre in the respiratory tract of mice, hamsters and non-human primates.45,53–55 In each of these animal models, we have demonstrated that infection with SARS
206
(a)
(c)
Severe Acute Respiratory Syndrome
(b)
(d)
CoV elicits serum neutralizing antibodies and that primary infection prevents the replication of challenge virus administered a month later in all three models.45,54,55 There was no evidence of enhanced disease or enhanced replication following challenge in any of the animal models.45,54,55 In the mouse model, passive transfer of neutralizing antibody transfers protection from pulmonary virus replication upon subsequent challenge (Table 21.1).45 These data suggest that vaccines that elicit a neutralizing antibody response should be protective. A variety of different vaccine strategies are being explored for SARS, including inactivated vaccines,56 vectored vaccines expressing the SARS CoV spike protein,57–59 DNA vaccines60 subunit vaccines (purified SARS spike protein expressed in a recombinant baculovirus), vaccine-like particles,
Figure 21.1 Histopathology, IHC, and ISH of mouse lung tissues harvested on day 2 following infection. (a) Focal and mild peribronchiolar mononuclear inflammatory infiltrate. Haematoxylin and eosin stain; original magnification, ¥158. (b) SARS CoV antigens in multiple bronchiolar epithelial cells. Immunoalkaline phosphatase staining, naphthol fast red substrate with light haematoxylin counterstain; original magnification, ¥158. (c and d) SARS CoV nucleic acids in multiple bronchiolar epithelial cells. Immunoalkaline phosphatase staining, naphthol fast red substrate with light haematoxylin counterstain; original magnification, c, ¥100; d, ¥250. (Subbarao et al., 2004). Reproduced with permission from the Journal of Virology.
and live attenuated vaccines. Efficacy of protection is evaluated by measuring the level of replication of the challenge virus and/or the development of histopathological evidence of disease in animal models because none of the currently available animal models mimic the clinical illness that was seen in SARS. An inactivated vaccine (unpublished data), DNA vaccine60 and vaccinia virus vaccine expressing the SARS spike protein57 were immunogenic and provided complete protection against pulmonary virus replication in mice. The DNA vaccine encoding the spike protein of SARS CoV mediated protection through neutralizing antibodies (Fig. 21.2).60 A live attenuated parainfluenza virus vaccine expressing the SARS spike protein was immunogenic and efficacious in hamsters and African green monkeys,58,61 and evaluation of
Table 21.1 Passive transfer of immune serum protects naïve mice from replication of challenge SARS-CoV in the respiratory tract Experiment1
Passive transfer serum2
Neutralizing Ab titer in passive transfer serum3
Mean prechallenge neutralizing Ab titre in recipient mice
Virus replication in mice 2 days following challenge Lungs
Nasal turbinates
No. infected/no. tested
Mean (± SE) virus titre4
No. infected/no. Mean (± SE) tested virus titre4
1
Immune Non-immune
1 : 284 £1 : 4
1 : 28 £1 : 4
0/3 2/3
£1.5 ± 05 3.9 ± 1.21
2/3 2/3
3.2 ± 0.72 2.4 ± 0.32
2
Undiluted immune 1 : 10 dilution of immune Non-immune
1 : 1024 1 : 274
1 : 231 1 : 22
0/3 1/3
£1.8 ± 06 2.0 ± 0.17
1/3 2/3
2.0 ± 0.17 3.3 ± 0.73
£1 : 4
£1 : 4
3/3
7.3 ± 0.14
3/3
5.6 ± 0.55
1 The dose of challenge SARS CoV administered intranasally to mice was 103 TCID in experiment 1 and 104 TCID in experiment 2. 50 50 2 200 ml of serum pooled from immunized or uninfected mice were administered to recipient mice by intraperitoneal injection in experiment 1 and 500 ml of indicated
serum preparations were administered to mice in experiment 2. 3 Titre of antibody that neutralized infectivity of 100 TCID of SARS CoV. 50 4 Virus titres are expressed as log TCID /g of tissue. 10 50 5 Virus not detected; this value represents the lower limit of detection of infectious virus in a 10% w/v suspension. 6 Virus not detected; this value represents the lower limit of detection of infectious virus in a 5% w/v suspension. 7 p < 0.05.
Vaccines
Reproduced from Subbarao et al. 2004,45 with permission from the Journal of Virology.
207
Vaccine: (a)
Control
SDCD
TCID50 (log10/g tissue)
7 6 5 4 3 2 1
0 Depletion: CD4 CD8 CD90 Control
+ – – –
+ + + –
– + – – 8
(b) TCID50 (log10/g tissue)
7
– – – +
+ – – –
5 4 3 2 1 SDCD
p = 0.0002
p = 0.0286 8
14 12 10 8 6 4 2 0 Ig: Control
– – – +
6
TCID50 (log10/g tissue)
Luminescence (¥ 105)
16
+ + + –
p = 0.49
0 T cells: Control (c)
– + – –
SDCD
7 6 5 4 3 2 1 0 Ig: Control
SDCD
Figure 21.2 Immune mechanism of protection from a DNA vaccine encoding the SARS spike protein gene. T-cell depletion, adoptive transfer and antibody passive transfer. (a) Monoclonal mouse anti-CD4, CD8 and CD4/CD8/CD90 were used to deplete T and NKT cells in SDCD and control vaccinated mice (n = 4).19 Mice were then challenged with SARS CoV (Urbani) 48 hours later. Viral replication in the lungs was measured as described in Fig. 3. Data represent mean ± SE, and statistical analysis was performed as above. A statistically significant difference was observed for each group relative to its control (p = 0.002). (b) Lack of protection against SARS CoV replication in the lungs after adoptive T-cell transfer from vaccinated mice. Each naïve mouse (n = 4) received 3 x 107 purified T-cells from a donor mouse confirmed to respond to the vaccine (immune) or from a non-immune mouse (control). The recipient mice were challenged 24 hours after adoptive T cell transfer. Viral titre in lung was measured. There was no statistically significant difference between these two groups (p = 0.39). (c) Protection against SARS CoV replication in the lungs after passive transfer of immune IgG from vaccinated mice. Purified IgG from SDCD and control vaccinated donor mice (n = 4) was passively transferred into recipient naïve mice (n = 4). Serum from recipient mice was collected 1 day prior to challenge, and neutralization activity was confirmed using the S pseudotyped lentiviral vector (left panel). Recipient mice (n = 4 per group) were challenged 24 hours after IgG transfer, and TCID50 units per gram of lung tissue was measured in verocell monolayers. The non-parametric two-tailed t-test (Mann Whitney) revealed a statistically significant difference of 0.000072 between these two groups. Reproduced from (Yang et al., 2004) with permission from Nature.
Vaccines parainfluenza virus vectors expressing the S, M, N and E proteins alone and in combination established that the S protein alone conferred protective immunity in hamsters.61 These preclinical data demonstrate that vaccines that elicit neutralizing antibodies in animal models confer protection from replication of virus in the lower respiratory tract upon challenge and indicate that SARS CoV differs from diseases such as HIV and hepatitis C that are difficult to prevent with neutralizing antibody alone. Preclinical evaluation of candidate vaccines will include immunogenicity and efficacy studies in one or more animal models, including mice,45 hamsters,55 ferrets62 and non-human primates.53,54,59 SARS CoV replication has been reported in experimentally infected mice45 and cats,62 and illness and pneumonitis have been reported in experimentally infected ferrets,62 hamsters55 and monkeys.53,54 Unfortunately, the accelerated fulminant disease seen in FIPV is not sufficiently well understood to design preclinical studies to ascertain this risk with SARS CoV vaccines. However, a careful assessment of the quality and quantity of the immune response to vaccines must be undertaken because of these concerns. This includes assay of neutralizing antibody and ELISA titres and the ability of antibodies induced by the candidate vaccine to protect naïve animals from challenge with SARS CoV on passive transfer. Efficacy of protection conferred by candidate vaccines against challenge with wild-type SARS CoV can be evaluated in animal models with restriction of virus replication or absence of pathologic findings as end-points. For example, SARS CoV replicates in the lungs and nasal turbinates of mice but does not cause disease.45 In this model, efficacy of a vaccine can be measured by the ability of the vaccine to prevent replication of the challenge virus in the respiratory tract. The inability to recover virus from tissues can be confirmed by the absence of viral genetic material by RT-PCR. In animals such as hamsters,55 ferrets62 or cynomolgus
209
macaques53 that manifest pathologic findings with SARS CoV challenge, the endpoint in efficacy studies can be the absence of pneumonitis as well as restriction of virus replication. Each vaccine platform has advantages and disadvantages and associated concerns that will have to be addressed during preclinical and clinical development. For example, it will be important to characterize both the quantity and quality of the immune response elicited by inactivated and subunit vaccines to guard against enhanced disease such as that which was seen with inactivated respiratory syncytial virus and measles virus vaccines63 and to ensure that the inactivated vaccine can induce high quality neutralizing antibodies. Primeboost strategies will need to be explored for DNA vaccines because they have not been highly immunogenic in humans. The greatest concern with live attenuated vaccines may lie in the fact that coronaviruses can recombine and the risk of recombination may be magnified if a live SARS CoV vaccine is persistently shed from the GI tract. Vaccines that show promise in preclinical trials will probably proceed rapidly to clinical trials that will focus on assessment of the safety and immunogenicity of candidate vaccines. The nature and epidemiology of the SARS CoV pathogen may preclude efficacy trials in humans, so efficacy data in animal models and assessment of the immunogenicity in humans will be critical end-points in the evaluation of candidate SARS vaccines. Health-care workers may be selected as a study population for vaccine evaluations because they are at high risk for exposure to SARS CoV. In the event that a promising vaccine becomes available, recommendations for its use will probably be determined by the epidemiology of the disease. If the primary purpose of a vaccine is for use in outbreak control, short-term protection will be sufficient and DNA or inactivated vaccines may be appropriate. In such outbreak settings, vaccines that rapidly induce protective
210
Severe Acute Respiratory Syndrome
immunity are preferable. If SARS becomes endemic, long-term protection such as that conferred by a live vaccine will be desirable. The obvious target populations for vaccine use will include persons that are occupationally exposed to SARS such as biomedical researchers who work with the virus and health-care workers who care for patients with SARS. Depending on the nature and dosing schedule of the vaccine and the kinetics of the protective immune response that is conferred, target groups may also include close contacts of patients with SARS.
Summary Attempts to develop vaccines against SARS have been undertaken using a variety of strategies and several candidate vaccines have been evaluated in animal models. The virus replicates in epithelial cells of the respiratory tract of animals and induces a neutralizing antibody response. Efficacy of protection is evaluated by measuring the level of replication of the challenge virus and/or the development of histopathologic evidence of disease in animal models because none of the currently available animal models mimic the clinical illness that was seen in SARS. Vaccines that elicit neutralizing antibodies in animal models confer protection from replication of virus in the lower respiratory tract upon challenge. Important lessons from the field of animal coronavirus vaccine development must be considered during the development and evaluation of SARS vaccines. The extent of genetic and antigenic heterogeneity of the SARS CoV in its natural reservoir and the identification of that reservoir are significant unanswered questions that will influence the design and use of SARS vaccines.
References 1 Drosten C, Gunther S, Preiser W et al. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. N Engl J Med 2003;348: 1967–76.
2 Ksiazek TG, Erdman D, Goldsmith CS et al. A novel coronavirus associated with severe acute respiratory syndrome. N Engl J Med 2003;348: 1953–66. 3 Peiris JS, Lai ST, Poon LL et al. Coronavirus as a possible cause of severe acute respiratory syndrome. Lancet 2003;361: 1319–25. 4 Poutanen SM, Low DE, Henry B et al. National Microbiology Laboratory and Canadian Severe Acute Respiratory Syndrome Study. Identification of severe acute respiratory syndrome in Canada. N Engl J Med 2003;348: 1995–2005. 5 Lee N, Hui D, Wu A. A major outbreak of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348: 1986–94. 6 Tsang KW, Ho PL, Ooi GC et al. A cluster of cases of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348: 1977–85. 7 Fowler RA, Lapinsky SE, Hallett D et al. Critically ill patients with severe acute respiratory syndrome. JAMA 2003:290: 367–73. 8 Lew TW, Kwek TK, Tai D et al. Acute respiratory distress syndrome in critically ill patients with severe acute respiratory syndrome. JAMA 2003:290: 374–80. 9 Donnelly CA, Ghani AC, Leung GM et al. Epidemiological determinants of spread of causal agent of severe acute respiratory syndrome in Hong Kong. Lancet 2003;361: 1761–6. 10 Ng SKC. Possible role of an animal vector in the SARS outbreak at Amoy Gardens. Lancet 2003:362: 570–2. 11 Chan KH, Poon, LLM, Cheng V et al. Detection of SARS coronavirus in patients with suspected SARS. Emerg Infect Dis 2004;10: 294–9. 12 Seto WH, Tsang D, Yung RW et al. Effectiveness of precautions against droplets and contact in prevention of nosocomial transmission of severe acute respiratory syndrome (SARS). Lancet 2003;361: 1519– 20. 13 Murphy BR. Mucosal immunity to viruses. In: Ogra PL, Mestecky J, Lamm M et al., eds. Mucosal Immunology, 2nd edn. San Diego: Academic Press, 1999: 695–707. 14 Peiris JS, Chu CM, Cheng VC et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;361: 1767–72. 15 Leung WK, To K.-F, Chan PK et al. Enteric involvement of severe acute respiratory syndrome-associated coronavirus infection. Gastroenterology 2003;125: 1011–17. 16 Yu IT, Li Y, Wong TW et al. Evidence of airborne transmission of the severe acute respiratory syndrome virus. N Engl J Med 2004;350: 1731–9.
Vaccines
17 Arbour N, Day R, Newcombe J et al. Neuroinvasion by human respiratory coronaviruses. J Virol 2000;74: 8913–21. 18 Navas-Martin S, Weiss SR. SARS: Lessons learned from other coronaviruses. Viral Immunol 2003;16: 461–74. 19 Chim SSC, Tsui SKW, Chan KCA et al. Genomic characterisation of the severe acute respiratory syndrome coronavirus of Amoy Gardens outbreak in Hong Kong. Lancet 2003;362: 1807–8. 20 Consortium CSME. Molecular evolution of the SARS coronavirus during the course of the SARS epidemic in China. Science 2004;303: 1666–9. 21 Guan Y, Peiris JSM, Zheng BJ et al. Molecular epidemiology of the novel coronavirus that causes severe acute respiratory syndrome. Lancet 2004;363: 99–104. 22 Marra MA, Jones SJ, Astell CR et al. The genome sequence of the SARS-associated coronavirus. Science 2003;300: 1399– 404. 23 Rota PA, Oberste MS, Monroe SS et al. Characterization of a novel coronavirus associated with severe acute respiratory syndrome. Science 2003;300: 1394–9. 24 Ruan YJ, Wei CL, Ee AL et al. Comparative fulllength genome sequence analysis of 14 SARS coronavirus isolates and common mutations associated with putative origins of infection. Lancet 2003;361: 1779–85. 25 Guan Y, Zheng BJ, He YQ et al. Isolation and characterization of viruses related to the SARS coronavirus from animals in southern China. Science 2003;302: 276–8. 26 Hon KL, Leung CW, Cheng WT et al. Clinical presentations and outcome of severe acute respiratory syndrome in children. Lancet 2003;361: 1701–3. 27 Enjuanes L, Smerdou C, Castilla J et al. Development of protection against coronavirus induced diseases. In: Talbot PJ, Levy GA, eds. Corona- and Related Viruses. New York: Plenum Press, 1995: 197–211. 28 Berger A, Drosten C, Doerr HW et al. Severe acute respiratory syndrome (SARS)-paradigm of an emerging viral infection. J Clin Virol 2003;29: 13–22. 29 Wesley RD, Woods RD, Cheung AK et al. Genetic analysis of porcine respiratory coronavirus, an attenuated variant of transmissible gastroenteritis virus. J Virol 1991;65: 3369–73. 30 Diego MD, Rodriguez F, Alcaraz C et al. Characterization of the IgA and subclass IgG responses to neutralizing epitopes after infection of pregnant sows with the transmissible gastroenteritis virus or the antigenically related porcine respiratory coronavirus. J Gen Virol 1994;75: 2585–93.
211
31 Cavanagh D. Severe acute respiratory syndrome vaccine development: experiences of vaccination against avian infectious bronchitis coronavirus. Avian Pathol 2003;32: 567–82. 32 Fehr D, Holznagel E, Bolla S et al. Placebocontrolled evaluation of a modified live virus vaccine against feline infectious peritonitis: safety and efficacy under field conditions. Vaccine 1997;15: 1101–9. 33 Scott FW, Corapi WV, Olsen CW. Evaluation of the safety and efficacy of Primucell-FIP vaccine. Feline Health Topics 1992;7: 6–8. 34 Olsen CW. A review of feline infectious peritonitis virus: molecular biology, immunopathogenesis, clinical aspects and vaccination. Vet Microbiol 1993;36: 1–37. 35 Pederson NC, Boyle JF. Immunologic phenomena in the effusive form of feline infectious peritonitis. Am J Vet Res 1980;41: 868–76. 36 Vennema H, de Groot RJ, Harbour DA et al. Early death after feline infectious peritonitis virus challenge due to recombinant vaccinia virus immunization. J Virol 1990;64: 1407–9. 37 Weiss RC, Scott FW. Antibody-mediated enhancement of disease in feline infectious peritonitis: comparison with dengue hemorrhagic fever. Comp Immunol Microbiol Infect Dis 1981;4: 175–89. 38 Weiss RC, Scott FW. Pathogenesis of feline infectious peritonitis: nature and development of viremia. Am J Vet Res 1981;42: 382–90. 39 Weiss RC, Scott FW. Pathogenesis of feline infectious peritonitis: pathologic changes and immunofluorescence. Am J Vet Res 1981;42: 2036–48. 40 Corapi WV, Darteil RJ, Audonnet J-C et al. Localization of antigenic sites of the S glycoprotein of feline infectious peritonitis virus involved in neutralization and antibodydependent enhancement. J Virol 1995;69: 2858–62. 41 Olsen CW, Corapi WV, Ngichabe CK et al. Monoclonal antibodies to the spike protein of feline infectious peritonitis virus mediate antibody-dependent enhancement of infection of feline macrophages. J Virol 1992;66: 956–65. 42 Franks TJ, Chong PY, Chui P et al. Lung pathology of severe acute respiratory syndrome (SARS): a study of 8 autopsy cases from Singapore. Hum Pathol 2003;34: 743–8. 43 To KF, Tong JHM, Chan PKS et al. Tissue and cellular tropism of the coronavirus associated with severe acute respiratory syndrome: an insitu hybridization study of fatal cases. J Pathol 2004;202: 157–63. 44 Haagmans BL, Kuiken T, Martina BE et al. Pegylated interferon-alpha protects type 1 pneumocytes against SARS coronavirus infection in macaques. Nat Med 2004;10: 290–3.
212
Severe Acute Respiratory Syndrome
45 Subbarao K, McAuliffe J, Vogel L et al. Prior infection and passive transfer of neutralizing antibody prevent replication of SARS coronavirus in the respiratory tract of mice. J Virol 2004;78: 3572–7. 46 Almazan F, Gonzalez JM, Penzes Z et al. Engineering the largest RNA virus genome as an infectious bacterial artificial chromosome. Proc Natl Acad Sci USA 2000;97: 5516–21. 47 Casais R, Thiel V, Siddell SG et al. Reverse genetic system for the avian coronavirus infectious bronchitis virus. J Virol 2001;75: 12359–69. 48 Yount B, Curtis KM, Fritz EA et al. Reverse genetics with a full-length infectious cDNA of severe acute respiratory syndrome coronavirus. Proc Natl Acad Sci USA 2003;100: 12995–3000. 49 Yount B, Denison MR, Weiss SR et al. Systematic assembly of a full-length infectious cDNA of mouse hepatitis virus strain A59. J Virol 2002;76: 11065–78. 50 Haijema BJ, Volders H, Rottier PJM. Live, attenuated coronavirus vaccines through the directed deletion of group-specific genes provide protection against feline infectious peritonitis. J Virol 2004;78: 3863–71. 51 Collins PL, Murphy BR. Respiratory syncytial virus: reverse genetics and vaccine strategies. Virology 2002;296: 204–11. 52 Murphy BR, Collins PL. Live-attenuated virus vaccines for respiratory syncytial and parainfluenza viruses: applications of reverse genetics. J Clin Invest 2002;110: 21–7. 53 Kuiken T, Fouchier RAM, Schutten M et al. Newly discovered coronavirus as the primary cause of severe acute respiratory syndrome. Lancet 2003;362: 263–70. 54 McAuliffe J, Vogel L, Roberts A et al. Replication of SARS coronavirus administered into the respiratory tract of African green, rhesus
55
56
57
58
59
60
61
62
63
and cynomolgus monkeys. Virology 2004;330: 8–15. Roberts A, Vogel L, Guarner J et al. Severe acute respiratory syndrome coronavirus infection of golden Syrian hamsters. J Virol 2005;79:503–11. Stadler K, Roberts A, Becker S et al. An adjuvanted inactivated vaccine induces sterilizing immunity against SARS coronavirus in mice. Emerg Infect 2005 (submitted). Bisht H, Roberts A, Vogel L et al. SARS coronavirus spike protein expressed by a highly attenuated recombinant vaccinia virus induces neutralizing antibody and protectively immunizes mice. Proc Natl Acad Sci USA 2004;101: 6641–6. Bukreyev A, Lamirande EW, Buchholz UJ et al. Mucosal immunization of nonhuman primates with an attenuated parainfluenza virus expressing the SARS coronavirus spike protein for the prevention of SARS. Lancet 2004;363: 2122–7. Gao W, Tamin A, Soloff A et al. Effects of a SARS-associated coronavirus vaccine in monkeys. Lancet 2003;362: 1895–6. Yang Z.-Y, Kong W-P, Huang Y et al. A DNA vaccine induces SARS coronavirus neutralization and protective immunity in mice. Nature 2004;428: 561–4. Buchholz UJ, Bukreyev A, Yang L et al. Contributions of the structural proteins of severe acute respiratory syndrone coronavirus to protective immunity. Proc Natl Acad Sci USA 2004;101: 9804–9. Martina BEE, Haagmans BL, Kuiken T et al. SARS virus infection of cats and ferrets. Nature 2003;425: 915. Murphy BR, Chanock RM. Immunization against viral diseases. In: Knipe DM, Howley PM, Griffin DE et al., eds. Fields Virology, Vol. 1. Philadelphia: Lippincott Williams & Wilkins, 2001: 435–68.
Chapter 22
Counting the Economic Cost of SARS YC Richard Wong and Alan Siu
Introduction On 12 March 2003, the World Health Organization (WHO) issued a global alert of a new infectious disease that was named severe acute respiratory syndrome (SARS). Within weeks it had spread to 29 countries across the world, with a cumulative total of 8096 cases and 774 deaths (see Chapter 2). The most affected areas were in Mainland China, Hong Kong, Taiwan, Singapore and Canada. Within Mainland China the areas most affected were Guangdong, Beijing, Shanxi and Inner Mongolia. Here we assess the economic impact of SARS with a special focus on Hong Kong.
How SARS affected the economy In the short term, SARS mainly affected economic growth by reducing aggregate demand. The most directly affected sectors were the service industries that required face to face interaction between service providers and customers — such as retail sales, tourism, business travel, and transportation. The drop in demand rippled through the domestic economy and increased unemployment. The effects also spread to other parts of the world through the effects of international trade and investment flows. As the disease spread to the community, the population chose to stay home as far as possible to reduce the risk of contracting the disease. Consumption activities were cur-
tailed. The lack of information about the nature of the disease and the true extent of its spread worsened and exaggerated fears and panic among the population. Private consumption spending dropped and consumer confidence evaporated. Retail sales were the most seriously affected sector of these economies. There were no reports, however, that fear had reached levels that led people to refuse to go to work on any significant scale. The supply side of the economy was therefore not affected. Service exports, in particular tourismrelated exports, were severely hit. The negative effect on tourism was not confined to SARS-infected areas alone, but also spread throughout the whole east and south-east Asian region and beyond. In some economies, the fear of travelling also prompted a reduction in tourism-related service imports. Although this could have a positive effect on domestic GDP, it led to negative spillover effects on economies elsewhere in the world. The airlines were of course severely impacted by the drop of outbound and inbound travellers. Some investment plans could be delayed due to heightened uncertainties in the short run. The rescheduling of investment activities does not affect the long-term production capacity of an economy. On the other hand, domestic investment plans and foreign investment inflows into the region could be significantly reduced if SARS was perceived to have long-term adverse economic consequences. Market perception of
213
214
Severe Acute Respiratory Syndrome
how the future economic outlook would impact SARS-infected economies can be gauged from the behaviour of stock markets. A factor that might offset the negative impact on aggregate demand is the rebound in private spending once the SARS outbreak abated. Consumer spending can be substituted over time to some degree, especially consumption of durable and semidurable goods. Domestic consumption can recover quickly once public confidence is restored. Business travellers are likely to respond more quickly to an improvement in sentiments; however, recovery of tourism is likely to take longer. Increased government spending can mitigate the negative impact of SARS on aggregate demand. But fiscal spending is unlikely to be effective in reviving an economy in the face of significant reductions in private spending. Well-targeted public spending in preventing bankruptcy of otherwise sound businesses and providing relief to workers who suffer significant income losses is likely
to do more to cushion the most damaging effects of SARS.
Initial forecasts of economic losses due to SARS Beginning in 2002, signs of economic recovery were in evidence. The US dollar was weakening and exports from the region, particularly Mainland China, were picking up very strongly. Since growth in many parts of the region was primarily driven by exports, these were hopeful signs that a recovery in the domestic economy would also follow as consumption picked up. As the SARS outbreak exploded in a number of east and south-east Asian countries, the short-term economic growth outlook in the region dimmed. The conditions of a sustained economic recovery into 2003 began to look less favourable. In April 2003, the World Bank projected that the output growth of non-Japan East Asia to reach 5.0% rate of growth in 2003, almost
Table 22.1 East Asia and selected economies: GDP growth projections 2002
East Asia Developing East Asia South-east Asia Indonesia Malaysia Philippines Thailand Transition China Vietnam Small economies East Asia NIEs Hong Kong (SAR) Korea Singapore Taiwan (China) Japan 1
2003
2004
actual1
revised2
forecast1
actual2
forecast1
revised forecast2
5.8 6.6 4.4 3.7 4.2 4.6 5.2
6.0 6.9 4.6 4.3 4.1 4.4 4.5
5.0 6.0 3.9 3.3 4.2 4.0 4.5
5.9 7.8 5.3 4.5 5.3 4.7 6.8
5.7 6.3 4.7 4.0 5.5 4.5 5.0
5.9 7.0 5.5 5.4 6.0 4.5 5.8
8.0 6.0 1.2 4.5 2.3 6.3 2.3 3.5 0.3
8.3 7.0 2.6 4.7 1.9 7.0 2.2 3.6 -0.3
7.2 7.0 3.2 3.6 2.0 4.9 1.7 3.0 0.6
9.3 7.2 4.2 3.0 3.2 3.1 1.1 3.3 2.7
7.2 7.0 3.3 4.8 4.0 5.5 4.9 4.0 1.6
7.8 7.5 3.4 4.4 4.6 4.4 4.5 4.3 1.8
Figures from World Bank (April 2003). Figures from World Bank (November 2004). Source: World Bank (2003). East Asia is sum of Developing East Asia and Newly Industrialized Economies (NIEs). 2
Counting the Economic Cost of SARS
one percentage point lower than the original estimate of 5.8% rate of growth in 2002 (see Table 22.1). The growth rates for Mainland China, Hong Kong, Singapore and Taiwan were all forecasted to drop by a quarter to more than half a percentage point. The epidemic was also expected to affect the growth prospects of other economies in the region. As it turned out, the SARS impact on the GDP growth rates of the East Asian economies was rather limited. The actual GDP growth rate of non-Japan East Asia was estimated by the World Bank in November 2004 to be 5.9% in 2003, as compared with the revised estimate of 6.0% in 2002. The Asian Development Bank provided some interesting simulations of the impact of SARS on the economies (see Table 22.2). The study constructed two scenarios. One assumed that SARS would have a serious economic impact in the second quarter of 2003 only, and the other assumed that the impact would last for two quarters. Their results showed that GDP growth in east and south-east Asian economies was likely to be reduced by around 0.2–1.8 percentage points in 2003 if SARS persisted for one quarter in individual economies. If the impact of SARS extended into two quarters (i.e. the second and third quarters of 2003), GDP growth was likely to be reduced by 0.5–4.0 percentage points in individual economies. The two scenarios depicted very large losses in output and income. The estimated income loss ranged from US$12.3 to 28.4 billion for east and south-east Asia as a whole under the two scenarios. Analysts and private rating agencies were also busy revising downward their shortterm forecasts for growth prospect. Their forecasts of Hong Kong’s GDP growth rate was reduced by 0.4–2.9% (see Table 22.3). On average, Hong Kong’s output growth was cut by 1.2 percentage point by private sector analysts, translating into a drop of around US$1.9 billion in spending on final goods and services in the domestic economy. Their estimates were low compared to those made by the Asian Development Bank, which put the losses in the range
215
of US$2.9–6.4 billion. Hong Kong’s Financial Secretary announced on 9 April 2003 that the official forecast of a 3% real GDP growth rate made a month earlier could not be met; he did not produce an alternative forecast.
Effects on retail sales The impact of SARS on retail sales in some of the most affected economies in the period January–August 2003 are presented in Table 22.4. The impact of SARS is clearly detectable in Hong Kong, Taiwan, Singapore and Mainland China. The economic impact is not so obvious in Vietnam or Canada. The Canadian figures were probably less dramatic because the SARS outbreak was largely confined to Toronto. In Mainland China, retail sales dropped from RMB349 billion in March 2003 to RMB341 billion (-2.3%) and RMB346 billion (-0.86%) in April and May 2003, respectively. In Hong Kong, retail sales dropped from HK$14 171 million in March 2003 to HK$12 723 million (-10.2%) and HK$13 780 million (-2.8%) in April and May 2003, respectively. In Taiwan, retail sales dropped from NT$218 421 million in March 2003 to NT$215 165 million (-1.5%) and NT$210 705 (-3.5%) in April and May 2003, respectively. In Singapore, retail sales dropped from SG$1863 million in March 2003 to SG$1777 million (-4.6%) in April and rebounded to SG$1987 million (6.7%) in May 2003. Drops in private consumption spending were more severe in economies where more people were infected by the disease and took a longer time for it to come under control. The economic impact on large continental economies was less severe than that on smaller city economies where the loss of tourist spending was more important. Containing the disease at an early stage turned out to be critical for averting a crisis in public confidence that could cause huge economic losses due to declines in consumer spending. Consumer spending quickly rebounded
216
Economies
Estimated reduction in annual GDP growth
Estimated reduction in annual GDP levels
Estimation of 2003 GDP growth (%)
Scenario 1: SARS lasts 1 quarter (percentage point)
Scenario 2: SARS lasts 2 quarters (percentage point)
Scenario 1: SARS lasts 1 quarter (US$ billion)
Scenario 2: SARS lasts 2 quarters (US$ billion)
ADO 2003*
Scenario 1: SARS lasts 1 quarter
Scenario 2: SARS lasts 2 quarters
East Asia PRC Hong Kong, China Republic of Korea Taipei, China
0.4 0.2 1.8 0.2 0.9
1.0 0.5 4.0 0.5 1.9
9.1 2.3 3.0 1.3 2.5
20.7 5.8 6.6 3.0 5.3
5.6 7.3 2.0 4.0 3.7
5.3 7.3 0.8 3.8 2.8
4.7 7.0 -1.4 3.5 1.8
South-east Asia Indonesia Malaysia Philippines Singapore Thailand
0.5 0.5 0.6 0.3 1.1 0.7
1.4 1.4 1.5 0.8 2.3 1.6
3.2 0.7 0.5 0.2 1.0 0.8
7.7 2.0 1.3 0.6 2.0 1.8
4.0 3.4 4.3 4.0 2.3 5.0
3.4 3.2 3.8 3.7 1.9 4.3
2.5 2.3 2.9 3.2 0.7 3.4
* Asian Development Outlook 2003. Source: Asian Development Bank (2003).
Severe Acute Respiratory Syndrome
Table 22.2 Estimation of the impact of SARS on selected Asian economies, 2003
Counting the Economic Cost of SARS
217
Table 22.3 Private sector forecasts of Hong Kong’s GDP, 2003 Output growth Previous (percent)
Latest (percent)
Difference (percentage point)
Consensus forecast (April 2003) Economist Intelligence Unit (EIU) Bloomberg News Survey1
2.8 2.5 2.9
2.4 0.3 2.1
-0.4 -2.2 -0.8
Major forecasts by private companies JP Morgan Chase Hongkong and Shanghai Banking Corporation BNP Paribas Peregrine Citibank DBS Bank Internationale Nederlanden Group (ING) Goldman Sachs Lehman Brothers Morgan Stanley Merrill Lynch ABN Amro Bank
3.2 1.6 1.5 2.8 1.8 3.0 3.0 2.5 2.7 4.6 4.0
0.3 0.5 0.9 1.0 1.2 1.5 1.7 2.0 2.1 3.1 3.5
-2.9 -1.1 -0.6 -1.8 -0.6 -1.5 -1.3 -0.5 -0.6 -1.5 -0.5
Average
2.8
1.6
-1.2
1
Median of forecasts by 10 economists in a Bloomberg News survey conducted in April 2003.
once the disease came under control and public confidence was restored. Table 22.4 shows that by July and August, retail spending had almost returned to their pre-crisis levels in most of the severely affected economies. Part of the rapid rebound could be attributed to inter-temporal substitution effects as consumers made up for delayed expenditures. Many retail sales outlets were also cutting prices in a bid to clear unsold inventories quickly.
Effects on tourism and travel As visitor arrivals fell, hotel occupancy rates dropped and many airlines had to cancel flights. Gross tourism receipts account for a non-trivial percentage of GDP in east and south-east Asian economies. Gross receipts overstate the contribution of tourism expenditures to GDP, because imported goods account for some tourist purchases. Table 22.5 presents figures of the impact of SARS on visitor arrivals in the month around the
time when SARS began to strike. Almost all of the Asian economies were severely hit; the worst were Hong Kong, Singapore and Taiwan. Table 22.6 shows a month-by-month account of visitor arrivals since January 2003 in some of the most directly hit economies where SARS appeared, namely, Hong Kong, Mainland China, Taiwan and Singapore. Although the drop in visitor arrivals was severe, the figures began to rebound shortly after the SARS outbreak came under control. Figure 22.1 shows that in Hong Kong, the number of aircraft movement began to drop after the end of March and reached a nadir in early May. By the end of July, aircraft movement had almost recovered to its original level. The drop and subsequent rebound of hotel occupancy rates shown in Table 22.7 reflects the effects of changes in the number of visitor arrivals. For the month of May when fear of the SARS outbreak was at its peak, hotel occupancy rates dropped to
218
Month
January February March April May June July August
Hong Kong
China
Taiwan
Singapore
Vietnam*
HKD mn
y-o-y % change
RMB bn
y-o-y % change
NTD mn
y-o-y % change
SGD mn
y-o-y % change
17 484 12 680 14 171 12 723 13 780 13 600 14 633 14 720
9.9% -12.6% -6.1% -15.2% -11.1% -6.5% -2.5% 1.2%
391 371 349 341 346 358 356 361
10.0% 8.5% 9.3% 7.7% 4.3% 8.3% 9.8% 9.9%
233 240 200 553 218 421 215 165 210 705 220 761 235 577 227 057
9.0% -0.1% 2.9% 2.1% -3.0% 4.1% 6.9% 4.6%
2419 1578 1863 1777 1987 1900 2031 1867
-1.4% -18.2% -8.2% -2.3% 13.3% 5.0% 9.2% 2.4%
* Note: Year-to-date figures. Source: CEIC data, Statistics Canada.
VND bn
50 311 75 588 99 336 123 640 148 647 174 467 200 975
Canada y-o-y % change
10.5% 11.8% 10.3% 9.6% 9.7% 10.9%
CAD mn
y-o-y % change
22 541 21 242 24 591 25 946 29 011 27 532 27 984 27 485
4.7% 5.9% 3.7% 1.7% 4.3% 2.1% 5.2% 1.5%
Severe Acute Respiratory Syndrome
Table 22.4 Retail sales in major SARS-affected economies, 2003
Counting the Economic Cost of SARS
219
Table 22.5 The impact of SARS on tourism in Asia
China Hong Kong Indonesia Japan South Korea Malaysia The Philippines Singapore Taiwan Thailand
Gross tourism receipts as % of GDP (2002)
y-o-y % change in visitor arrivals (latest monthly figure*)
1.7 6.2 3.8 0.1 1.1 7.5 2.2 5.0 1.5 6.1
-30 -65 -38 -25 -29 -36 -10 -67 -82 -45
* Note: March figures for Malaysia and the Philippines, May for Taiwan, and April for the rest. Source: CEIC data.
700
Number of movements
600 500 400 300 200 100 0 3/1
4/1
5/1
6/1
7/1
8/1
Date Figure 22.1 Aircraft movement at Hong Kong International Airport (1 March–31 August 2003).
18% in Hong Kong, 34% in Singapore, and 22% in Taiwan. While there is no doubt that the downturn of the tourist industry in these economies was severe, the true impact was less severe than initial fears. In a Special Report released by the World Travel & Tourism Council in 2003, it was estimated that economic losses of industry GDP in Vietnam,
Singapore, Hong Kong and China would be 15%, 43%, 41% and 25%, respectively. These estimates were based on assumptions that the SARS crisis would last for 6 months in China, 4 months in Hong Kong and Singapore, and 3 months in Vietnam, with residual impacts until the end of 2004. Fortunately these fears did not materialize and the final impact was less severe.
220
Month
January February March April May June July August
Hong Kong
China
Taiwan
Singapore
Person
y-o-y % change
Person
y-o-y % change
Person
y-o-y % change
Person
y-o-y % change
1 545 978 1 408 139 1 347 386 493 666 427 254 725 236 1 291 828 1 644 878
31.0% 26.2% 3.9% -64.8% -67.9% -38.2% -5.6% 9.6%
8 484 300 7 376 002 7 851 398 5 649 200 5 435 957 6 526 017 7 769 433 8 843 955
14.5% 3.7% -6.5% -30.1% -31.0% -18.0% -8.5% -0.7%
237 979 259 861 258 023 110 632 40 256 57 131 154 174 200 614
19.8% 21.7% -0.2% -50.7% -81.9% -74.0% -27.2% -11.1%
641 733 613 395 559 945 203 562 177 806 316 514 540 373 601 819
7.5% 2.4% -14.5% -67.3% -70.7% -47.4% -20.5% -10.4%
Source: CEIC data.
Severe Acute Respiratory Syndrome
Table 22.6 Visitor arrivals, 2003
Counting the Economic Cost of SARS
221
Table 22.7 Hotel room occupancy rates, 2003 Month
January February March April May June July August
Hong Kong
Taiwan
Singapore
Percent
y-o-y % change
Percent
y-o-y % change
Percent
y-o-y % change
82 81 79 22 18 34 71 88
1.2% 8.0% -8.1% -74.7% -78.3% -57.0% -13.4% 2.3%
56 68 61 37 22 34 58 68
7.2% 16.2% -4.9% -38.5% -62.7% -44.6% -5.2% 8.8%
71 74 71 35 34 56 71 74
0.0% 0.9% -9.8% -54.7% -53.1% -24.4% -5.9% 0.8%
Source: CEIC data.
Effects on GDP and unemployment The combined effect of drops in domestic consumption and tourist spending had a perceptible impact on GDP growth rates in the first and second quarters of 2003. Table 22.8 shows that year-on-year GDP growth rates in 2003Q1 and 2003Q2 were respectively –0.1% and –6.3% in Hong Kong, 0.9% and –2.0% in Taiwan, and 1.2% and –5.6% in Singapore. The effects were less discernible in Mainland China because of its very high baseline growth rate. The economies in Hong Kong, Taiwan and Singapore were weak to begin with and were only starting to recover in 2002 before SARS struck. The labour markets were relatively weak too and were experiencing historically high unemployment rates. Unemployment rates rose throughout the SARS outbreak period (see Table 22.9). Many severely affected sectors either laid off workers or undertook work sharing among workers in the face of reduced demand. The unemployment rates in Hong Kong, Taiwan and Singapore increased marginally in the first and second quarters of 2003. The unemployment rate in Hong Kong rose from 7.2% in 2002Q4 to 7.5% in 2003Q1 and 8.6% in 2003Q2. It rose from 5.1% in 2002Q4 to 5.2% in 2003Q1 and back to 5.1% in 2003Q2 in Taiwan and from 4.3%
in 2002Q4 to 4.5% in 2003Q1 and 4.6% in 2003Q2 in Singapore.
Effects on goods exports and imports Fear and panic among the population did not seem to have disrupted work and the production of goods was largely unaffected. Most of the SARS-affected economies were heavily dependent on export-oriented manufacturing production. For this reason, goods export was an important barometer of the health of these economies. Hong Kong was an exception in that its manufacturing production activities had been transferred offshore, but its economy remains heavily reliant on the vitality of the offshore manufacturing base that it was servicing. In the case of Hong Kong, the most relevant barometer is not only domestic exports but also the re-export of goods from its offshore manufacturing base. Figures on the export of goods continued to grow in Hong Kong, Mainland China, Taiwan and Singapore throughout the SARS crisis. Table 22.10 shows year-on-year growth rates for export figures month-bymonth. There is some volatility in monthby-month figures but on the whole there is no obvious sign of SARS having an impact on Hong Kong or Mainland China. In Singapore and Taiwan there was some evidence of a slow-down in the growth rates of
222
Quarter
2001Q1 2001Q2 2001Q3 2001Q4 2002Q1 2002Q2 2002Q3 2002Q4 2003Q1 2003Q2
Hong Kong
China
Taiwan
Singapore
HKD mn
y-o-y % change
RMB bn
y-o-y % change
NTD mn
y-o-y % change
SGD mn
y-o-y % change
307 424 310 521 323 873 328 077 295 775 303 049 321 081 327 475 295 347 283 871
-0.3% -0.7% -2.0% -2.5% -3.8% -2.4% -0.9% -0.2% -0.1% -6.3%
1990 2305 2428 3009 2147 2511 2687 3173 2396 2720
9.5% 8.1% 7.3% 10.0% 7.9% 8.9% 10.6% 5.5% 11.6% 8.3%
2 434 264 2 253 711 2 371 310 2 447 339 2 457 062 2 321 719 2 459 477 2 510 553 2 479 703 2 275 903
1.4% -2.9% -3.9% -1.1% 0.9% 3.0% 3.7% 2.6% 0.9% -2.0%
39 395 38 453 38 190 38 039 38 831 40 366 39 034 39 833 39 304 38 099
7.4% -0.1% -7.5% -11.9% -1.4% 5.0% 2.2% 4.7% 1.2% -5.6%
Source: CEIC data.
Severe Acute Respiratory Syndrome
Table 22.8 Gross domestic product (current prices)
Counting the Economic Cost of SARS
223
Table 22.9 Unemployment rates (seasonally adjusted) Quarter
2001Q1 2001Q2 2001Q3 2001Q4 2002Q1 2002Q2 2002Q3 2002Q4 2003Q1 2003Q2
Hong Kong
Taiwan
Singapore
Percent
y-o-y difference (%)
Percent
y-o-y difference (%)
Percent
y-o-y difference (%)
4.5 4.5 5.2 6.2 7.0 7.7 7.4 7.2 7.5 8.6
-1.1 -0.6 0.4 1.8 2.5 3.2 2.2 1.0 0.5 0.9
3.8 4.4 4.9 5.3 5.3 5.2 5.1 5.1 5.2 5.1
0.9 1.5 1.9 2.0 1.5 0.8 0.2 -0.1 -0.1 0.0
2.4 2.8 3.7 4.4 4.5 4.3 4.3 4.3 4.5 4.6
-1.0 -0.7 1.1 1.7 2.1 1.5 0.6 -0.1 0.0 0.3
Source: CEIC data.
goods exports during the period of SARS outbreak, but this could also be attributed to seasonal factors. The slow-down of goods exports in Vietnam and Canada is largely due to economic trend effects. Imports of goods were somewhat modestly impacted reflecting perhaps the reduction in consumer spending rather than production activities (see Table 22.11). The effects are most noticeable in Taiwan and Singapore during the SARS crisis, but less so for Hong Kong, Mainland China and Vietnam. The slow-down of exports in Canada reflects economic trend effects that are independent of the SARS crisis.
Case of Hong Kong’s cross-border manufacturing operations Hong Kong’s goods exports are dominated by re-exports, and are roughly ten times its domestic exports. Most of the re-exports are goods manufactured across the border in the Pearl River Delta in Guangdong province in the Mainland. Hong Kong’s imports are also re-exported to the Pearl River Delta to be used as intermediate inputs. The value of goods so traded was enormous. In 2002, total imports and exports of goods and services in Hong Kong was 293% of
GDP. Hong Kong is not only one of the world’s most open economies, it is also the business centre for managing a vast regional cross-border manufacturing base. A recent study by the Hong Kong Centre for Economic Research (2003) estimated that in the year 2001, 63 000 Hong Kongbased manufacturers and traders were engaged in manufacturing activities in the Mainland. Among them 59 000 operated or owned factory facilities there were employing more than 11 million workers. The overwhelming proportion of these activities was located in Guangdong with some 10 million workers. These 63 000 manufacturers and traders employed 477 000 workers in Hong Kong. Hong Kong’s total labour force was only 3.5 million in 2001. The study also estimated that as many as 1 million additional jobs in Hong Kong were in producer services that were indirectly engaged in supporting the manufacturing base in the Mainland. In other words, some 43% of the Hong Kong labour force is either directly or indirectly engaged in managing and supporting a cross-border manufacturing base. The role of Hong Kong as the management, financing, transportation and trading hub of this vast export-oriented
224
Month
January February March April May June July August September
Hong Kong
China
Taiwan
Singapore
Vietnam
Canada*
HKD mn
y-o-y % change
USD mn
y-o-y % change
NTD mn
y-o-y % change
SGD mn
y-o-y % change
USD mn
y-o-y % change
CAD mn
y-o-y % change
135 028 105 148 143 087 134 146 143 022 142 343 157 574 151 171 157 518
26.7% 10.4% 15.4% 9.0% 13.6% 14.0% 7.6% 7.0% 6.4%
29 776 24 456 32 091 35 617 33 842 34 476 38 110 37 415 41 941
37.2% 27.8% 34.7% 33.3% 37.3% 32.5% 30.5% 27.1% 31.4%
348 652 340 430 436 645 397 902 391 291 401 354 399 702 422 174 427 747
3.1% 20.9% 9.0% 4.8% 2.5% 4.7% 7.3% 15.2% 11.3%
20 905 17 540 20 874 20 166 19 528 19 886 20 902 20 370 22 599
18.1% 19.9% 13.8% 3.8% 5.1% 7.5% 5.7% 5.5% 20.5%
1615 1450 1598 1614 1798 1695 1818 1720 1622
61.5% 53.9% 28.6% 24.2% 33.7% 13.5% 22.2% 14.7% 1.7%
35,131 34 937 35 334 33 562 32 586 32 540 33 191 32 177 33 761
7.1% 3.3% 5.9% -2.9% -3.6% -3.6% -5.0% -7.8% -3.9%
* Note: Seasonally adjusted figures. Source: CEIC data, Statistics Canada.
Severe Acute Respiratory Syndrome
Table 22.10 Exports of goods, 2003
Table 22.11 Imports of goods, 2003 Month
China
Taiwan
Singapore
Vietnam
Canada*
HKD mn
y-o-y % change
USD mn
y-o-y % change
NTD mn
y-o-y % change
SGD mn
y-o-y % change
USD mn
y-o-y % change
CAD mn
y-o-y % change
133 225 115 984 154 212 141 320 145 796 146 369 159 883 152 918 162 164
21.4% 18.3% 14.6% 8.4% 9.4% 11.5% 5.6% 6.0% 7.3%
31 016 23 780 32 549 34 601 31 604 32 336 36 514 34 621 41 652
63.5% 49.4% 44.8% 34.4% 40.9% 40.0% 35.3% 27.3% 39.8%
329 948 292 162 378 978 363 132 331 429 357 396 357 324 358 130 375 354
28.1% 27.6% 6.3% 6.3% 1.3% 8.5% 0.2% 16.5% 3.9%
17 249 16 053 19 265 18 229 16 858 18 556 18 448 18 321 19 102
9.5% 9.3% 9.8% 0.9% -2.9% 1.7% -0.7% 2.2% 13.0%
1623 1570 2253 2358 2299 2037 2133 1950 2173
29.8% 41.1% 60.5% 54.1% 48.6% 14.9% 24.1% 18.2% 27.0%
29 951 30 106 30 038 28 960 28 640 27 972 28 398 26 838 27 722
8.1% 3.6% 4.8% -0.2% -2.0% -6.7% -5.0% -12.6% -8.5%
* Note: Seasonally adjusted figures. Source: CEIC data, Statistics Canada.
Counting the Economic Cost of SARS
January February March April May June July August September
Hong Kong
225
226
Severe Acute Respiratory Syndrome
cross-border manufacturing base means that its economy was potentially susceptible to serious disruption if the SARS outbreak in Guangdong paralysed parts of these global supply chains. The potential impact on GDP and employment could be quite severe for Hong Kong. The fear of a fall in goods exports and reexports is not borne out by the cross-border truck traffic figures (see Fig. 22.2), which showed no obvious decrease throughout the period of the SARS outbreak. Evidently the production and movement of goods went unperturbed. The external trade figures in the first and second quarters of 2003 continued to grow robustly for goods. For the first and second quarters of 2003, the value of total exports of goods rose by 17.6% and 12.2%. Over the same period in 2002 the value of re-exports surged by 20.3% and 14.4%, while the value of domestic exports decreased by 10.4% and 11.3%. During the same two periods, the value of imports of goods increased by 18.0% and 10.0%, resulting in a visible
trade deficit of HK$16.0 billion and HK$9.8 billion. Thus, the external trade sector has performed quite well in the first and second quarters of 2003. There were no reports of major disruption in work in factories in the Pearl River Delta. The production lines were normal despite the SARS outbreak. Since most factory hands employed in the manufacturing operations in the Pearl River Delta were staffed by migrant workers who lived in employerprovided hostels, this probably helped to insulate the workers from the disease that was spreading in the community. Even the curtailment of business travel to the region by overseas companies appeared to have been mitigated by the return of businessmen after the SARS outbreak was over and local businessmen who travelled to meet with their overseas clients.
Reactions in the stock markets The stock market reaction to the SARS outbreak was relatively moderate and short-
Number of trucks
15,000
10,000
5,000
Inbound
Outbound
0 3/1
4/1
5/1
6/1
7/1
8/1
9/1
Date Figure 22.2 Cross-border truck traffic — Lok Ma Chau (1 March–8 September 2003). Source: Unpublished daily data provided by Customs and Excise Department, Hong Kong SAR Government.
Counting the Economic Cost of SARS
lived. The Hang Seng Index only dropped by 2.71% between March 12 and March 31 when WHO raised the global alert. Other stock markets, except for Taiwan, had risen over the same period (Table 22.12). The Hong Kong market surged by 3.7% on April 29 after WHO said the worst of the SARS outbreak appeared to be over in Singapore, Hong Kong, Canada and Vietnam. Share price indices dropped in Taiwan and Shanghai during the period 15–30 April in response to reports of fresh outbreaks of SARS cases that were related to visitors involving the two cities; but the effects were again short-lived. Airline shares suffered steep losses (see Table 22.13). The share-price of Cathay Pacific dropped by 13.64% between 12 and 31 March and by a further 8.95% between March 31 and April 15. Air Canada share prices dropped by 20.75% and 47.14% in the corresponding period and the airline had to file bankruptcy proceedings. China Airlines and Singapore Airlines both suffered shares price drops of 11.73% and 8.38% in the period March 12–31. China Eastern Airlines with a hub in Shanghai dropped by 12.40% in the period April 15–30 along with the rest of the Shanghai stock market. Both the short duration and small amount of price drops in the overall stock markets suggest that the market did not expect the SARS outbreak to have long-lasting and widespread economic effects. The most severely affected shares were those directly related to the tourist, travel and retail sectors.
Fiscal stimulus In response to the negative demand shock created by SARS, governments in the region announced fiscal stimulus packages that were targeted at providing relief to the hardest hit sectors. Table 22.14 provides a summary of the fiscal policy packages announced by respective governments. The relative size of the fiscal packages in relation
227
to the size of the economy varied from a low of 0.1% to a high of 2.0%. Such spending contributed to a worsening of the fiscal balance and the impact on the fiscal deficit as a percentage of GDP varied from a low of 0.2% to a high of 1.8%.
Concluding remarks Initial alarmist reports and estimates on the impact of SARS were fortunately not borne out. The economic impact of SARS turned out to be relatively short term in its effects on consumption, tourism and travelrelated services. Fear and panic subsided quickly once the outbreak abated. The stock market’s overall response mirrored these developments quite well, with most of the short-term negative effects concentrating on sectors that were severely hit by the negative demand shock. There was no major disruption to external trade in goods nor was there a supply side shock in the affected economies. Although it may not be definitive, at the time of writing, there was no anecdotal evidence that investment into the region was slowing or reversing. For Hong Kong, the offshore manufacturing base in the Pearl River Delta was unaffected and goods continued to flow through Hong Kong normally. The possibility that the virus will return cannot be ruled out. But the economies will be much better prepared, especially with research confirming that the disease can be effectively contained through isolation and quarantine measures. The spread of the disease in the hospitals is likely to be better handled with enhanced infection isolation facilities, infection control procedures, and heightened vigilance. The fact that the disease had raged on for several months in Mainland China without getting world attention is revealing of the damaging effects of information control in relatively closed societies. That the disease unfortunately spread to Hong Kong with its first outbreak in a hotel frequented by
228
Index
12–31 March
31 March– 15 April
30 May– 30 June
30 June– 30 July
Hong Kong: Hang Seng Index Shanghai Stock Exchange, A–share Taiwan Stock Exchange Weighted Singapore Straits Times S&P TSX Composite
-2.71%
-0.03%
0.99%
8.83%
0.95%
5.82%
-1.78%
6.90%
2.45%
8.03%
-6.71%
3.65%
-5.72%
-0.63%
3.25%
-0.16%
3.85%
-7.57%
9.83%
6.94%
9.16%
2.53%
2.16%
-0.68%
6.47%
6.57%
1.84%
2.28%
1.51%
4.16%
1.80%
Source: Datastream.
15–30 April
30 April– 30 May
12 March– 30 April
12 March– 30 May
12 March– 30 June
12 March– 31 July
12 March– 28 November
7.91%
14.20%
39.57%
7.02%
0.90%
0.26%
-7.51%
-4.16%
5.26%
12.57%
22.88%
40.65%
7.26%
4.03%
10.76%
18.03%
26.60%
44.18%
3.93%
5.74%
10.13%
12.11%
16.53%
25.93%
Severe Acute Respiratory Syndrome
Table 22.12 Stock market indices (percentage change, 2003)
Table 22.13 Share price movements of selected airlines (percentage change, 2003) 12–31 March
31 March– 15 April
Cathay Pacific Air Canada China Airlines Singapore Airlines China Eastern Airlines
-13.64% -20.75% -11.73% -8.38%
-8.95% -47.14% 2.47% 1.14%
0.40%
3.56%
Source: Datastream.
15–30 April
30 April– 30 May
30 May– 30 June
30 June– 31 July
12 March– 30 April
12 March– 30 May
12 March– 30 June
12 March– 31 July
12 March– 28 November
8.67% 16.22% -4.47% 6.78%
4.78% 29.46% 9.83% 1.59%
6.60% -19.76% 4.26% 8.33%
4.29% 8.21% 13.65% 3.85%
-14.55% -51.32% -13.59% -1.05%
-10.45% -36.98% -5.10% 0.52%
-4.55% -49.43% -1.05% 8.90%
-0.45% -45.28% 12.46% 13.09%
35.45% -59.25% 24.60% 27.75%
-12.40%
1.96%
-8.55%
2.34%
-8.93%
-7.14%
-15.08%
-13.10%
-22.02%
Counting the Economic Cost of SARS
Stock
229
230
Severe Acute Respiratory Syndrome
Table 22.14 Recently announced fiscal policy packages, 2003 Date package announced
Singapore Hong Kong Taiwan China Malaysia Korea
April 17 April 23 May 2 May 8 May 21 June 4
Headline size
Change in FY03 Deficit (% of GDP)
US$ bn
% of GDP
0.1 1.5 1.4 2.7 1.9 3.5
0.1 0.9 0.5 0.2 2.0 0.7
0.3 1.8 0.5 0.2 0.5 1.7
Source: Press Reports.
international travellers and the consequent spread to Vietnam, Singapore and Canada was critical in getting global public attention. The free flow of information and the consequent public reaction it provoked probably played an important role in containing the spread of the disease. The Chinese government responded quickly to contain the spread of the disease in the wake of international public pressure. Research that led to early identification of the virus and a better understanding of the attributes of the disease also calmed the panic engendered by the fear of the unknown. If any recurrence of the disease is effectively controlled, the economies in the region will not be significantly affected. SARS demonstrated to the Chinese leadership that GDP numbers are built on public and international confidence and has apparently led to greater openness on health matters in mainland China. The region as a whole also stands to benefit from a more open China whose rapid growth in the past two decades has been an important driver of regional trade expansion.
Acknowledgements The research was fully supported by a grant from the University Grants Committee of the Hong Kong Special Administrative Region, China (Project No. AoE/H-05/99). Eva Chan and June Sieh provided excellent research assistance.
References 1 Asian Development Bank. Asian Development Outlook 2003. Hong Kong, China: Oxford University Press for the Asian Development Bank, 2003. 2 Hong Kong Centre for Economic Research. Made in PRD: The Changing Face of HK Manufacturers. Hong Kong: Federation of Hong Kong Industries, 2003. 3 Asian Development Bank. SARS: Economic Impacts and Implications. Manila, Asian Development Bank, May 2003. 4 World Bank. East Asia Update: Looking Beyond Short-Term Shocks. Development Bank, Washington, DC: The World Bank, April 2003. 5 World Travel and Tourism Council. Special SARS Analysis: Impact on Travel and Tourism. London: World Travel and Tourism Council, May 2003.
Chapter 23
Preparing for a Possible Resurgence of SARS Umesh D Parashar, Angela Merianos, Cathy Roth and Larry J Anderson
Introduction Interruption of the global spread of SARS was announced by the World Health Organization on 5 July 2003.1 The success in containing this new emerging disease affirmed the power of a well-co-ordinated global response to control a rapidly spreading infectious disease. However, the speed with which SARS was able to spread globally when undetected and the relatively high mortality rate associated with infection signal the need for continued vigilance for the return of SARS and research to understand SARS better and prevent it. It is not known if or how SARS might return, but it could reemerge from several potential sources. SARS could spread to humans from the original animal reservoir as it presumably did in Guangdong Province, China, or from other wildlife species susceptible to infection.2,3 SARS-associated coronavirus (SARS CoV) has been detected for weeks after illness in stool specimens of infected patients and such persistent infection could lead to spread to other persons and re-establish spread in humans.4 It is also possible but unlikely that ongoing spread of SARS has been missed by global surveillance systems if it is circulating at a very low level in the community. Last, a break in technique could lead to laboratory-acquired infection, as illustrated by the recent SARS cases in Singapore, Taiwan, and China.5,6 While the first two incidents did not lead to any secondary transmission, seven individuals
were infected in two chains of transmission resulting from one of the individuals with laboratory-acquired infection in China, clearly illustrating the potential for laboratory-acquired SARS CoV infection to spread into the community. The response to SARS needs to be rapid and decisive. Control of SARS relies on the prompt detection and isolation of case-patients, stringent infection control within health-care settings, the appropriate use of personal protective equipment by health-care workers, and contact tracing, evaluation, monitoring, and, possibly, quarantine to prevent ongoing transmission. Since the spread of SARS is typically focal in nature and often limited to specific facilities or settings within a community, the public health response has to be tailored to characteristics of the community outbreak and the resources available to control the outbreak. Although the strategies to control and prevent SARS rely on time-tested, fundamental public health measures, preparedness to implement these measures rapidly, efficiently, and effectively is key to achieving the required communityspecific response. Preparing health systems to detect, contain, and control SARS will have broader benefits in responding to other new and emerging infectious diseases. In this chapter, we review the key lessons learned from the global SARS outbreak and describe the strategies and measures for detecting and controlling SARS.
231
232
Severe Acute Respiratory Syndrome
Lessons from SARS SARS serves as a reminder that dangerous new diseases can emerge and spread rapidly worldwide. The health, social, and economic consequences of the outbreak and the challenges in containment have taught several valuable lessons that can have longterm positive effects for public health and health care. First and foremost is the need for international co-ordination and co-operation for containing global infectious disease outbreaks. WHO provided leadership in coordinating an intense and rapidly evolving global response that was not only key to preventing the spread of SARS but also played a vital role in the rapid identification of the etiological agent, SARS CoV, and determination of its complete genomic sequence.7–13 Outbreak response teams that included personnel from global partner organizations of the Global Outbreak Alert and Response Network were rapidly mobilized and deployed to SARS-affected areas. Information sharing between these teams and global public health agencies was facilitated by frequent, regularly scheduled teleconferences and through the use of secure websites. Leading scientists and clinicians from around the world participated in an unprecedented collaborative effort to share experiences in patient management and to identify the best strategies for prevention and control of the outbreak. This intense spirit of global co-operation was one of the hallmarks of the global response to SARS. A clear organizational structure with well-defined roles and responsibilities and strong leadership and political will are necessary for an effective response to public health emergencies. Because SARS impacted all sectors of society — political, economic, social, health care, and others — strong leadership was required from the highest level of government to co-ordinate the response to allocate resources appropriately and to ensure dissemination of consistent information in a timely manner. Control of SARS required that policymak-
ers, health-care and public health professionals, community leaders, and the public worked together in a co-operative fashion. Responsible reporting by the media played a major role in disseminating information rapidly, and there was unprecedented use of the Internet in providing situation updates to the public as well as specific guidance to health professionals, policy makers and affected industries. The resources of health-care and public health systems in areas affected by the SARS outbreak were stretched to their limits and even unaffected areas were severely taxed by their efforts to investigate incoming travellers for SARS, even though most of these individuals were eventually discarded as cases. The outbreak demonstrated the need to renew public health and clinical care systems and develop surge capacity. Shortages were experienced in the availability of trained public health personnel to lead and co-ordinate an emergency response and of nursing, medical, and other health-care staff to maintain appropriate standards of patient care and infection control. In some areas, the burden of SARS led to closure of health-care facilities and hasty reconstruction of alternative facilities. Strengthening the public health and clinical care infrastructure and capacity is vital to preparations for a possible resurgence of SARS and outbreaks of other diseases. Finally, the SARS outbreak has demonstrated — once again — the importance of preparing and planning for response efforts to urgent public health threats. In many settings, infrastructure and resources put in place as part of preparedness planning for conditions such as pandemic influenza and bioterrorism were effectively utilized to mount rapidly a response to SARS. The complexity of a response to a SARS outbreak dictates that preparedness be broad-based and include (1) case identification and surveillance, (2) infection control and case isolation, (3) contact tracing, evaluation, and management, (4) laboratory diagnostics, (5) education, and (6) communication. There is also a need for counselling and
Preparing for a Possible Resurgence of SARS
other social support for patients, their families, and communities. Also, there is a need for support systems for health-care workers, since they are at high risk of infection and, at the same time, responsible for the care of others with SARS.
Key measures for containment of a SARS outbreak In the absence of a vaccine, effective drugs, or natural immunity to SARS CoV, the key to controlling SARS is the classic public health control strategy of case identification and containment. During the SARS outbreak that began in November 2002, these measures proved effective and were associated with cessation of transmission throughout the world by July 2003. The level of SARS CoV transmission during an outbreak is dynamic; thus, response activities, by necessity, must also be dynamic. The key to understanding transmission dynamics and knowing when to escalate the response at the local level is a surveillance system that provides ready access to timely information on the number of new case-patients, the likely source of exposure for case-patients, case-patients not previously identified as contacts, and contacts with high-risk exposures to known case-patients (potential prospective cases). Although jurisdictions will need to adjust the types and level of response measures to local conditions and resources, they will also need to co-ordinate with adjacent jurisdictions to ensure consistency among responses and to minimize confusion or mistrust that may derive from inexplicable differences in outbreak control strategies.
Early identification of SARS cases Rapid and early identification of SARS cases will allow for the prompt institution of infection control measures and, thus, is key to the containment of SARS. SARS has some characteristic clinical features (e.g. SARS patients usually do not develop upper respiratory tract symptoms, most are hospitalized
233
with X-ray changes of atypical pneumonia, and many develop lymphopenia),14–17 but these features are not sufficiently distinct to permit a reliable clinical diagnosis. In addition, currently available diagnostic tests for SARS do not detect infection reliably early in illness. Consequently, the diagnosis of SARS is difficult, especially during outbreaks of other respiratory illnesses, such as influenza, that can be confused with SARS. Since most SARS patients have had a clear pre-illness exposure to another SARS patient, persons with atypical pneumonia, or a setting where SARS transmission is occurring, the risk of potential exposure can be used as the key to detecting SARS. When there is no known SARS transmission in the world, surveillance for clusters of lower respiratory tract infection with fever in acute health-care settings is the approach recommended by WHO to alert health-care workers to the possibility of SARS and the need for enhanced infection control. Casefinding among sporadic cases of pneumonia can be guided by exposure to the possible sources of a SARS re-emergence — laboratory exposures, locations where animal reservoirs of SARS CoV may exist, and/or persistent infection in humans. The most likely sites of disease recurrence are the zone of emergence of the previous SARS epidemic in southern China (and possibly similar ecological niches) and locations where the most cases occurred previously. These areas require a more intense level of human surveillance, and there is a need for linked studies of the ongoing risk of SARS CoV re-emergence in humans and wildlife in southern China. In other locations where the risk of SARS is low, the diagnosis of SARS should focus on unusual clusters of atypical pneumonia (e.g. clusters in health-care workers) or on persons hospitalized with pneumonia with compelling clinical and epidemiologic evidence that SARS may be the cause of their illness. If SARS does recur, then SARS patients or known sites of SARS CoV transmission become the most likely source of exposure. As the risk of exposure and likelihood that an illness will be SARS
234
Severe Acute Respiratory Syndrome
increases, the diagnosis should be considered earlier in the illness (e.g. first sign of fever and/or respiratory symptoms) so that control measures can be implemented rapidly and further spread prevented. Since the existence of SARS cases anywhere in the world has local, national, and global implications, it is essential that the health-care and public health communities exchange information on individual SARS cases and the status of SARS transmission in the community and worldwide. The healthcare community needs to report cases rapidly and the public health community needs to alert the global community and rapidly analyse surveillance data and determine transmission patterns.
Prompt isolation of SARS case-patients Once a potential case of SARS is detected, appropriate infection control measures have to be implemented immediately. During the 2003 global epidemic, a substantial proportion of SARS cases resulted from delays in clinical recognition and isolation of patients, and transmission in health-care facilities was a major factor in the spread of the disease. In areas with extensive outbreaks, the virus spread most readily among hospital workers caring for SARS patients (21% of all cases worldwide) and among other patients and visitors to hospitals.18 For example, in Toronto, 77% of the patients in the first phase of the outbreak were infected in the hospital setting, and half of all SARS cases in Toronto were in health-care workers.19 Factors that probably contribute to the disproportionate rate of transmission in health-care settings include: (1) a higher viral load in respiratory secretions during the second week of illness when patients are likely to be hospitalized;20 (2) use of ventilators, nebulizers, endotracheal intubation, and other droplet- and aerosol-generating devices and procedures;21 and (3) frequent exposures of workers to patients, their secretions, and
potentially contaminated environments. In addition, delays in diagnosis because of atypical presentations of SARS fuelled new clusters in health-care settings or prolonged established outbreaks. SARS patients often require hospitalization because of the severity of their illness. Those who have less severe illness have sometimes been hospitalized to ensure that strict isolation procedures are followed. In other settings, patients have been cared for in residential facilities, after evaluation of their suitability for this purpose. Transmission of SARS CoV appears to occur primarily through droplets, close contact, and, possibly, fomite exposures, but limited airborne transmission cannot be excluded in some instances. Therefore, patients who require hospitalization should be admitted to an airborne-infection isolation room or a specially adapted SARS unit or ward where they can be cared for safely and appropriately, although even contact and droplet precautions appear to confer some protection.22 In some locations, a lack of isolation rooms and/or a need to concentrate infection control efforts and resources may lead to a strategy of cohorting patients in individual rooms on the same floor. This strategy physically isolates SARS patients from non-SARS patients and also makes it possible to dedicate resources and appropriately trained staff to their care. Experience in some countries, such as Taiwan and Toronto, demonstrated that such cohorting of SARS patients effectively interrupted virus transmission. Certain medical procedures, such as intubation and ventilator therapy, which generate aerosols, require use of higher standards of infection control and additional personal protective equipment.
Management of contacts of SARS case-patients Contact tracing, the identification of persons who potentially have been exposed to a disease, is essential to prevent the spread of SARS. Contact tracing provides a means
Preparing for a Possible Resurgence of SARS
of focusing control efforts on persons who are at high risk of SARS, identifying infected persons early in the course of their illness, and implementing control measures before the virus can be spread to others. Contact tracing, evaluation of contacts for possible illness, and regular, frequent monitoring for possible symptoms of SARS should be an immediate high priority to maximize the chance to control an outbreak rapidly. Many countries with extensive disease transmission during the 2003 global SARS epidemic quarantined contacts to quell the spread of virus.23–25 However, since transmission of SARS CoV is not believed to occur prior to the onset of symptoms, quarantine theoretically should be unnecessary if, at first recognition of illness, the contact notifies public health officials and is placed in isolation. However, SARS frequently has an insidious onset (e.g. with fever, headache, malaise, and myalgias), and the patient may fail to recognize or deny the likelihood of SARS or may fail to alert health officials of his or her illness and, thus, may spread the disease to others before control measures are implemented. Quarantine of exposed persons is used to prevent inadvertent exposures by separating exposed from unexposed persons and by decreasing the interval between the onset of symptoms and the institution of control measures. Since most SARS patients have a clear close contact exposure to other SARS patients or to settings with SARS transmission, contact tracing and monitoring, sometimes with restrictions, was a highly effective outbreak control strategy. Monitoring of contacts of SARS case-patients should be performed carefully and with respect for human dignity and the social, psychological, financial, and personal needs of contacts. Monitoring the compliance of persons placed in quarantine is extremely cost- and labourintensive, and, in some SARS-affected areas, thousands of persons were placed in home quarantine. To ensure that public health and personal resources are not needlessly strained by unnecessary tracing and moni-
235
toring of contacts, an efficient approach to identify high priority contacts should be developed. This can be accomplished to an extent by evaluating how the definition of a ‘close contact’ was applied during the 2003 SARS outbreak and then by evaluating its predictive value against the outbreak definition of a ‘probable case’ and ‘laboratory criteria for SARS CoV infection’.
Other vital components of SARS containment efforts Laboratory diagnostics Laboratory diagnostics are essential for detecting and documenting the reappearance of SARS CoV, responding to and managing outbreaks, and managing concerns about SARS in patients with other respiratory illnesses. The identification of SARS CoV led to the rapid development of enzyme immunoassays and immunofluorescence assays for SARS CoV antibody and reversetranscriptase polymerase chain reaction (PCR) assays for SARS CoV RNA. These assays can be very sensitive and specific for detecting antibody and RNA but can be insensitive for detecting infection in patients, especially early in the disease. Most patients in the early stages of SARS illness have a low titre of virus in respiratory and other secretions and require time to mount an antibody response.20 The diagnostic sensitivity of PCR tests is greater during the second week of illness, and stool specimens appear to have a slightly greater yield than nasopharyngeal aspirates. Limited data suggest that serum and respiratory specimens are the best specimens for SARS PCR diagnostics during the first week of illness and respiratory and stool specimens are best for PCR diagnosis later in the illness. SARS antibody results may be positive as early as 8–10 days after onset of illness and are often positive by 14 days, but sometimes they are not positive until 28 days after onset of illness. Diagnostic assays for other respiratory pathogens selected on the basis of the local
236
Severe Acute Respiratory Syndrome
epidemiology of respiratory disease may be helpful in differentiating SARS from other illnesses, but SARS patients can sometimes be infected simultaneously with SARS CoV and another respiratory pathogen. In patients with dual infections, factors such as the strength of the epidemiological exposure link to SARS, the specificity of the diagnostic tests, and the compatibility of the clinical presentation and course of illness for the alternative diagnosis should be considered when determining whether to continue considering SARS in the differential diagnosis. The diagnosis of probable SARS will trigger hospital-based and public health investigations, case isolation, and contact tracing, and it will heighten infection control within affected health-care facilities and, possibly, vigilance for SARS in the international community. In the postoutbreak period, false alarms of new cases of SARS CoV disease have resulted in renewed public concern, including loss of confidence by international financial markets, until SARS was discounted. Accordingly, clinicians, virologists, and public health physicians must work together to triage, investigate, and manage clusters or sporadic cases of acute lower respiratory tract infection to minimize the risk of false-positive SARS diagnoses. In addition, all sporadic cases and new clusters of laboratoryconfirmed SARS should be verified independently by another laboratory experienced in SARS diagnosis.
community and general awareness among all parties about the possibility of a SARS outbreak and the steps that would be indicated in such an event. The emergence of SARS anywhere in the world will probably generate immediate and intense media attention and require an enormous effort to respond to the demand from the public, the media, policymakers, and health-care providers for information and guidance. It will be important to develop key messages that will inform the public and guide them toward a response to SARS that facilitates control and minimizes unnecessary social disruption.
Education Education and training are integral to improving preparedness to respond to a SARS outbreak. Public health and health-care workers need training about the clinical aspects of SARS, features of SARS that would allow early detection of cases, appropriate use and interpretation of laboratory tests, and best practices for effective use of infection-control strategies, including use of personal protective equipment. Laboratory workers should be instructed in best practices for performing diagnostic tests for SARS, including the use of recommended biosafety precautions. Educating the public about SARS, the control of SARS, and how the public can help to control spread is also important.
Information technology Communication Rapid and frequent communication of crucial information about SARS — such as the level of the outbreak worldwide and recommended control measures — is vital to contain the spread of SARS. Specific communication needs and key messages will vary substantially by level of SARS activity. In the global absence of SARS, communications messages and materials are designed to maintain vigilance in the health-care
An efficient means to link clinical, epidemiological, and laboratory data on SARS cases and to disseminate this information locally, nationally, and globally is key for an effective SARS response. Unfortunately, in many outbreak settings in the 2003 outbreak, the lack of integrated information management systems made outbreak control less efficient and, in some instances, may have actually delayed the containment and control of SARS. Rapid identification, tracking,
Preparing for a Possible Resurgence of SARS
evaluation, and monitoring of contacts of SARS case-patients is also key for early detection of symptoms in persons at greatest risk of SARS, and development of a data management system to facilitate this process is vital. Contact tracing can be particularly challenging and resource intensive in large-scale outbreaks or among highly mobile populations, such as international travellers. Innovative contact tracing databases were used in Hong Kong and Singapore and a number of institutions are working to develop improved information management systems for the field. The tracking of contacts of SARS cases on conveyances (e.g. airplanes) will require rapid availability of electronic passenger manifests that provide information on the proximity of the contact to the case-patient. This information needs to be rapidly assimilated and disseminated for notification and monitoring of contacts.
Conclusions We have learned a great deal about SARS that is helping us prepare for the possibility that it will return. The experience has taught us that SARS can be controlled by the classic public health control strategy of early case identification and prompt institution of containment measures. Because of the multifaceted nature of a SARS response effort, the preparation for and response to an outbreak of SARS requires co-ordination and co-operation among public health authorities and other emergency response entities at all levels of government, at the local, national, and global levels. A predetermined organizational structure that addresses planning, operations, logistics, finance, and administration is essential for maximizing the use of limited resources, monitoring the status of an outbreak, and consolidating the control of a large number of individual resources. The success of efforts to detect rapidly, respond to, and contain an outbreak also depends in large part on the availability of information systems
237
that integrate all facilities and personnel involved in the response, expedite real-time communication and flow of information, aid in logistics planning and resource management/allocation, and facilitate decision making and operational co-ordination. Preparation and planning are essential to ensure an effective response to public health threats such as SARS, and investments made in the public health and health-care systems to increase preparedness for SARS will also pay dividends in preparedness to confront other emerging infectious diseases and other threats to public health.
References 1 World Health Organization. Update 96 — Taiwan, China: SARS transmission interrupted in last outbreak area. http://www.who.int/csr/ don/2003_07_05/en/ 2 Guan Y, Zheng BJ, He YQ et al. Isolation and characterization of viruses related to the SARS coronavirus from animals in Southern China. Science 2003;302: 276–8. 3 Martina BE, Haagmans BL, Kuiken T et al. SARS virus infection of cats and ferrets. Nature 2003;425: 915. 4 Leung WK, To KF, Chan PK et al. Enteric involvement of severe acute respiratory syndrome-associated coronavirus infection. Gastroenterology 2003;125:1011–17. 5 Lim PL, Kurup A, Gopalakrishna G et al. Laboratory-acquired severe acute respiratory syndrome. N Engl J Med 2004;350: 1740–5. 6 World Health Organization. Update 7 — China’s latest SARS outbreak has been contained, but biosafety concerns remain. http:// www.who.int/csr/don/2004_05_18a/en/ 7 Stohr K, World Health Organization Multicentre Collaborative Network for SARS Diagnosis. A multicentre collaboration to investigate the cause of severe acute respiratory syndrome. Lancet 2003;361: 1730–3. 8 Peiris JSM, Lai ST, Poon LLM et al. Coronavirus as a possible cause of severe acute respiratory syndrome. Lancet 2003;361: 1319–25. 9 Drosten C, Gunther S, Preiser W et al. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. N Engl J Med 2003;348: 1967–76. 10 Ksiazek TG, Erdman D, Goldsmith C et al. A novel coronavirus associated with severe
238
11
12
13
14
15
16
17
18
Severe Acute Respiratory Syndrome
acute respiratory syndrome. N Engl J Med 2003;348: 1953–66. Kuiken T, Fouchier RAM, Schutten M et al. Newly discovered coronavirus as the primary cause of severe acute respiratory syndrome. Lancet 2003;362: 263–70. Marra MA, Jones SJM, Astell CR et al. The genome sequence of the SARS-associated coronavirus. Science 2003;300: 1399–404. Rota PA, Oberste MS, Monroe SS et al. Characterization of a novel coronavirus associated with severe acute respiratory syndrome. Science 2003;300: 1394–9. Lee N, Hui D, Wu A et al. A major outbreak of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348: 1986–94. Tsang KW, Ho PL, Ooi GC et al. A cluster of cases of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348: 1977–85. Poutanen SM, Low DE, Henry B et al. Identification of severe acute respiratory syndrome in Canada. N Engl J Med 2003;348: 1995–2005. Booth CM, Matukas LM, Tomlinson GA et al. Clinical features and short-term outcomes of 144 patients with SARS in the Greater Toronto area. JAMA 2003;289: 2801–9. World Health Organization. Cumulative number of reported probable cases of severe acute respiratory syndrome (SARS). Available at http://www.who.int/csr/sars/country/en/
19 Centers for Disease Control and Prevention. Update: severe acute respiratory syndrome — Toronto, Canada, 2003. MMWR 2003;52: 547–50. 20 Peiris JSM, Chu CM, Cheng VCC et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;361: 1767–72. 21 Wong TW, Lee CK, Tam W et al. Cluster of SARS among medical students exposed to a single patient, Hong Kong. Emerg Infect Dis 2004;10: 269–76. 22 Seto WH, Tsang D, Yung RWH et al. Effectiveness of precautions against droplets and contact in prevention of nosocomial transmission of severe acute respiratory syndrome (SARS). Lancet 2003;361: 1519–20. 23 Liang W, Zhu Z, Guo J et al. Severe acute respiratory syndrome, Beijing, 2003. Emerg Infect Dis 2004;10: 2. 24 Centers for Disease Control and Prevention. Efficiency of quarantine during an epidemic of severe acute respiratory syndrome — Beijing, China, 2003. MMWR 2003;52: 1037–40. 25 Centers for Disease Control and Prevention. Use of quarantine to prevent transmission of severe acute respiratory syndrome — Taiwan, 2003. MMWR 2003;52: 680–3.
Chapter 24
Lessons for the Future: Pandemic Influenza Robert G Webster and David S Fedson
Introduction Severe acute respiratory syndrome (SARS) is the first new plague of the twenty-first century, but it is certain that others will follow. The Foreword to this volume and a recent report of the Institute of Medicine of the National Academy of Sciences puts SARS and other emerging infectious diseases into perspective. ‘In the highly interconnected and readily traversed “global village” of our time, one nation’s problem soon becomes every nation’s problem, as geographical and political boundaries offer trivial impediments to such threats.’ 1 From a historical perspective, influenza pandemics have swept the world at irregular intervals.2,3 In the past century, there were three pandemics: Spanish influenza in 1918, Asian influenza in 1957, and Hong Kong influenza in 1968. The most devastating was Spanish influenza that killed 50–100 million people worldwide.4 More recent pandemic threats include the reemergence of Russian influenza in 1977 and several outbreaks of human infection with avian influenza viruses since 1997. Influenza virologists consider the next pandemic imminent. Thus, it is useful to ask what lessons have been learnt from the SARS experience that are relevant for global preparedness for pandemic influenza.
SARS and influenza: similarities and differences Perhaps the most important question at this time is whether SARS will behave like influenza. From a clinical viewpoint, several differences are already apparent, as discussed elsewhere in this volume. From an epidemiololgical viewpoint, the mode of transmission of SARS coronoavirus (CoV) appears to be different and the virus is less infectious; its reproductive number (Ro) is between 2 and 4, whereas the Ro for influenza is ~10.5,6 Sources of SARS CoV certainly continue to exist in the northern hemisphere in animal reservoirs and now in many laboratories. The initial outbreak of SARS coincided with the re-emergence of human cases of avian influenza A (H5N1), and many elements of the approach to influenza surveillance and control by the World Health Organization (WHO) and national health authorities were immediately applied to SARS. If outbreaks of SARS do not re-emerge in human populations, the fact that SARS CoV has been contained by the full application of classical measures of public health — comprehensive surveillance, meticulous attention to hygiene and rigorous contact tracing and management (e.g. quarantine)— coupled with modern methods of scientific and public communication7 will become more convincing. It is unlikely, however, that these measures alone will be sufficient to control pandemic influenza.
239
240
Severe Acute Respiratory Syndrome
In this chapter we review the SARS experience in order to determine what has been learnt and, more importantly, what still needs to be learnt and applied in order to confront the next influenza pandemic. We focus on four areas: (1) organizing the public health response, (2) improving human and animal surveillance, (3) ensuring supplies of affordable antiviral agents and vaccines, and (4) addressing the economic and political dimensions of SARS and their implications for pandemic influenza.
Organizing the public health response The foundation for organizing the international public health response to pandemic influenza was set forth in 1999 in the WHO Influenza Pandemic Preparedness Plan.8 The plan describes the responsibilities and actions to be taken by WHO during the interpandemic period, when a pandemic threat emerges and during and following the pandemic. In addition to co-ordinating the international flow of scientific and public information, WHO has defined different phases and levels of alert for a pandemic and stipulated the actions it will take. Similarly, WHO has published its plan for alert, verification and public health management of SARS in the post-outbreak period.9 The WHO SARS plan defines three levels of activity that take account of an area’s recent experience with SARS and the local potential for resurgence. In addition to providing clinical and laboratory case definitions, the WHO plan outlines an approach to public health management of a SARS alert, focusing on international reporting requirements and levels of surveillance for animal and human populations, especially groups at high risk such as health-care workers. The implementation of the public health response to SARS by WHO and by many of its Member States has been impressive. For example, on 5 June 2003, the European Commission published the results of its survey of measures being undertaken by both
current and future Member States to control the SARS outbreak.10 In a matter of a few months, European countries were found to have provided guidance and adopted rapid measures to detect cases, implemented contact tracing and isolation procedures, outlined measures for infection control and the protection of health-care workers, provided guidance and information to international travellers and the public, specified measures for screening travellers arriving from affected areas, provided guidance on the organization of laboratory services and the handling of clinical specimens, and organized numerous channels for communicating this information. In Canada, a country seriously affected by SARS, the public health response was even more impressive.11 The same can be said for Hong Kong12 and other countries, as documented elsewhere in this volume. Many countries have developed detailed plans for a public health response if SARS should return.13 Plans for the public health response to pandemic influenza have emerged gradually over the past few years. Although WHO has called on all Member States to prepare national pandemic preparedness plans and many countries have begun doing so, relatively few plans have been published.14,15 Nonetheless, many countries have been engaged in an ongoing process of pandemic preparedness, and these activities were quickly adapted to the challenge of SARS.16 Many of the planning documents that have been prepared to address the possible reemergence of SARS can be adapted to meet other threats such as bioterrorism or pandemic influenza. It should be remembered, however, that a public health emergency such as SARS, once recognized, always gets attention and resources. A future influenza pandemic will place a far greater burden than SARS on public health and health-care systems throughout the world. Unfortunately, however, no country has given pandemic preparedness planning anything like the level of attention and resources given to SARS.
Lessons for the Future: Pandemic Influenza
Improving human and animal surveillance for SARS CoV and influenza viruses Human surveillance Growing international awareness of the seriousness of emerging infectious diseases led WHO to establish the Global Outbreak and Response Network in 2000. On 11 February 2003 the network was alerted to an outbreak of atypical pneumonia of unknown cause in China’s Guangdong Province. One week later, two cases of avian influenza A H5N1 were diagnosed in Hong Kong, leading many to believe these cases were the first signs of a new pandemic. The outbreak in China was quickly shown not to be due to influenza, and instead was shown to be due to a new coronavirus. As detailed elsewhere in this volume, in the absence of a diagnostic test, human surveillance initially was based on epidemiological and clinical criteria. Many suspect cases were later confirmed as SARS by RTPCR or serological testing. Nonetheless, epidemiological and clinical criteria alone provided a remarkably sound basis for the eventual success in controlling outbreaks of the disease. Surveillance of influenza in human populations was begun by WHO in 1947. Initial efforts focused on virus isolation and serological testing, and the results of these studies became the basis for the annual decision by WHO and its Collaborating Centers on the strain composition of influenza vaccines. WHO’s virological surveillance currently is based on a network of 112 National Influenza Centers, not all of which are fully functional.17 In the late 1960s, information also began to be gathered by sentinel practitioners on the impact of influenza in patients.18 Systems for practice-based surveillance have now been established in 18 countries in Europe and in North America, and those in European countries are now co-ordinated by the European Influenza Surveillance Scheme (EISS). In addi-
241
tion, methods for estimating the extensive impact of influenza on hospitalizations and mortality also have been developed.19,20 Ongoing surveillance of human populations for outbreaks of serious respiratory disease will continue to provide the first signals for the re-emergence of SARS and pandemic influenza. In both instances, laboratory confirmation will be essential, and extensive networks are now in place to transport patient specimens rapidly to laboratories where the viruses can be identified. However, the networks are not complete, and for both diseases there remain large areas of the world where human surveillance is non-existent and regional laboratories are incapable of processing diagnostic specimens. Should either disease appear in one of these areas, it will spread rapidly to distant sites before any control measures can be undertaken. Thus, anticipating the re-emergence of the two diseases may be crucially dependent on effective surveillance for SARS CoV and influenza viruses in their animal hosts.
Animal surveillance The SARS coronavirus and influenza viruses are both zoonotic agents. Coronaviruses are widely distributed in both avian and mammalian species. Soon after a new coronavirus was shown to the causative agent for SARS, it was isolated from Himalayan palm civets (Paguma larvata) and other animals in live markets in China. Serological evidence of infection also was found in raccoon dogs (Nyctereutes procuyoinboides) and in humans working in the live animal markets.21 The markets were closed for a short while, but later reopened, provoking much concern from health officials in the region. The natural reservoirs of influenza viruses are aquatic birds.22 Live poultry markets provide influenza viruses with ideal environments for genetic reassortment and interspecies transfer, as was shown in 1997 when the markets in Hong Kong were linked to human infection with an avian
242
Severe Acute Respiratory Syndrome
influenza A H5N1 virus.23–26 In the United States, a similar link has been established for the spread of H5N2 and H7N2 influenza viruses in commercial poultry.27 The surveillance activities of influenza virologists in Hong Kong and their colleagues in the WHO Collaborating Center in Memphis, Tennessee have provided the basis for control programmes in Hong Kong. Enormous efforts have been undertaken in the poultry markets to interrupt virus transmission, including the elimination of live ducks, geese (original sources of the virus) and quail (an intermediate source) and the initiation of monthly ‘clean’ days when all markets are simultaneously emptied and cleaned.28,29 The use of an inactivated H5 vaccine in poultry and improvements in biosecurity seem to have had some effect, but avian influenza viruses have continued to emerge. Moreover, similar interventions have not been undertaken in the live poultry markets in mainland China, and precursors of H5N1 influenza viruses continue to circulate in aquatic birds in the region, becoming sources for newly emergent H5N1 viruses.30,31 These viruses have continued to circulate and reassert themselves in Asia.32,33 This has led to the emergence of a dominant virus genotype that caused a major regional outbreak of H5N1 avian influenza in 2004, affecting 8 counties in Asia and leading to the deaths of over 100 million poultry through disease or slaughter, in attempts to stamp out the outbreak. The avian H5N1 influenza virus was transmitted to humans in Vietnam and Thailand causing severe disease and death.34–36 Because live poultry and animal markets are ideal sites for influenza and coronaviruses to exchange viral genes and be transmitted to other species, including humans, closing these markets might be considered. If this were done, the markets would probably move ‘underground’ and escape any kind of regulation. A better approach would be to improve the management of existing markets. Public health officials in Hong Kong have established such a programme
for influenza and it could also serve as a model for controlling coronaviruses. However, public health officials in mainland China missed an opportunity to do this for SARS CoV37 and they have yet to implement Hong Kong-like control programmes for avian influenza in their live poultry markets. Eventually, improved economic conditions, including widespread purchase of household refrigerators, will make these markets unnecessary, but it will take years for this to happen.
Ensuring supplies of affordable antiviral agents and vaccines Antiviral agents Intensive research has begun on SARS CoV in order to identify targets for antiviral activity and screen for agents that show promising antiviral activity. It is too early to know if this research will be successful, but if it is, there will be great pressure to ensure that effective antiviral agents for SARS are made available to all who need them. Experience with antiviral agents for influenza indicates that this might be extremely difficult. Two types of antiviral agents are available for influenza: the ion channel (M2) inhibitors amantadine and rimantidine and the neuraminidase inhibitors zanamivir (Relenza™) and oseltamivir (Tamiflu™). Both types of agents are effective in prophylaxis against infections with type A influenza viruses and both limit the course of illness in those who are infected.38,39 Therapeutic use of amantadine and rimantidine, however, is regularly accompanied by the emergence of antiviral resistance, thus limiting their use. Resistance to the neuraminidase inhibitors also occurs, but is much more difficult to achieve. Although both types of agents are imperfect, they could be of great benefit in the event of a pandemic. Unfortunately, they are in very limited supply. Only one antineuraminidase agent
Lessons for the Future: Pandemic Influenza
(oseltamivir) is currently being marketed actively by its manufacturer, and the lead time necessary to increase production capacity is at least 18 months. If these agents are to be available for use in a pandemic, they will have to be stockpiled well in advance of a pandemic threat. In spite of recent experience with anthrax bioterrorism and the recognized need for stockpiles of antimicrobial agents, few countries have implemented programmes to stockpile antiviral agents for an influenza pandemic.
Vaccines Research needed to understand the genetic mutability of the SARS CoV, the immunopathogenesis of the disease, the immune correlates of protection and the potential for immune escape is proceeding rapidly, but most of the basic questions have yet to be answered.40 If a SARS vaccine looks promising, demonstrating its safety in animal models of infection will be essential because of the known enhancement of coronavirus disease in some animal species following vaccination with certain coronavirus vaccines. It will be impossible to demonstrate directly the safety of a SARS vaccine in humans in the absence of a return of the disease. If SARS does return, a vaccine believed to be safe from animal studies probably will be used straightaway, without a Phase III efficacy trial, and vaccination effectiveness will be determined in subsequent observational studies. If reemergent SARS becomes widespread, particularly in developing countries, getting supplies of an affordable vaccine to these countries will be an enormous challenge. Unlike SARS, there is an effective and not very expensive vaccine for influenza.41 Moreover, the use of influenza vaccine is increasing, especially in rapidly developing countries. In year 2000, approximately 34% of the almost 250 million doses of influenza vaccine distributed worldwide were used in countries outside North America, Western
243
Europe and Australia and New Zealand.42 Approximately 85% of all doses distributed were produced in nine (now eight) countries, indicating that there is a substantial flow of vaccine from producer to nonproducer countries.43 When the next pandemic threat appears, vaccine companies will almost certainly produce a monovalent vaccine and, with the addition of an alum adjuvant, they eventually might be able to increase their overall production to 3 to 6 billion doses. This assumes that vaccine seed strains can be prepared that grow efficiently in the production systems of the vaccine manufacturers. However, in the six years following the 1997 isolation of avian influenza A H5N1 viruses from fatal cases in Hong Kong, no commercially viable H5N1 vaccine seed strains could be produced using conventional reassortant techniques. Fortunately, in early 2003, when a new influenza A H5N1 virus was isolated from human cases in Hong Kong, the techniques of reverse genetics were used to prepare high-growth reference strains within a few weeks.44,45 It is generally agreed that it will be necessary to use reverse genetics to prepare seed strains for the next pandemic vaccine. However, the intellectual property rights for reverse genetics are owned by two institutions. Moreover, several countries might be reluctant to allow companies to produce vaccines from seed strains that are considered to be ‘genetically modified organisms’. Given these difficulties, it would be prudent to introduce reverse genetics into the routine production of interpandemic influenza vaccines before a pandemic threat appears. Yet, vaccine companies have little incentive to pay the royalties necessary to produce reverse-geneticsengineered interpandemic vaccines when they will still sell for the same low prices. It is too early to know whether intellectual property rights will cause similar problems for producing a SARS vaccine. This seems unlikely because most of the patent applications have come from public institutions
244
Severe Acute Respiratory Syndrome
and these institutions are likely to find it difficult to manage their intellectual property for strictly commercial purposes.
Addressing the economic and political dimensions of SARS and their implications for pandemic influenza Economists have yet to determine the full costs of SARS to the international economy, but early estimates have been as high as $100 billion, with countries in east and south-east Asia and Canada bearing the brunt of the costs.46 Travel-related service industries were hit immediately, but exportoriented manufacturing was also affected, and direct foreign investment in Asian countries was threatened. Following the WHO declaration that SARS had been controlled, economic life in the affected countries recovered quickly. If SARS should re-emerge and threaten to spread worldwide, its economic effects will almost certainly be more severe and longer lasting. For pandemic influenza, the global economic costs have not been estimated, but for the immediate pandemic period they will probably dwarf those seen with SARS. In the United States alone, the combined direct and indirect economic costs of a pandemic have been estimated at $71 to $166.5 billion, not including the costs of disrupting social and commercial life.47 The international political response to SARS has been extraordinary. The foundation for this response was laid in the 1980s (if not before) when the framework for international health regulations began to break down.48 The acceleration of globalization and the appearance of HIV/AIDS and other emerging infectious diseases revealed the inadequacies of traditional approaches to preventing their spread from one country to the next. Protecting commercial activities in developed countries became less important than paying attention to the human rights of people in developing countries and improving their access to essential
drugs and vaccines. As the concept of global public goods for health began to take hold, a broader framework for global health governance began to unfold, with nongovernmental organizations and public– private partnerships for health taking their places alongside WHO and national governments in the governance of global health.48 In response to these changes, in 1995 the World Health Assembly directed WHO to undertake a broad revision of its International Health Regulations, giving specific attention to developing ‘criteria to define a public health emergency of international concern’ and outlining the ‘core capacities . . . needed to operate national systems for disease surveillance and response. . . ’.49 When the World Health Assembly met in May 2003 in the midst of the SARS crisis, the Director-General of WHO observed that SARS had ‘given concrete expression to . . . the inadequacy of the current Regulations and the urgent need for WHO and its international partners to undertake specific actions not addressed by the Regulations’.50 The World Health Assembly was asked to urge each Member State to establish ‘a national standing task force . . . with operational responsibilities . . . to ensure the speed of both reporting to WHO and consultation with national authorities when urgent decisions must be made’. After informing the governments concerned, the Director-General was instructed, ‘to alert the international community of (international) public health threats’, and ‘to conduct on-the-spot studies (to ensure) that appropriate control measures are being employed’.50 These instructions were a natural followon to the actions taken earlier by WHO once the implications of SARS had become apparent. WHO issued its global warning on SARS on 12 March 2003 and the first of several travel advisories on 15 March. These announcements were unprecedented in that they were directed not to the Member States but to the international community at large, and represented actions not specified in
Lessons for the Future: Pandemic Influenza
WHO’s Constitution.48 Remarkably, opposition to the WHO announcements by several countries questioned their appropriateness, not their legitimacy. These developments provide precedents for what could be a more activist role by WHO in response to a pandemic influenza.
Lessons learnt from SARS and their implications for pandemic influenza For WHO, the SARS experience has demonstrated (1) the weaknesses of many national surveillance systems and the crucial importance of prompt and open reporting of disease outbreaks, (2) the inadequate surge capacity of health-care facilities and staffing, (3) the necessity for international scientific and medical collaboration at the highest levels, (4) the essential contributions of responsible media and electronic communication to global understanding and vigilance, and (5) the absolutely decisive role of political leadership, commitment and solidarity at the national and international level.51 As WHO stated in its report to the World Health Assembly, ‘The lessons learned so far are useful in assessing global capacity to respond to other infectious disease threats, including the next influenza pandemic. . . ’.51 In 2002, WHO gave notice of its expanded influenza programme when it published its Global Agenda on Influenza Surveillance and Control.52 After extensive consultation with influenza experts in several fields, the agenda was organized around four main objectives: (1) enhancing and integrating virological and disease surveillance, (2) increasing knowledge of the health and economic burden of influenza, (3) increasing influenza vaccine use, and (4) accelerating national and international action on pandemic preparedness. In a move that will have far-reaching effects, in May 2003 the World Health Assembly passed a resolution on the prevention and control of influenza pandemic and annual epidemics.53 Subse-
245
quently, WHO published draft guidelines on the use of vaccines and antiviral agents during a pandemic.54 After being sidetracked by SARS, WHO staff gradually are resuming the implementation of the global agenda on influenza. Although SARS and pandemic influenza share many common features, it is important to recognize that the global response to SARS was one of reaction to an existing crisis, whereas the response to a future pandemic threat must be one of anticipation. For SARS, the absence of a diagnostic test, effective antiviral agent or vaccine forced health officials to rely on surveillance networks, including those used for influenza, and rigorous control measures, many of which were common in the pre-antibiotic era. Virological and ecological studies quickly demonstrated the intimate connections between its animal and human hosts, lessons long familiar to influenza experts. These findings provide compelling evidence that preventing the re-emergence of SARS CoV in human populations might depend on interrupting its transmission from animal to human hosts, in much the same way that control of avian influenza has been largely achieved in the poultry markets of Hong Kong. Thus, WHO’s emphasis on strengthening national surveillance systems must not overlook importance of animal as well as human surveillance for both SARS CoV and influenza. The SARS experience has provided valuable lessons in what can be accomplished with modern techniques of communication amongst scientists and to the public. However, SARS also has demonstrated the difficulties faced by health-care systems in responding to the demand for hospital care and in managing its impact on hospital staffing. These problems are familiar to those who deal with influenza. In response to economic pressures, many health-care systems have reduced their hospital bed capacities and as result have experienced greater problems in responding to winter
246
Severe Acute Respiratory Syndrome
outbreaks of interpandemic influenza.55 SARS has provided a stark warning of what to expect on a much wider scale with the next pandemic. Unlike SARS, health-care systems already possess the necessary tools — antiviral agents and vaccines — with which to confront pandemic influenza. Yet, knowing these tools exist is no guarantee that they will be accessible and affordable to those who will need them. The global production capacity for antiviral agents is extremely limited and cannot be scaled up after a pandemic threat appears. Hence, there have been repeated calls for creating stockpiles of antiviral agents. Unfortunately, no one has proposed a workable plan for how this could be accomplished. For pandemic vaccines, prospects are more hopeful. In many countries, recommendations for influenza vaccination in inter-pandemic years have been expanded.42,56 The recommendations should have a positive impact on vaccine production capacity and, more importantly, on the social infrastructure needed for vaccine delivery. Although stockpiles of pandemic vaccines cannot be created beforehand and it will take at least 6 months if not more for them to become available, vaccine companies currently have the theoretical capacity to produce several billion doses of monovalent, low dose haemagglutinin, alumadjuvanted pandemic vaccines.43 If the regulatory and intellectual property rights issues for reverse genetics can be sorted out during interpandemic years, there is reason to hope that pandemic vaccine production could move ahead swiftly. Whether pandemic vaccines will be distributed equitably is another matter. It is almost certain that political leaders of vaccine-producing countries will not allow the export of pandemic vaccines until their domestic needs have been met. Once this is achieved, determining which of the nonvaccine-producing countries will be supplied first will depend on political, not commercial decisions. Failure to deal equitably with the problem of vaccine supply
for ‘have’ and ‘have not’ countries carries with it the potential for an international diplomatic disaster. It is probably too much to expect equitable decisions on vaccine allocation to be made by national political leaders in the weeks immediately following the emergence of a pandemic threat. Thus a workable approach to the political dimensions of pandemic vaccine supply must be developed in advance. This will require international mediation at the highest level. Negotiating a successful solution to this problem will be doubly important because it will have immediate applicability to the global distribution of antiviral agents and vaccines against SARS.57 Many have said that SARS has been a ‘wake-up’ call to the international health community. It has taught scientists, health officials and political leaders valuable lessons on how to collaborate and communicate on a sudden health crisis. But SARS also has highlighted how much is yet to be learnt about successfully preparing for a future health crisis such as pandemic influenza.
Acknowledgements The writing of this article was supported in part by Public Health Service Grant AI29680 to RG Webster and by Cancer Center Support (CORE) Grant CA-21765 from the National Institutes of Health and by the American Lebanese Syrian Associated Charities (ALSAC). We thank Carol Walsh for manuscript preparation.
References 1 Smolinski MS, Hamburg MA, Lederberg J. Microbial Threats to Health: Emergence, Detection and Response. Washington, DC: The National Academies Press, 2003. 2 Potter C. Chronicles of influenza pandemics. In: Nicholson KG, Webster RG, Hay AJ, eds. Textbook of Influenza. Oxford: Blackwell Science, 1998. 3 Nguyen-van-Tam JS, Hampson AW. The epidemiology and clinical impact of pandemic influenza. Vaccine 2003;21: 1762–8.
Lessons for the Future: Pandemic Influenza
4 Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 ‘Spanish’ influenza pandemic. Bull Hist Med 2002;76: 105–15. 5 Riley S, Fraser C, Donnelly CA et al. Transmission dynamics of the etiological agent of SARS in Hong Kong: impact of public health interventions. Science 2003;300: 1884–5. 6 Lipsitch M, Cohen T, Cooper B et al. Transmission dynamics and control of severe acute respiratory syndrome. Science 2003;300: 1966– 70. 7 Eysenbach G. SARS and population health technology. J Med Internet Res 2003;5: e14. 8 World Health Organization. Influenza pandemic preparedness plan. The role of WHO and guidelines for national and regional planning. Geneva: World Health Organization, 1999: 1–65. Available at http://www.who.int/emc/ diseases/flu/index.html. Accessed 3 June 2002. 9 World Health Organization. Alert, verification and public health management of SARS in the post-outbreak period. 2003: 1–9. Available at http://www.who.int.csr/sars/postoutbreak/ en/. Accessed 2 October 2003. 10 European Commission. Measures undertaken by Member States and Accession Countries to control the outbreak of SARS. Report by the Commission, 5 June 2003. 2003: 1–23. Available at http://europa.eu.int/comm/ health/ph_threats/com/sars/sars_measures_ en.pdf Accessed 2 October 2003. 11 Government of Canada. Learning from SARS. Renewal of Public Health in Canada. A report of the National Advisory Committee on SARS and Public Health, October 2003. Available at http://www.hc-sc.gc.ca/ english/protection/warnings/sars/learning. html. Accessed 4 October 2003. 12 Hong Kong SARS Expert Committee. Available at http://www.SARS-expertcom.gov.HK. Accessed 14 October 2003. 13 Centers for Disease Control and Prevention. Draft: Public health guidance for community-level preparedness and response to Severe Acute Respiratory Syndrome (SARS). Available at http:// www.cdc.gov/ncidod/sars/sarsprepplan.htm. Accessed 20 October 2003. 14 Zambon M, Joseph C. The PHLS plan for pandemic influenza 2001: 1–30, Available at http://www.hpa.org.uk/ infections/publications/pdf/pandemicplan. pdf. Accessed 4 October 2003. 15 Paget WJ, Aguilera JF Influenza planning in Europe. Eurosurveillance 2001;6: 136–40. 16 Jennings L, Lush D. National pandemic planning must be an ongoing process. In: Kawaoka Y (ed). Options for the Control of Influenza V. Amsterdam: Elsevier BV 2004; 230–34.
247
17 World Health Organization. WHO Global Influenza Programme: survey on capacities of national influenza centres, JanuaryJune 2002. Wkly Epdemiol Rec 2002;77: 350–8. 18 Fleming DM, van der Velden J, Paget WJ. The evolution of influenza surveillance in Europe and prospects for the next 10 years. Vaccine 2003;21: 1749–53. 19 Thompson WW, Shay DK, Weintraub E et al. Mortality associated with influenza and respiratory syncytial virus in the United States. JAMA 2003;289: 179–86. 20 Simonsen L, Blackwelder WC, Reichert TA et al. Estimating deaths due to influenza and respiratory syncytial virus. JAMA 2003;289: 2499–500. 21 Guan Y, Zheng BJ, He YQ et al. Isolation and characterization of viruses related to the SARS coronavirus from animals in southern China. Science 2003;302: 276–8. 22 Webster RW, Bean WJ, Gorman OT et al. Evolution and ecology of influenza A viruses. Microbiol Rev 1992;56: 152–79. 23 Claas EC, Osterhaus AD, VanBeek R et al. Human influenza A H5N1 virus related to a highly pathogenic avain influenza virus. Lancet 1998;351: 472–7. 24 Suarez DL, Perdue ML, Cox N et al. Comparisons of highly virulent H5N1 influenza A viruses isolated from humans and chickens from Hong Kong. J Virol 1998;72: 6678–88. 25 Lin YP, Shaw M, Gregory V et al. Avian-tohuman transmission of H9N2 subtype influenza A viruses: Relationship between H9N2 and H5N1 human isolates. Proc Natl Acad Sci USA 2000;97: 9654–8. 26 Bridges CB, Lim W, Hu-Primmer J et al. Risk of influenza A (H5N1) infection among poultry workers, Hong Kong, 1997–1998. J Infect Dis 2002;185: 1005–10. 27 Senne DA, Pearson JE, Panigraphy B. Live poultry markets: A missing link in the epidemiology of avian influenza. In: Proceedings of the Third International Symposium on Avian Influenza, Madison, WI: University of Wisconsin, 1993: 7–19. 28 Sims LD, Ellis TM, Liu KK et al. Avian influenza in Hong Kong 1997–2002. Avian Dis 2003;47: 832–8. 29 Sims LD, Guan Y, Ellis TM et al. An update on avian influenza in Hong Kong 2002. Avian Dis 2003;47: 1083–6. 30 Liu M, He S, Walker D et al. The influenza virus gene pool in a poultry market in south central China. Virology 2003;305: 267–75. 31 Webster RG, Guan Y, Peiris M et al. Characterization of H5N1 influenza viruses that continue to circulate in geese in south eastern China. J Virol 2002;76: 118–26.
248
Severe Acute Respiratory Syndrome
32 Guan Y, Peiris JSM, Lipatov AS et al. Emergence of multiple genotypes of H5N1 avian influenza viruses in Hong Kong SAR. Proc Natl Acad Sci USA 2002;99: 8950–5. 33 Sturm-Ramirez KM, Ellis T, Bousfield B et al. Re-emerging H5N1 influenza viruses in Hong Kong in 2002 are highly pathogenic to ducks. J Virol 2004;78: 4892–901. 34 World Health Organization. Avian influenza A (H5N1). Wkly Epidemiol Rec 2004;79: 65–76. 35 Hien TT, Liem NT, Dung NT et al. Avian influenza A (H5N1) in 10 patients in Vietnam. N Engl J Med 2004;350: 1179–88. 36 Li KS, Guan Y, Wang J et al. Genesis of a highly pathogenic potentially pandemic H5N1 influenza virus in eastern Asia. Nature 2004;430: 209–13. 37 Enserink M. SARS in China: China’s missed chance. Science 2003;301: 294–6. 38 Hayden FG. Perspectives on antiviral use during pandemic influenza. Philos Trans R Soc [Biol], 2001;356: 1877–84. 39 Kaiser L, Wat C, Mills T et al. Impact of oseltamivir treatment on influenza-related lower respiratory complications and hospitalizations. Arch Intern Med 2003;163: 1667–72. 40 DeGroot AS. How the SARS vaccine effort can learn from HIV — speeding towards the future, learning from the past. Vaccine 2003;21: 4095–104. 41 Nichol KL. The efficacy, effectiveness and cost-effectiveness of inactivated influenza vaccines. Vaccine 2003;21: 1769–75. 42 van Essen GA, Palache AM, Forleo E et al. Influenza vaccination in 2000: recommendations and vaccine use in 50 developed and rapidly developing countries. Vaccine 2003;16: 1780–85. 43 Fedson DS. Pandemic influenza and the global vaccine supply. Clin Infect Dis 2003;36: 1552–61. 44 Webby RJ, Perez, DR, Coleman JS et al. Responsiveness to a pandemic alert: use of reverse genetics for rapid development of influenza vaccines. Lancet 2004;363: 1099–103. 45 Stephenson I, Nicholson KG, Wood JM, Zambon MC, Katz JM. Confronting the avian influenza threat: vaccine development for a potential pandemic. Lancet Infect Dis 2004;4: 499–509. 46 Monaghan K. SARS: down but still a threat. Unclassified report of the National Intelligence Center, Central Intelligence Agency.
47
48
49
50
51
52
53
54
55
56
57
2003:.1–25. Available at http://www. cia.gov/nic/pubs/other_products/SARS/ICA0 3_09.htm. Accessed 4 October 2003. Meltzer MI, Cox NJ, Fukuda K. The economic impact of pandemic influenza in the United States: priorities for intervention. Emerg Infect Dis 1999;5: 659–71. Fidler DP. SARS: political pathology of the first post-Westphalian pathogen. J Law Med Ethics 2003;31: 485–505. World Health Assembly. Revision of the International Health Regulations. Report by the Secretariat. Fifty-sixth World Health Assembly. A56/25. 24 March 2003. Available at http:// who.int/gb. Accessed on 2 October 2003. World Health Assembly. Revision of the International Health Regulations. Report by the Secretariat. Fifty-sixth World Health Assembly. A56/25 Add.1. 16 May 2003. Available at http://www.who.int/gb. Accessed on 2 October 2003. World Health Assembly. Revision of the International Health Regulations. Severe acute respiratory syndrome (SARS). Report by the Secretariat. Fifty-sixth World Health Assembly. A56/48. 17 May 2003. Available at http:// www.who.int/gb. Accessed on 2 October 2003. Stohr K. The global agenda on influenza surveillance and control. Vaccine 2003;21: 1744–8. World Health Assembly. Prevention and control of influenza pandemics and annual epidemics. Fifty-sixth World Health Assembly. WHA56.19. 28 May 2003. Available at http://www.int/gb. Accessed on 2 October 2003. World Health Organization. Draft WHO guidelines on the use of vaccines and antivirals during influenza pandemics. Wkly Epidemiol Rec 2003;47: 394–404. Available at http://www.who.int/influenza. Accessed 2 October 2003. Glaser CA, Gillian S, Thompson WW et al. Medical care capacity for influenza outbreaks, Los Angeles. Emerg Infect Dis 2002;8: 569–74. Schabas RE. Mass influenza vaccination in Ontario: a sensible move. CMAJ 2002;164: 36–7. Fedson DS. Vaccination for pandemic influenza and severe acute respiratory syndrome: common issues and concerns. Clin Infect Dis 2003;36: 1562–3.
Chapter 25
Lessons Learnt Albert DME Osterhaus and Malik Peiris
The emergence of SARS in humans by the spill-over of a previously unrecognized animal coronavirus (SARS CoV) (Chapter 11) and its subsequent dissemination over several continents with devastating consequences for both human health (Chapter 2) and the global economy (Chapter 22) has again underlined the continuing threat posed by newly emerging infectious diseases. Although the rapid and efficient response of the scientific community, co-ordinated by the World Health Organization (WHO) provided public health organizations with the appropriate tools to combat and contain SARS in record time (see Chapters 2, 7, 13), this may not be the case with future outbreaks of other newly emerging infections with a higher humanto-human transmissibility.
Predisposing factors for the emergence of virus infections The development of an impressive arsenal of effective antibiotics and vaccines in the 1950s, 1960s and 1970s and the eradication of smallpox by a worldwide campaign orchestrated by the WHO at the end of the twentieth century lulled many policymakers into believing that infectious diseases were a scourge of the past. However, the increasing number of infectious disease threats worldwide in the past few decades argue otherwise.1–3 These include the emergence of Nipah virus in Malaysia, avian flu (H5N1) in Asia, variant CJD in the United
Kingdom and hantavirus pulmonary syndrome in North America. Increased media coverage may explain part of this upsurge in global concern. It certainly has resulted in the need for better communication as part of the management of such emerging disease outbreaks. However, it is clear that many changes in modern society and lifestyle have facilitated the emergence and dissemination of virus and other infections among humans and animals alike. Changes in our social environment, agriculture, animal husbandry practices, trade, technology, travel and our impact on the global ecology and climate (Table 25.1) provide microbes new niches to exploit, more opportunities for crossing species barriers; and once the species barriers are breached, more opportunity for efficient spread.1–3 A mix of the above-mentioned changes played an important role in the emergence of SARS as a new human infectious disease arising from the animal world. Increased wealth and population density in the area where the infection emerged (Guangdong, China)4 led to an increased demand and consumption of the exotic wildlife species that carried the virus. The demand for such exotic foods combined with the cultural belief that meat and fish for human consumption has to be freshly killed (rather than frozen) led to the growth of large markets providing a wide diversity of live wild and farmed game animals. These markets provided an ideal opportunity for virus transmission and amplification within them and
249
250
Severe Acute Respiratory Syndrome
Table 25.1 Emerging virus infections in the last decennia, facilitated by changes in: Social environment Behaviour Mobility Demography Socio-economic status Public health measures
(Morals, taboos, iv drug abuse) (Transport systems, air travel) (Urbanization, population density) (Wealth, poverty) (Disinvestments, breakdown)
Technology Medical Food production
(Blood transfusion, transplantation, vaccination) (Increased scale, recycling)
Ecology Contacts with animals Agriculture Fisheries Environmental pollution Global warming
(Wildlife, domestic animals, zoo animals) (Deforestation) (Predator migration) (Immune suppression) (Impact on flora and fauna)
Viruses Mutation, recombination reassortment
(Host range, transmissibility, pathogenicity)
vastly increased the opportunity for interspecies transmission, including transmission to humans. Late recognition and inadequate implementation of public health measures at the early stages of the epidemic as well as extensive travel by land and air allowed the virus to spread rapidly beyond the region of its initial emergence. Genetic changes in the virus facilitated its adaptation to the new host. The virus found a new source of amplification within hospitals depleted of adequate isolation rooms for management of infectious diseases — a consequence of the perception that infectious disease was a problem of the past.
How do we prepare for future emerging infections? A mix of social, technological and ecological changes in our global society predisposes for the increased emergence of new infections, most of which originate from animal reservoirs.1–3 But it is often very difficult to influence or reverse some of these structural changes of modern society. Due to the complex nature of the factors and their interactions responsible for infectious
disease emergence, it is virtually impossible to predict when or where future emerging infectious threats will arise or which virus will be involved. Therefore, we have to invest more seriously in effective preparedness and response strategies. Currently the most serious threat of an outbreak with severe public health consequences is the emergence of another pandemic outbreak of influenza (Chapter 24) and the widespread outbreak of avian flu H5N1 in Asia and its associated threat to human health is an issue that clearly needs to be addressed. Although progress has been made towards better pandemic preparedness, it should be recognized that neither the world nor any individual country is yet adequately prepared for the next influenza pandemic. Preparedness for other emerging diseases (other than influenza and SARS) is clearly even more rudimentary in nature. Experiences from control efforts implemented for SARS and many of the other major outbreaks of past decades have shown that the key elements of a successful control strategy should be based on effective early warning systems for emerging human and animal infectious diseases
Lessons Learnt
using epidemiological and laboratory surveillance networks. Syndromic surveillance in humans and the rapid identification of newly emerging pathogen viruses are crucial in this regard. The timely sharing of information about unusual disease syndromes or disease patterns of humans and animals is vital. In this global village, a threat emerging anywhere is a potential threat everywhere. The international organizations that perform this function need support, both with resources and the willingness to share information. It is relevant to note that in the case of SARS, the Global Public Health Intelligence Network (GPHIN) picked up the first indications of unusual respiratory disease activity in Guangdong in November 2002 (Chapter 2). A salutary lesson from the experience of SARS was that hospitals can serve as amplifiers of disease transmission. This appears to be an even bigger problem with the practice of invasive ‘high-technology’ medicine. Good training in infection control practice (see Chapter 19), adequate isolation rooms and wards and availability of appropriate personal protective equipment to deal with patients with transmissible diseases are needs that have to be addressed. The lack of surge capacity for dealing with an infectious disease outbreak was apparent in many cities. This was not confined to cities hit hard by SARS; even in the absence of confirmed SARS or local transmission, the need to investigate and isolate large numbers of travellers with suspected SARS stretched the resources of many major cities with travel links with affected areas (Chapter 18). Generic plans for responding to an outbreak need to be devised in advance and tested in mock exercises on a regular basis. The experience of SARS demonstrated that when faced with a significant health threat and when provided with the leadership of the WHO, international collaboration between clinicians, microbiologists, epidemiologists and others does work effectively (Chapters 2 and 7). Indeed this was critical to the success in controlling SARS.
251
Even prior to establishing the nature of the aetiological agent responsible for an outbreak, epidemiological analysis can play a major role in containment of disease transmission. Techniques of mathematical modelling of disease outbreaks provides greater refinement and allows insights on the effectiveness of key interventions (Chapter 14). However, the key to these approaches and other strategies for outbreak control is the collection and analysis of relevant epidemiological data (e.g. risk factors, contact tracing) on a real-time basis. While this seems simple in concept, it is more challenging in implementation. However, this is a challenge that must be met inthe future. The ethical issues related to the application of measures such as quarantine (community containment of healthy contacts), contact investigation and confidentially, and travel advisories were highlighted in efforts to control SARS. These relate to the balance between individual freedoms and the common good and are issues that need to be discussed and debated well in advance of the next global infectious disease threat.5 The rapid identification of the aetiological agent allows targeted diagnosis and surveillance to be established upon which further control measures can be based. It also provides a better understanding of the modes of transmission and pathogenesis and permits development of specific antiviral drugs, therapeutic antibodies and vaccines. A renewed urgency for the search for endemic pathogens of humans (e.g. undiscovered pathogens causing endemic pneumonia, encephalitis, etc.) would be a worthwhile goal of itself.6 The development of methods for detecting novel pathogens causing endemic diseases will allow us to be better prepared when faced with an unknown microbial threat in future. Novel technologies for pathogen detection are increasingly available and need development. However, it is worth noting that classical virology techniques (culture, electron microscopy) provided the breakthrough in the search for the aetiology of SARS. Thus while
252
Severe Acute Respiratory Syndrome
novel techniques need to be explored, developed and exploited, conventional methodologies should not be neglected. It is striking that the attention refocused on coronaviruses since the advent of SARS has already led to the discovery of three novel coronaviruses — two in humans7–9 and one in animals.10 Since most novel emerging infections are zoonoses, close and effective collaboration between organizations responsible for human and veterinary health at the national and international levels is a prerequisite for better preparedness to meet future emerging infectious disease threats. It is important to obtain a better understanding of the microbial ecology of animals and birds most intimately associated with humans (e.g. livestock, pets, poultry and other animals sold for human consumption in live animal markets). This not only involves pathogens causing disease in these animals but also viruses that are carried by apparently healthy animals. Greater understanding of viruses carried by wildlife would also be a worthwhile goal. Nipah, Hendra and SARS CoV all have a wildlife reservoir. Ecological studies arising from concern with Nipah virus has led to the discovery of a number of novel viruses including Tioman, Menangle and Australian bat lyssavirus. Some of these are now known to cause disease in humans or in livestock.11 Three of four episodes of SARS re-emergence since the end of the global outbreak in mid-2003 was associated with laboratory infections. On one occasion, this was associated with secondary and tertiary transmission with fatal outcome in some patients. This community outbreak was effectively controlled. These laboratory infections occurred in biosafety level 3 or 4 laboratories. It is notable that there were no laboratory infections during the peak of the SARS outbreak in early 2003 although a number of diagnostic laboratories were handling SARS-infected specimens and carrying out diagnostic tests (including virus culture in some laboratories) for SARS CoV.
It appears that unsafe procedures were responsible for the infections in all three incidents. These laboratory incidents highlight the need for meticulous training and adherence to BSL-3 procedures (Chapter 9). SARS was the first new ‘plague’ of the twenty-first century and provided a dramatic illustration of the impact of a new emerging infectious disease. It had a profound effect on those afflicted, on health-care workers, on society in general, politicians and on the economy (Chapter 22). The reverberations of its impact still linger. It was fortunate that the disease had a number of characteristics that rendered it amenable to control by methods of classical contact training and patient isolation. The next major infectious disease crisis, whether it arises from nature (as is far more likely) or is man-made (through bioterror or bioerror), may be less amenable to control by these same means. For example, pandemic influenza will almost certainly not be controllable by case finding and isolation. Better investment in scientific research on all the aspects of infectious diseases will be required to provide the basis for effective strategies to combat future emerging infectious diseases. To echo what Joshua Lederberg said in the forward to this monograph, if SARS will lead to better levels of overall preparedness to counter emerging infectious disease threats of the future, all the pain and sacrifice associated with this incident may have not been in vain.
References 1 Osterhaus ADME. Catastrophes after crossing species barriers. Phil Trans R Soc Lond [Biol] 2001;356: 791–3. 2 Kuiken T, Fouchier R, Rimmelzwaan G et al. Emerging viral infections in a rapidly changing world. Curr Opin Biotechnol 2003;14: 641–6. 3 Smolinski MS, Hamburg MA, Lederberg J. Microbial threats to health: emergence, detection, and response. Institutes of Medicine of the National Academies: Washington, DC: National Academies Press, 2003.
Lessons Learnt
4 Zhong NS, Zheng BJ, Li YM et al. Epidemiology and cause of severe acute respiratory syndrome (SARS) in Guangdong, People’s Republic of China, in February, 2003. Lancet 2003;362: 1353–8. 5 Fidler DP. Germs, governance and global public health in the wake of SARS. J Clin Invest 2004;113: 799–804. 6 van den Hoogen BG, de Jong JC, Groen J et al. A newly discovered human pneumovirus isolated from young children with respiratory tract disease. Nat Med 2001;7: 719–24. 7 Van der Hoek L, Pyrc K, Jebbink MF et al. Identification of a new human coronavirus. Nat Med 2004;10: 368–73. 8 Fouchier RA, Hartwig NG, Bestebroer TM et al. A previously undescribed coronavirus associ-
253
ated with respiratory disease in humans. Proc Natl Acad Sci USA 2004;101: 6212–6. 9 Woo PC, Lau SK, Chau CM et al. Characterization and complete genome sequence of a novel coronavirus, coronavirus HKU1, from patients with pneumonia. J Virol 2005;79: 884–95. 10 Poon LL, Chu DKW, Chan KH et al. Identification of a novel coronavirus in bats. J Virol 2005;79: 2001–9. 11 Mackenzie JS, Field HF, Guyatt KJ. Managing emerging diseases borne by fruit bats (flying foxes), with particular reference to henipaviruses and Australlian bat lyssavirus. J Appl Microbiol 2003;94 Suppl: 59S–60S.
Appendix
Representative compounds with inhibitory activity against SARS CoV or other CoVs in vitro
Compound (reference)
Virus (cell type)
Inhibition assay
50% Inhibitory concentration/comment
SARS (FRhK-4) SARS (Vero) SARS (Vero-E6) SARS (Vero-E6)
CPE CPE CPE, virus yield CPE
SARS (Vero-E6) SARS (FRhK-4)
CPE CPE
50 mg/ml. Inhibition at 48 hr but not at 96 hr >1000 mg/ml Inactive at 2000 mg/ml No activity at non-cytotoxic concentrations (<200 mg/ml) >200 mg/ml at 48 hr 50–100 mg/ml at 72 hr
Amantadine (Leibowitz 1993)84
MHV (L2)
Plaque
(Chu 2004)45
SARS (FRhK-4)
CPE
(Tan 2004)54 Glycyrrhizin (Cinatl 2003)53 (Chen 2004)56 Pyrazofurin (Cinatl 2003)53 Interferon-a 2 (Stroher 2004)55
SARS (Vero-E6)
CPE
200 mM. Inhibition not based on early effect due to altered endosomal pH Not active at 4 times peak serum concentrations Inactive at 1000 mg/ml
SARS (Vero) SARS (FRhK-4)
CPE CPE
300–600 mg/ml. Selectivity >33 >400 mg/ml
SARS (Vero)
CPE
4.2 mg/ml. Selectivity index 12
SARS (Vero-E6)
1000–2000 IU/ml
SARS (Vero) SARS (Caco2)
CPE, virus yield, protein CPE CPE
SARS (Vero-E6)
CPE, virus yield
SARS (Vero) SARS (Caco2) SARS (FRhK-4)
CPE CPE CPE
SARS (FRhK-4) SARS (FRhK-4) SARS (Vero-E6)
CPE CPE CPE
Ribavirin (Chu 2004)45 (Cinatl 2003)53 (Stroher 2004)55 (Tan 2004)54 (Chen 2004)56
65
(Cinatl 2003)
Interferon-b (Hensley 2004)85 (Cinatl 2003)65 (Chen 2004)56 Lopinavir (Chu 2004)45 (Chen 2004)56
4950–6500 IU/ml 880–1530 IU/ml 50 IU/ml with pretreatment; 500 IU/ml at 1 hr post infection 95–105 IU/ml 9–21 IU/ml 625 IU/ml pretreatment; 2500–10 000 IU/ml with infection 4 mg/ml. Inhibition at 48 hr but not 96 hr 1–4 mg/ml at 48 hr 4–8 mg/ml at 48 hr
255
256
Appendix
Compound (reference)
Virus (cell type)
Inhibition assay
50% Inhibitory concentration/comment
(Wu 2004)46a b-D-N4hydroxycytidine (Barnard 2004)84c Rimantadine (Chen 2004)56 Chloroquine (Keyaerts 2004)84a Niclosamide (Wu 2004)84a
SARS (Ver-E6)
CPE
50 mm
SARS (Vero)
Virus yield
EC90 of 6 mM
SARS (FRhK-4)
CPE
8–16 mg/ml at 48 hr Selectivity 4–8
SARS (Ver-E6)
CPE
8.8 mM. Selectivity 30
SARS (Vero-E6)
1–3 mM
siRNAs to replicase 1A gene (He 2003)86 Baicalin (Chen et al 2004)56
SARS (FRhK-4)
Antigen, virus yield CPE, antigen, RNA
CPE CPE CPE
12.5–25 mg/ml at 48 hr 100 mg/ml at 48 hr Neutralizing activity at 0.37 nM
Human monoclonal IgG1 to S protein (Sui 2004)18
SARS (FRhK-4) SARS (Vero-E6) SARS (Vero-E6)
CPE, cytopathic effect; FRMK-4, fetal rhesus monkey kidney-4.
50–92.5% reduction in RNA copies depending on siRNA
Index
Page references to figures and tables are shown in bold active immune responses in pigs 94 acute interstitial pneumonitis 47 Acute Physiology and Chronic Health Evaluation (APCHE) scores 23 acute respiratory distress syndrome (ARDS) 37, 47, 192, 203 radiograph 36 aetiology 50–5, 111, 251 air travel role in containment 15, 18, 127, 154–7 role in spread 2, 3, 14, 100, 102–3, 119, 169 alanine aminotransferase (ALT) 26, 33 ALT see alanine aminotransferase alveolar proteinosis 47 amplification premises 9 anti-inflammatory role, IFN-a 195 antibodies IgG 94 serum neutralising 206 SIgA 93 antibody titres 187 antibody–mediated disease enhancement 187, 205, 206 antiretrovirals 196 antiviral agents 184–98 administration routes 187 aspartyl proteinase inhibitor 191 case series outcomes 193–4 cellular autophagy inhibitor 191 clinical development 6–7, 185–6 clinical trials, problems 185–6 post-exposure prophylaxis 186 immunomodulators 7, 188–9 glucocorticoids 188 inhibitory compounds in vitro 255 interferons 192–6, 255 affects on non-SARS CoVs 195 interaction with viruses 195 interferon-alpha (IFN-a) 186, 192, 195 interferon-beta (IFN-b) 186, 192 use in Chinese outbreak 195 manufacture lead time 242–3
p38 MARK inhibitor 188–9 preclinical testing 184–5 protease inhibitors 191, 196 aspartyl proteinase inhibitor 191 HIV protease inhibitors 184, 196 picornavirus 3C proteinase inhibitor 196 ribavirin 6–7, 39, 184, 188, 191–6, 196 target selection 189, 190, 190–1 tumour necrosis factor (TNF) inhibitors 189 antiviral resistance 189 ‘Asian flu’ see influenza A, Asian aspartyl proteinase inhibitor 191 barotrauma 23, 37 basic reproduction rate see case reproduction number (R0) ‘bird flu’ see influenza A, avian bovine CoV (BCoV) see group 2 bovine CoV (BCoV) bronchiole epithelial regeneration 74 bronchiolitis obliterans organizing pneumonia (BOOP) 22, 47 bronchoalveolar lavage (BAL) 76 Canada Ministry of Health 13, 15 Toronto outbreak 13, 100–1, 105, 240 cardio–respiratory fitness test 39 case fatality rate (CFR) 117, 117–18 case reproduction number (R0) 103, 120–5, 167, 239 see also effective reproduction number (Rt) case series outcomes, antiviral agents 193–4 cases, sex, age and outcome analysis 16–17, 103 cat, domestic (Felis domesticus) 82, 101 cattle calf diarrhoea 90–2, 91 haemorrhagic diarrhoea 90, 91 shipping fever 90–2, 91 winter dysentery 90, 91 causes of death 23 CDC see Shenzhen Centre for Disease Control
257
258
Index
cDNA clones 205 cell culture, virus isolation 52 cellular autophagy inhibitor 191 chicken infectious bronchitis virus (IBV) see group 3 infectious bronchitis virus (IBV) children clinical presentations 31, 32 contact infection of respiratory diseases 135 diagnosis 31–2 long term sequelae 34 manifestations 30–1, 37 outcome and prognosis 33–4 treatment protocol 32–3, 33 China Canton 1 economic effects (see costs, economic) epidemic pattern 120 Foshan 1 Guangdong province 1, 13, 50, 79–80, 81, 100–1, 119, 165 Guangzhou 1, 50, 105 Hong Kong (see Hong Kong Special Administrative Region (HKSAR)) information control 227, 230 Shenzhen animal market human workers immunity 80, 81, 101, 241 possible source mammals 80, 81, 241–2 Shenzhen Centre for Disease Control (CDC) 7, 21 chlamydia 3, 50–1 civet, Himalayan palm (Paguma larvata) 7, 62.80, 81, 101 coronavirus 101 clinical development, antiviral agents 185–6 clinical presentations 22, 31, 32 colon biopsy 25 computerised tomography CT scan contact infection of respiratory diseases 135 contacts, tracing and quarantining 127–9 coronavirus(es) animal vaccines 92–7 enteric CoV vaccines 93–5 FIPV vaccines 95, 205 respiratory CoV vaccines 95 in animals 72, 84–97 canine coronavirus (CCoV) 84 ‘classical’ human coronavirus diseases 134–5 cofactors in infection 88, 96 evolution 84–95 genetic groups 84, 85 haemagglutinin (HE) 84, 92 Himalayan palm civet CoV 101 human coronavirus 229E 134, 136 human coronavirus OC43 134 human enteric CoV (HECoV) 92 IFN-alpha treatment 93 life-cycle 189–90 long vs short spike (EM) 86 mouse hepatic virus (MHV) 189 pathogenesis 84–95
proteinase, crystal structure 191 tissues infected 86–8, 87 see also CoV groups; respiratory virus diseases, seasonality corticosteroids 184, 188 costs, economic 18–19, 129, 165, 213–27 cross border manufacturing, Hong Kong 223–6, 226, 227 exports and imports 213, 221–3, 224, 225 fiscal stimulus 227, 230 GDP, growth projections, East Asia 214, 214–15, 216, 217, 221, 222 influenza pandemic, future 244–5 information control, China 227 retail sales 215, 218 service industries 213–14 stock markets 226–7, 228–9 tourism and travel 217–19, 219–21 unemployment 221, 223 costs, social and political 18–19, 129, 165, 232 creatinine phosphokinase (CPK) 23, 26 cross-infection, control 48 cross-species transmission 90 CT scan thorax 22, 23 see also HRCT cytokines 28 detection and containment mechanisms, future 14–18 early identification 233–4 health care workers, infection clusters 233 lower respiratory tract, infection clusters 233 role of new technology 15–18 diagnosis in children 31–2 diarrhoea 5–6, 23–4, 25, 74, 105 diffuse alveolar damage (DAD) 73–4 DNA-N protein vaccination 185 DNA-S protein vaccination 185 ‘droplet and contact precautions’ 176–83 ‘droplet nuclei,’ virus transmission 135 effective reproduction number (Rt) 125 ELISA assay 5 emergence of virus infections predisposing factors 249–50, 250 preparing for the future 250–2 see also resurgence, future detection and containment measures enzyme immunoassay (EIA) 53, 68–9 eosinophilic pneumonitis 47 epidemic centres (epi-centres) 9 epidemiology 100–6, 129 Singapore 140 faeces, SARS CoV in 65, 74, 204 feline CoV see group 1 feline CoV femoral head, avascular necrosis 34, 40 fibrosis 38, 47, 73
Index
GDP, growth projections, East Asia 214, 214–15, 216, 217, 221, 222 gene sequence analysis 59 genome, SARS coronavirus 5, 7, 58–62, 111 open reading frame (ORF) 60 organisation 58–60, 59 variation 60–2, 204 Global Influenza Surveillance Network, WHO 13, 15, 50, 245 Global Outbreak Alert and Response Network, WHO (GOARN) 13, 15, 232, 241 global perspective 13–19 Global Public Health Intelligence Network (GPHIN) 13 GOARN see Global Outbreak Alert and Response Network GPHIN see Global Public Health Intelligence Network group 1 coronaviruses 84–90 antigenic relationships 90 cross-species transmission 90 group 1 feline CoV 89–90 feline enteric coronavirus (FECoV) 84, 89–90 feline infectious peritonitis virus (FIPV) 84, 89–90, 205 IFN-alpha treatment 92 incubation period 89 group 1 porcine CoV 84–9 porcine epidemic diarrhoea virus (PEDV) 84–6 porcine respiratory coronavirus (PRCV) 84–9, 205 clinical signs 88 corticosteroid treatment, effect on infection 89 infection route and dose 88 interaction with TGEV 88 mutations 88 respiratory bacterial co-infections 89 respiratory viral co-infections 88 vaccines 94–5 transmissible gastroenteritis virus (TGEV) 84, 86–8, 205 IFN-alpha treatment 92 mutations 86–8 vaccines 93–5 group 2 bovine CoV (BCoV) 90–2, 205 CoV disease syndromes in cattle 90, 91 cross-species transmission 92 vaccines, lack of 95 group 3 infectious bronchitis virus (IBV) chicken infectious bronchitis virus (IBV) 205 vaccines 95 haemolytic anaemia 192 Haemophilus influensae 2 hair thinning and shedding 34 health care workers future infection clusters 233 hospitals as infection amplifiers 176
259
role in early spread 2, 3, 9–10, 51 susceptibility 2, 3, 14, 17, 21, 37, 103, 119, 146, 172–3, 176–9 hepatitis B 26 ‘hidden’ SARS patients 27 high resolution computed tomography see HRCT histopathology 186–7 historical perspective 1–10 HIV protease inhibitors 184, 198 HKSAR see Hong Kong Special Administrative Region Hong Kong flu 8 Hong Kong Special Administrative Region (HKSAR) Amoy Gardens estate, sewage system 5, 5–6, 167 case reproduction number (R0) 167 community control elderly homes 167 home confinement 166 public education and risk communication 167 ‘step down isolation’ 167 contingency plans 166 cross border manufacturing 223–6 economic effects (see costs, economic) entry, screening and quarantine health declaration form 166 quarantine 166 temperature screening 166 epidemic pattern 1–10, 13, 105, 119, 120, 166 infection surge problems 168 information collation 166 initial outbreak 1–10, 13, 50–1, 100–1 Kwong Wah Hospital 2 Metropole Hotel 2 Prince of Wales Hospital 5 Princess Margaret Hospital 4 public health response 165–8, 240 role in prevention of epidemics 8–10 SARS Integrated Database (SARSID) 113 sequelae 165 surveillance systems 165–6 influenza 165 severe community-acquired pneumonia (SCAP) 165 see also Queen Mary’s Hospital, Hong Kong hospitals as infection amplifiers 176 host immunity 136–7 host immunomodulatory agents 184 HRCT 32, 37, 45–7 fibrosis 38, 47 ‘ground glass’ opacification 38, 46, 47 subpleural consolidation 46 human enteric CoV (HECoV) 92 humidity and virus survival 136 hypocalcaemia 192 hypomagnesaemia 192
260
Index
IFN-alpha treatment 93 IgG antibodies 94 immune hyperactive phase 28, 28 immune mechanism in mice 207, 208 immunoassay 4, 53 immunomodulators 188–9 inactivated viral vaccination 185 indirect immunofluorescence assay (IFA) 52–3, 68 infection control 176–83 Centers for Disease Control Guidelines 176 ‘droplet and contact precautions’ 176 general medical wards 177–9 infection control practices (ICPs) 177–9, 178 (see also Queen Mary’s Hospital, Hong Kong) infectious bronchitis virus (IBV) see group 3 infectious bronchitis virus (IBV) influenza A Asian 8 avian 3, 13, 50, 239 pandemic 1918 8 pandemic 1957 8 pandemic 1968 8 subtype H2N2 8 subtype H3N2 8, 50 subtype H5N1 3, 13, 50, 239, 242, 243 subtype H5N2 242 subtype H7N2 242 influenza B 13 influenza pandemic, future antiviral agents 242–3 manufacture lead time 242–3 influenza vs SARS 239–40 lessons from SARS 245 public health response 240 surveillance improvements animals 241–2 human 241, 245 vaccines 243–4 stockpiling 243–4, 246 influenza surveillance system, Hong Kong 165 interferon-alpha (IFN-a) 186, 192, 195 interferon-beta (IFN-b) 186, 196 internet, information dissemination via 232 ion channel inhibitors 242 isolation of patients 125–7
low incidence region, public health response 169–75 lower respiratory tract infection, clusters 233 Lowson, Dr James 1 lung damage by macrophages 6 lung parenchymal cells 72 lymphoid hyperplasia 76 lymphopenia 32, 76, 233 macaque, cynomologus (Macaca fascicularis) studies 72, 74–6, 185 vs human SARS 76 manifestations atypical 27–8, 146 cardiovascular 26 in children 30–1 enteric 23–4 haematological 24–6 hepatic 26 neurological 26–7 respiratory 21–3 media, information dissemination via 232, 236 medical supply stockpiling 163 methicillin resistant Staphylococcus aureus (MRSA) 10 methylprednisolone 36, 188, 192, 197 mice, intranasal infection 185 mitrogen-activated protein kinase (MAPK) p38 188 mortality rate, age specific 203 MRSA see methicillin resistant Staphylococcus aureus mucous membranes, importance in transmission 104 multinucleated giant cells 73 Mycoplasma pneumoniae 3 nasopharyngeal aspirates (NPA) 33, 64–6, 65, 197 neuraminidase inhibitors 185, 242 neutrophilia 24 northern blotting analysis 59 ‘one country–two systems,’ contribution to spread 1 open reading frames, predicted 60 outbreak surveillance system 162 oxygen saturation (SaO2) 44
journalists, early reporting of epidemics 1 Koch’s postulates 5, 53–4 laboratory-acquired infection biosafety 69, 103, 105, 231, 252 Singapore case 140 lactate dehydrogenase (LDH) 23, 32 leukopenia 32, 76 live attenuated vaccine, transmission risks 204 lopinivir 196
papain-like cysteine proteases (PLPs) 59 paramyxovirus 3, 52 passive transfer in mice 207, 208 pathogenesis 72–7, 186–7 pathology 72–7 patients children 30–4 elderly 26, 27 phases immune hyperactive 28, 28
Index
pulmonary destruction 28, 28 viral replicative 28, 28 picornavirus 3C proteinase inhibitor 196 plague 1 pneumocytes 72–6 pneumoenteric CoV infections 90–2 pneumomediastinum 23, 37 pneumonia, atypical see SARS epidemic pneumonia, SARS vs non-SARS 47–8 pneumothorax 23 porcine CoV see group 1 porcine CoV postmortem, cultural restrictions 72–3 preclinical antiviral testing 184–5 prodromal symptoms 21 public health response Hong Kong Special Administrative Region (HKSAR) 165–8 low incidence region 169–75 Singapore 139–68 United States of America (USA) 169–75 pulmonary destruction phase 28, 28 pulmonary function tests 34, 39 pulmonary haemorrhage 47 pulmonary oedema 47 pulmonary sequelae 36–9 pulse steroid therapy 36 Queen Mary’s Hospital, Hong Kong administrative leadership 179–80 communication via web-site 182–3 ‘Infection Control Link Nurse’ 179 infection control practices (ICPs) 179–83, 180 ‘basics’ in infection control 179–81, 181 intensive surveillance programme 181–2 logistics and staff welfare 182–3 low secondary rate in HCWs 176 microbial analysis 2 no nosocomial transmission 179 Protocol for Investigating a Single Suspected Case 182 staff education 182–3 ‘Wash hands, wear masks, control SARS’ 181 racoon dog (Nyctereutes procyonoides) 7, 62, 101 radiograph acute respiratory distress syndrome 36 ‘ground glass’ opacification 42, 43 pneumomediastinum 37, 44 subpleural consolidation 42, 43, 44 radiology 42–8 correlation with clinical parameters 44 diagnosis 42–4, 185–6 pneumonia, SARS vs non-SARS 47–8 Rattus norvegicus 1 Rattus rattus 1 reservoir concept 79 respiratory muscle weakness 39 respiratory virus diseases ‘classical’ coronavirus diseases 134–6 contact infection 135
261
‘droplet nuclei,’ virus transmission 135 human metapneumovirus (hMPV) 134, 187 influenza 131–2, 135, 136 seasonality in Hong Kong 132 seasonality in Netherlands 133 parainfluenza virus diseases 132–4, 135 respiratory adenovirus diseases 134 respiratory syncytial virus (RSV) 134 rhinovirus disease (common cold) 134 SARS 135 seasonality 131–7, 132 possible mechanisms 136–7 transmission routes 135–6 resurgence, future detection and containment measures 231–7 contact tracing and management 234 education 236 information technology, improving systems 236–7 international communication 236 isolation procedures 234 laboratory diagnostics 235 low level circulation 231 need for vigilance 137, 231 public health planning, broad-based approach 232 reverse genetic manufacture of vaccines 243–4 reverse transcription-polymerase chain reaction see RT-PCR ribavirin 6–7, 39, 184, 188, 191–6, 196 ritonavir 196 RT-PCR 3–4, 32–3, 53, 59, 64–7, 65, 66, 171, 204 rupintrivir 196 ‘SARS Alert’ Programme 176 SARS coronavirus (SARS CoV) animal reservoir 79–83, 101 causal model, putative 82 central nervous system 204 characteristics 23–6, 25, 55 confirmation criteria 69 cytopathic and immunopathic effects 203 ecology 82, 82–3 electron micrographs 75, 75–6 in faeces 5–6, 65, 74, 204, 231 genetic variation 60–2, 204 Initial detection and characterisation 4–8, 52–5.69 Koch’s postulates 5, 53–4 main protease (Mpro) 198 origin 79, 101–2 in other species 54–5, 204–5 papain-like protease 196 precursor 7–8, 79–83 respiratory shedding vs onset 119 serological test, development 82–3, 171 see also coronaviruses; genome; viral diagnosis SARS epidemic (pneumonia, atypical) case fatality rate (CFR) 117, 117–18
262
Index
cases, frequency distribution 112 cases, sex, age and outcome analysis 16–17, 103 containment 106, 127–8, 144, 233–7 differentiation from typical pneumonia 2 early spread 1–3, 50–0, 139–40 epidemic patterns 119, 120 hospital admission vs discharge time 116 hospital admission vs onset 114–16, 115, 116 hospital admission vs time of death 116–17, 117 incubation period 2, 103, 114, 115, 204 infectious period 103–4, 118 nosocomial transmission 2, 50, 101, 146, 179 pneumonia, SARS vs non-SARS 47–8 progress, global 102, 102–3 seasonality 135 second SARS outbreak averted 81 symptoms 1, 3, 6, 113–14, 233 ‘Will SARS return?’ (see resurgence, future detection and containment measures) see also transmission; transmission dynamics SARS Integrated Database see transmission, SARS Integrated Database (SARSID) seasonality, respiratory virus diseases 131–7, 132 secondary infection 103, 118, 147 see also case reproductive number sequelae adult 36–40, 165 children 34 musculoskeletal 40 neurological and psychiatric 39–40 pulmonary 36–9, 48 ribavirin, adverse reactions 39 serum neutralising antibodies 206 sewage system, role in early spread 5, 5–6, 167 Shenzhen Centre for Disease Control (CDC) 7, 21 SIgA antibodies 93 Singapore case analysis 142 containment strategies 144–61 broad case definition 149, 158, 162 command, control and co-ordination strategies 157–8, 159, 162 designation of SARS hospitals 152 hospitals, key containment measures 150–1, 150–3, 162–3 information collation 158, 159 ‘prevent–detect–isolate–contain’ 146, 149 epidemic patterns 100–1, 120, 140–5, 141, 156 fatality analysis 143 index cases 2, 139–40 infection clusters 141 infection surge problems 158–60 market outbreak containment 153–4 medical supply stockpiling 163 outbreak phases 144–5, 145, 148
outbreak surveillance system 162 ports, screening and quarantine 154–7, 163 Health Declaration Card 154 home quarantine 154 temperature screening 154–7 thermal scanners 154 public health response 139–63, 155 public reactions 160–1 superspreader cases 146–9, 147, 149 three level response system 161, 161–2 transmission pattern 144, 146 Spanish influenza A pandemic 8 spike glycoproteins 7, 86, 189–91, 204, 206 spleen intrafollicular hyalinosis 76 Staphylococcus aureus 2, 10 methicillin resistant (MRSA) 10 Streptococcus pneumoniae 2 superimposed infections 29 superspreading of SARS 2, 106, 125, 129, 146, 147, 149 surge capacity 158–60, 168, 245 T-cell depletion 208 T-lymphocyte counts 24 Taiwan 101, 105, 120 three level response system 161, 161–2 thrombocytopenia 24–5, 32 thromboembolism 25–6 tourism, effects on 19 transmissible gastroenteritis virus (TGEV) 84 transmission 24, 30, 96, 104–6 Singapore 144 transmission dynamics 111–30 case definition 113–14 case fatality rate (CFR) 117, 117–18 contacts, tracing and quarantining 127–9 epidemic patterns 119, 120, 127–9, 140–1, 141 intervention strategies and impact 125–7, 126 mathematical models 119–24, 122, 124, 128–9 patient database 112–13, 128 SARS Integrated Database (SARSID) 113 transmission routes 118–19, 135–6, 146 WHO case definition criteria 113 treatment protocol, children 33 tri-phasic nature 28, 28–9 tumour necrosis factor (TNF) 189 turkey diarrhoea 92, 93 United States of America (USA) air travel, role in spread 169 Centers for Disease Control and Prevention (CDC) international staff deployment 170, 170–2 web-site guidelines 172, 173 communication 172–4 laboratory response 170–1 low incidence 169 public health response 169–75
Index
surveillance 171 broad case definition 171 health alert notices 171 Urbani, Dr Carlo 2, 51, 100 urine, SARS CoV 65, 74 vaccines 203–10 DNA protein 185, 206 immune response to 185, 206, 209 inactivated virus 185, 206 live attenuated 204, 206 transmission risks 204 reverse genetics manufacture 243–4 serum neutralising antibodies 206 vectored, spike protein 206 vaccines for animal CoVs 92–7 enteric CoV vaccines 93–5 FIPV vaccines 95 respiratory CoV vaccines 95 ventilator induced lung damage 76 Vietnam 2, 3, 13, 100, 105 viral clearance gastrointestinal tract 203–4 respiratory tract 203 viral diagnosis 64–70 antibody detection 68 biosafety 69 confirmation criteria for SARS CoV 69 culture 67–8 enzyme immunoassay (EIA) 53, 68–9 histopathological course 186–7 inclusion bodies 72–3 indirect immunofluorescence assay 52–3, 68
263
serum antigen 67 virus neutralization test (NT) 69 see also RT-PCR viral replicative phase 28, 28 virus infections, emergence predisposing factors 249–50, 250 preparing for the future 250–2 see also resurgence, future detection and containment measures virus survival and humidity 136 wet markets 9 wild animal trade 81–3, 101 wildlife reservoirs 79–83, 96, 101 ‘Will SARS return?’ 137 World Health Assembly 245 World Health Organisation (WHO) case definition criteria 21, 31–2, 113, 140 containment declaration 14, 102, 111, 227, 232 contingency plans 240, 245 global alert(s) 13, 14, 18, 51, 100, 102, 106, 148, 227 Global Influenza Surveillance Network 13, 15, 50–2, 245 Global Outbreak Alert and Response Network (GOARN) 13, 15, 232, 241 initial responses 2, 13–18 international response coordination 129, 167–8, 169–71, 251 Yersinia pestis 1